050-2792vietnam vol2.qxd 8/24/04 9:37 AM Page 1 Vietnam Reading and Mathematics Assessment Study Preface and acknowledgement In 2000, the Vietnam Ministry of Education and Training (MoET) launched a large scale monitoring study of primary education. The study tested pupils and their teachers in the last grade of primary education (Grade 5), using a sample cross-sectoral survey in two key subject areas, reading comprehension in Vietnamese and mathematics. This volume is the second of three. Each volume contains different aspects of the study carried out in Vietnam in 2001. Volume 1 presents the study, summarizes the main issues and proposes policy recommendations. Volume 2 presents detailed findings of the study structured around ten chapters: Chapter 1 Setting for the Study Chapter 2 What were the levels of achievement of Grade 5 pupils in Reading and Mathematics? Chapter 3 What were the baseline data on Grade 5 pupils? Chapter 4 What were the characteristics of Grade 5 teachers and their classrooms? Chapter 5 What were the characteristics of School Heads and schools? Chapter 6 How do the conditions of schooling in Vietnam compare with the Ministry's own benchmark standards? Chapter 7 How equitably allocated were educational inputs to primary schools? How equitably distributed was pupil achievement between schools? Chapter 8 Effective primary schools in Vietnam Chapter 9 Initial explorations of relationships among independent variables Chapter 10 Summary and Conclusions Volume 3 provides full technical details of the design and conduct of the study. The study was conducted by the MoET in Vietnam. Many people were involved in the study. We would like to acknowledge our sincere thanks to the following people: the late Vice-Minister Le Vu Hung, who provided the oversight and guidance for the study; Vice-Minister Dang Huynh Mai responsible for primary education; senior directors and heads of departments of MoET, National Education Institutes, which provided specialists and researchers to conduct this study; provincial and district education offices, whose staff collected the data from over 3600 schools throughout Vietnam in an exemplary fashion; Dr. Nguyen Quoc Chi, National Manager of the Primary Education Project, who was instrumental in conducting the study and who provided special insights into the problems of primary education; Professor Dang Ba Lam, Director of the National Institute of Educational 1 Vietnam Reading and Mathematics Assessment Study Development (NIED), who provided a member of the data team; Professor Tran Kieu, Director of the National Institute of Educational Science (NIES), who furnished researchers for the questionnaire committee, and also Professor Do Dinh Hoan and Dr. Do Cong Vinh who worked on the study. Professor Hoan and his team were responsible for test development. Dr. Vinh organized the data editing, entry, cleaning and analysis. Members of the international support team included Dr. Kenneth Ross, Dr. Mioko Saito and Stephanie Dolata from IIEP/UNESCO; Professor Patrick Griffin of the Assessment Research Center of Melbourne University; Miyako Ikeda who guided the data entry, cleaning and merging of files for nearly 75,000 records and who conducted many of the subsequent statistical analyses, and finally the principal architect of the study who guided and helped the international advisory and the national researchers, Emeritus Professor Neville Postlewaite of Hamburg University. The report writing team would like to acknowledge detailed and thoughtful comments by Vincent Greaney (Lead Education Specialist), Marlaine Lockheed (Lead Education Specialist) and Luis Benveniste (Senior Education Specialist) from the World Bank and Richard Wolf (Professor and Chairman, Department of Measurement, Evaluation and Statistics, Teachers College, Columbia University, USA) and Al Beaton (Professor in Education, Boston College, MA, USA). Finally, we would like to thank the British Department for International Development and the Canadian Agency for International Development for their strong and generous support to the undertaking of this study. Christopher Shaw and Mai Thi Thanh World Bank in Vietnam August 2004 2 Vietnam Reading and Mathematics Assessment Study Chapter 1 THE SETTING FOR THE STUDY Introduction y the beginning of the 21st century, Vietnam was richly endowed B Vietnam's population in with human and natural resources but had been constrained in its 1999 was 76 million of development by almost a century of colonialism and thirty years of which 9 million were devastating war. The population numbered about 76 million and members of ethnic the population growth rate was 1.7 percent in the period 1989 to 1999. Of the minorities. 76 million persons, about 34 percent was under the age of 16. Seventy-six percent of the population lived in rural areas. Most people lived in the delta areas of the Mekong and Red rivers and along the long coastline. There were more than fifty ethnic groups (comprising about 9 million people) but most lived in the mountainous areas of the central plateau and in the northern part of the country. Vietnam is one of the world's top rice exporters. There are large reserves of Education viewed as coal, bauxite, and gemstones. Petroleum and natural gas have been found off link to economic the coast. Upon re-unification in 1975, Vietnam's initial development model development. was based upon a command economy and a highly centralised administration. By the year 2000, there was a great deal of decentralisation, and a state-managed market economy with a socialist orientation had been established. The primary political goal of the government is economic growth and development through transition to a market economy. Education and the economy are viewed as linked in Vietnam and consequently a high priority is placed on education. The literacy rate at the last count was said to be 91 percent for all persons over the age of ten-years. Only 9.8 percent of those over five years of age had never attended school. There are five years of primary school which are compulsory and children School system 5-4-3. begin school in September in the calendar year in which they become six. After primary school, there are four years of lower secondary school, three years of upper secondary school, and two to eight years of tertiary or professional education. The survival rate in primary school was 68 percent at the end of the1990s and the transition rate from Grade 5 primary to lower secondary school was 98 percent. 3 Vietnam Reading and Mathematics Assessment Study This report is concerned with primary education and more particularly with the last grade of primary education, namely Grade 5. This is the last grade of mass education and it was important to have information on the conditions of schooling as well as the achievement of pupils in the two key subjects of reading comprehension in Vietnamese and mathematics. The rest of this chapter therefore concentrates on the primary school. Primary education 70-80% government There are nearly 15,000 primary schools enrolling over 10 million pupils budget allocated to with about 1.9 million in Grade5. In 1999, the net enrolment ratio for primary education, of which 36% school was 88.5 percent but this average masked the fact that in the large to primary education. cities there was a 95.2 percent enrolment and in some mountainous areas it But, 90% paid from was only 56.5 percent. An effort is underway to achieve 100 percent local budget. enrolment at the primary school level in accordance with `Education for All' (EFA). The allotment of the government's budget to education was 17.4 percent in 1998. This was the equivalent of 3.5 percent of GDP. Of the MOET (Ministry of Education and Training) budget 36.4 percent went to primary education. About 90 percent of the total primary education is paid for from the local budget. There are 61 provinces and it is money levied at the provincial and school levels that finances much of education. Although the central Ministry sets the overall policy for primary education, including the development of a new curriculum, it is the provincial offices (and sometimes district offices) that control the allocation of teachers and facilities to schools. There was thought to be considerable variation in school and classroom resources not only among schools but also among provinces. One objective of the Grade 5 sample survey was to identify the extent of the inequity among schools and provinces in the conditions of schooling. Teachers poorly paid. Primary school teachers were paid about 50 US dollars a month but in some Teachers received schools, given extra allowances, this can be as much as USD130. This is low different forms of compared with the teacher salaries in several other Asian countries. Many teacher training in teachers teach extra classes or second shifts or have other forms of different eras. employment in order to supplement their incomes. In 1999, some 41 percent of primary teachers had not received the full pre-service teacher training of two years after graduation from secondary school. Forty-seven percent of all primary teachers had received the `standard' teacher training consisting of 12 years of general education and two years of pre-service teacher training. The remaining 12 percent of teachers had 12 years of general education and three or more years of pre-service teacher training. Again, as will be seen later, some of these figures had changed by 2001. Average pupil receives Three kinds of curriculum have been in operation for 25 years. The first 660 instruction hours consists of a 165 weeks full curriculum across the five years of primary per year. school and this requires 165 days in school per year. The second was a 120 4 Vietnam Reading and Mathematics Assessment Study weeks curriculum for ethnic minorities but this has now been discontinued. The third lasts 100 weeks and this is aimed at out-of-school youth wishing to attend primary school. The National Institute for Educational Sciences (NIES ­ an institute within the Ministry of Education) has been developing a new curriculum and as from 2001 it has begun to be gradually introduced into 9 compulsory subjects at the schools. In 2000, the average number of hours of instruction per day for Grade 5 most pupils was four hours. With 33 weeks per year this only allowed 660 hours of instruction per year. This means that there is no place for an overloaded curriculum. And yet, at the same time, there is more and more demand to increase the content of the curriculum. Exactly how the new curriculum will turn out remains to be seen. Currently there are the following compulsory subjects in the Grade 5 primary school curriculum: Vietnamese language, mathematics, moral education, nature and society, technology/crafts, health education, physical education, music, and art. In some schools, pupils have an option to study informatics and English as a foreign language. The major objectives for primary education in the year 2000 included achieving education for all, having all teachers receive thorough in-service training programs, improving the classroom and school resources, extending the school day so as to have more time to cover the curriculum, and ensuring a quality education for all. Major research questions posed by the Ministry of Education Since this study was the first national survey of educational achievement Ministry officials ever undertaken in Vietnam, the focus was on the questions that were of most decided on policy interest to the senior officials of the Ministry of Education and Training research questions for (MOET). A group of 36 senior members of the Ministry held discussions ­ study. both plenary and group ­ and decided on the research questions to be used for the survey. Selected general and specific questions have been listed in `Appendix 1.1: Policy research questions'. The research study was concerned only with the policy questions posed by the ministry. All dummy tables and instruments were based on the research questions. The policy questions were developed at the end of 1999. The data for the study were collected in April, 2001. It is clear that several developments took place subsequently within the ministry and the system of education. The current study has dealt with the originally posed policy questions. As a summary the major questions have been listed below. 1. What was the level of achievement of Grade 5 pupils overall and in the various domains of reading and mathematics? What was the level of Grade 5 teachers in reading and mathematics? 5 Vietnam Reading and Mathematics Assessment Study a) What percentages of pupils reached the different skill levels in reading and mathematics? b) What percentages of pupils reached benchmark levels in reading and mathematics? c) What were the total scores by region and province? d) What were the differences in achievement between: i) pupils on isolated, rural, and urban schools; ii) boys and girls; and iii) different socio-economic groups? e) Were the pupils `elite' (upper 5%) performances similar in different regions and socio-economic groups? To What extent did the performance `tails' (bottom 5%) differ across regions and socio- economic groups? f) What were the relationships between teachers and pupils performance on the tests of reading and mathematics? 2. Policy questions related to educational inputs a) What were the characteristics of Grade 5 pupils? b) What were the characteristics of Grade 5 teachers? c) What were the teaching conditions in Grade 5 classrooms and in primary schools? d) What aspects of the teaching function designed to improve the quality of education were in place? e) What was the general condition of school buildings? f) What level of access did pupils have to textbooks and library books? 3. Specific questions relating to a comparison of reality in the schools and the benchmarks set by the MOET and the Fundamental School Quality Levels Were the following benchmarks met? (total school enrolment, class size, classroom space, staffing ratio, sitting places, writing places, chalkboard, classroom furniture, classroom supplies, academic qualification of school heads, professional qualification of school heads, etc.) 4. Have the educational inputs to schools been allocated in an equitable fashion? a) What was the equity of material resource inputs among regions, among provinces and among schools within provinces? b) What was the equity of human resource inputs among provinces and among schools within provinces? c) How different was pupil achievement among regions, among provinces, and among schools within provinces? 5. Which were the variables most associated with the difference between the most effective and least effective schools? 6. What variables were most associated with achievement? The main focus of the study is, however, on the achievement of the pupils and 6 Vietnam Reading and Mathematics Assessment Study in all ensuing chapters in this book relationships will be made to achievement. Furthermore, it is the achievement chapter that has been presented first in this report and the other chapters follow in sequence. The design of the study A decision was taken to conduct a cross-sectional survey of all pupils in Tests covered both the Grade 5 in April 2001 in the whole country. At the same time it was old and new curricula. considered that since this was the first survey, the Ministry of Education and Training might well wish to conduct repeat surveys and that this survey might well be the benchmark data against which future comparisons will be made. The perceived future did have an influence on the design of this study because, for example, when the tests were developed account was also taken of the new curriculum to have sufficient anchor items between the two curricula so that comparisons will be able to be made at a future time. A second decision was taken that the survey would be a sample survey but Sampling needed to be that the sample should be drawn in such a way that the standard errors of very good. sampling would be very small for national estimates and small for provincial estimates. A third decision was taken to select two aspects of learning for the study: Reading and reading comprehension and mathematics. Mathematics selected as achievement focus. In general, the design of the study followed from the questions posed by the Ministry. There were pupil tests and a questionnaire as well as a teacher test and questionnaire and then a school head questionnaire. The sampling therefore was a sample of pupils and two Grade 5 teachers were selected from all grade 5 teachers in the school. For the most part the analyses were undertaken using the pupil as the unit of analysis. Sampling The size of a sample is a function of the amount of differences among schools and the accuracy of estimates that one wishes to have. The more there are differences among schools the more schools that have to be included in the sample to be sure to cover all of the variation. The more accuracy that one wishes to have the greater the number of schools and pupils that one needs in the sample. For the country as a whole it was decided to aim for great accuracy and certainly not more than 2.5 percent for one standard error of sampling for a percentage (or 0.1 of a standard deviation for a mean). For a province it was decided to have as good accuracy as one could with the money and resources available. 7 Vietnam Reading and Mathematics Assessment Study Sample to have 60 It was assumed that the intraclass correlation (a measure of the extent to schools per province and which pupils within schools were like each other) at the province level would 20 Grade 5 pupils per be not less than 0.30. To achieve standard errors of sampling of not more than school. 7.5 percent for a percentage (at the 95 percent confidence limit), this meant having 60 schools per province. With 61 provinces this meant 3660 schools. With a random sample of 20 pupils per school this meant 73,200 pupils. Clearly this was to be a very large survey. If an intraclass correlation of 0.50 was assumed, this would mean having 77 schools per province and therefore 1880 pupils per province and 114,680 pupils altogether. There was some small amount of evidence available from a previous study undertaken in five provinces that for those provinces the intraclass correlation was about 0.4. But since there had only been 14 schools in each province in the study, then the estimate was unstable. It was therefore agreed to `go with' 60 schools per province and hope that a province error would not exceed 10 percent. With 3660 schools in the sample it could be safely assumed that the national error would be very small. A school was defined as a school on the school census form. This was usually a school building consisting of a main school. In some cases however, there were main schools plus up to ten satellite campuses. But in this latter case the school plus campuses was regarded as one school. The sampling frame for all primary schools was taken from the 1999 special school census conducted by the Australian Council for Educational Research (ACER). The way in which the sample was drawn and the ensuing calculations and post-stratification have been reported in Volume 3 of this series and will not be given here. It should be noted that without the ACER census it would have been impossible to have a sampling frame. The MOET's annual school census only contains aggregated data to the province level. Using the latest (March, 2001) figures available for the enrolment in each province, the weighting of the sample was undertaken by Dr. Kenneth N. Ross of the International Institute for Educational Planning (IIEP, UNESCO). Thus, sampling weights were available and all estimates produced in this book for pupils, teachers and schools are: Unit Percentage Mean Pupil % pupils Mean of pupils Teacher % pupils having teachers with Average pupils in Vietnam has a the referred to characteristics teacher with reported characteristics School % pupils are in school with Average pupil is in a school with the referred to characteristics the referred to characteristics For each estimate in the book there is an associated standard error of sampling. This will be further explained at the beginning of later chapters. These standard errors are needed to enable readers to construct confidence limits for sample estimates of population characteristics, and to make 8 Vietnam Reading and Mathematics Assessment Study appropriate comparisons among different sample estimates. Suffice it to say here that the sampling errors for the national Vietnamese estimates were very small (less than one percentage point for percentage estimates) but for provinces they were higher. The intraclass correlations between schools within provinces varied from 0.20 to 0.81 and most were around 0.60. This means that in future that if generalisations are to be made from provincial samples to provincial populations a large sample of schools will be required for most provinces. There was the further problem of teachers. In order to avoid too much The sample involved disruption in the schools when testing, it was decided to include only two only 2 Grade 5 teachers Grade 5 teachers in the sample. Thus not every pupil was associated with his per school. or her reading or math teacher. Two teachers were drawn at random from the list of Grade 5 teachers for each school drawn in the sample. This meant that only 50.1 percent of pupils were associated with a teacher. This was considered to be a large enough sample for teacher/pupil analyses for the whole country. There was some worry that the teacher half sample was not the same as the non-teacher half sample. In other words, there was fear that some bias might have occurred in the selection of teachers. From Table 1.1 it can be seen that the differences between the two half samples were small. The two half samples were similar and it is reasonable to assume that the half sample analysed to examine the teacher pupil link was a reasonable approximation to all Grade 5 pupils, when considering teacher-pupil links in the data. It can be seen that whereas the pupils in the total sample had a mean score of Pupil-teacher link made 500 for mathematics, the pupils in the `pupils with teacher' group had a mean for about half of the of 498.3 and the `pupils without teacher' group had a mean of 501.5. The pupils. This group standard deviations were similar. The reading scores were slightly higher for nearly identical with pupils in the without teacher group. The values for sex, age, total possessions, group of pupils not parental education, and grade repeating were similar between two groups. linked to teachers. The `pupils with teacher' group had more pupils from ethnic minorities, slightly more advantaged home background and were in schools with slightly fewer resources. But, in general, it was felt that two groups were similar enough that the half sample with teachers would be good enough to yield good estimates of the real situation in Grade 5 classrooms. Table 1.1: Comparison of two half samples of pupils: with teachers and without teachers Math score Reading Age Years Times score Sex in Total Home School possessions parent grade Background resources Pupil Mean Pupil Mean % girls months education repeat Pupil with teacher 498.3 497.6 48.0 134.6 9.5 14.9 0.2 -0.080 10.4 Standard deviation 101.1 101.1 11.4 2.7 6.7 0.5 1.048 4.8 Pupil without teacher 501.5 502.1 48.1 133.6 9.9 15.5 0.2 0.089 11.5 Standard deviation 99.0 99.0 10.3 2.8 6.7 0.5 0.935 4.8 Total 500.00 500.00 48.1 134.1 9.7 15.2 0.2 0.010 11.0 Standard deviation 100.00 100.00 10.8 2.8 6.7 0.5 0.993 4.8 9 Vietnam Reading and Mathematics Assessment Study It must be recalled however when dealing with the linkages between pupils and teachers that they are based on the half sample above. But for the general descriptive statistics for all teacher variables are based on the whole sample of two Grade 5 teachers per school. Instrument development Both tests and questionnaires were developed The tests were developed by members of the National Institute for Educational Science (NIES) with the guidance and assistance of Professor Patrick Griffin from the University of Melbourne. There were two pupil tests and two teacher tests to be developed. Full details of the test development have been given in Chapter 2 of Volume 3. The questionnaires were developed by a team from NIES. Pupil tests: 60 multiple Pupil tests: A blueprint was drawn up for the pupil reading test and a choice items for each separate one for the pupil math test. (see Appendices 1.2 and 1.3). These test. blueprints were based on the existing curricula and the changes foreseen over the next few years ­ on the pre- and post-2000 curricula (commonly known in Vietnam as the old and new curricula). The reason for including those changes made between the pre and post 2000 curricula was to ensure that the test would be valid for 2001 and also later years. It was agreed that the test should have about 60 questions. This was a function of the amount of time that could be made available for testing in the schools. The items were written in terms of the skills required and reading and math for the two tests. In reading the domains of narrative, expository texts and document reading and in mathematics the domains were number, measurement and space. About three times more items were written than were required for the final test. A specific decision was taken to use multiple choice items. Although multiple choice items had not been used widely in Vietnamese primary schools, there were various reasons for the decision. These included; with multiple choice format, more items could be included in the tests, they allowed objectivity of marking, they had been used with success in the five province study, there was sufficient evidence from other countries having used multiple choice for the first time that, with the careful administration of practice items before the test, all pupils quickly adapted to responding to multiple choice items. Most of the items had four options consisting of one right answer and three wrong ones. The disadvantage of this was that a pupil who did not know the answer still had a 25 percent chance of guessing the right answer. The advantage was that the only three good wrong answers had to be created. These items were trialled in the province of Thanh Hoa, an average province, and after the item statistics were reviewed a selection of 60 items 10 Vietnam Reading and Mathematics Assessment Study per test made. Domains of reading and mathematics: The three different major domains of reading included in the tests were: Three major aspects of 1. Narrative. Based on continuous texts in which the aim was to tell a reading to be tested. story, whether fact or fiction. 2. Expository. Based on continuous texts that were designed to describe, explain, or otherwise convey factual information or an opinion to the reader. 3. Document. Based on structured information presented in the form of tables, maps, graphs, lists, or sets of instructions. The pupils were requested to search, locate, and process selected facts rather than read every word of a continuous text. Two of these domains and the approach to testing were new to the Vietnamese curriculum for a number of reasons. First, reading is taught as part of an integrated approach to language instruction in which reading materials is used for discussion and for writing lessons. Assessment of language is also conducted with an integrated approach. In this study reading was assessed on its own. Second, most reading in Vietnam year 5 focuses on narrative texts. Expository and document texts are encountered routinely in the secondary curriculum. These domains are also commonly used in international studies of reading achievement. Expository and document texts are being introduced into the new primary curriculum. These domains were included in the tests for all these reasons. In mathematics, the major domains covered were: Three major aspects of 1. Number: Operations and number line; rounding and place value, mathematics to be significant figures, decimals, fractions, percentages and ratio tested. 2. Measurement: Measurement of distance, length, area, capacity, money and time. 3. Space and Data: The ability to understand and use geometric shapes and to interpret bar, pie and line graphs and tables presenting data describing common phenomena for Grade 5 pupils. Following the main testing the items were again reviewed using Rasch and classical analyses. These are two ways of building tests and scaling them. The Rasch approach has the advantage that not all items are given equal weight. It 11 Vietnam Reading and Mathematics Assessment Study is also possible to scale all items onto one scale and in this case it was also possible to scale the teacher items and the pupil items onto one scale thus allowing the teachers and pupils to be compared on the same scale. Four items were dropped from the pupil reading test but all 60 were retained for the math test. The four items were dropped from the reading test because they were found not to be measuring the same construct as the other fifty six items. Three different, but complementary ways of Test scores: Three types of scores were produced for each test. The first was analysing the test data a skill levels categorisation. This allows teachers and curriculum developers were developed. to know what the pupils in specific classes or in the country as a whole can do. The skills of individuals or groups can be analysed and decisions taken on further work for them. In this sense, it is a developmental framework. The second is a functional categorisation. It provides benchmarks about how the primary school system is functioning for primary children to enter society and to prepare pupils for studying independently in secondary school. The third is a scale score that is used for the comparison of achievement among groups and for other explanatory type analyses. Skill Levels: The first was the percentage of pupils (or teachers) at a given skill level in reading or mathematics. The levels for reading were: Reading skill levels Level 1: Matches text at word or sentence level aided by pictures. Restricted to a limited range of vocabulary linked to pictures Level 2: Locates text expressed in short repetitive sentences and can deal with text unaided by pictures. Type of text is limited to short sentences and phrases with repetitive patterns. Level 3: Reads and understands longer passages. Can search backwards or forwards through text to for information. Understands paraphrasing. Expanding vocabulary enables understanding of sentences with some complex structure. Level 4: Links information from different parts of the text. Selects and connects text to derive and infer different possible meanings. Level 5: Links inferences and identifies an author's intention from information stated in different ways, in different text types and in documents where the message is not explicit. Level 6: Combines text with outside knowledge to infer various meanings, including hidden meanings. Identifies an author's purposes, attitudes, values, beliefs, motives, unstated assumptions and arguments. 12 Vietnam Reading and Mathematics Assessment Study Mathematics skill levels Level 1: Reads, writes and compares natural numbers, fractions and decimals. Uses single operations of +,-,x and : on simple whole numbers; works with simple measures such as time; recognises simple 3D shapes. Level 2: Converts fractions with denominator of 10 to decimals. Calculates with whole numbers using one operation (x,-,+ or : ) in a one- step word problem; recognises 2D and 3D shapes. Level 3: Identifies place value; determines the value of a simple number sentence; understands equivalent fractions; adds and subtracts simple fractions; carries out multiple operations in correct order; converts and estimates common and familiar measurement units in solving problems. Level 4: Reads, writes and compares larger numbers; solves problems involving calendars and currency, area and volume; uses charts and tables for estimation; solves inequalities; knowledge of angles in regular figures; understands simple transformations with 2D and 3D shapes. Level 5: Calculates with multiple and varied operations; recognises rules and patterns in number sequences; calculates the perimeter and area of irregular shapes; measurement of irregular objects; recognised transformed figures after reflection; solves problems with multiple operations involving measurement units, percentage and averages. Level 6: Problem solving with periods of time, length, area and volume; embedded and dependent number patterns; develops formulae; recognises 3D figures after rotation and reflection and embedded figures and right angles in irregular shapes, data from graphs and tables; These levels were hierarchical. Each of the skill levels is in the curriculum and reports were made on the percentages of pupils at each level of skill. That is to say, if 14 percent of pupils were at Level 3, they had reached Levels 1, 2, and 3 but had not reached 4, 5, and 6. Categories of pre-functional, functional (reading and mathematics) skills, and independent learning skills for lower secondary grades: Two panels of subject-matter specialists from the NIES decided on the probability of pupils answering each item correctly and these probabilities were then used to create points on a scale whereby pupils reaching a point were classified as pre function, functional (meaning that they had the reading and mathematics skills to enable them to cope in Vietnamese society) and finally independent 13 Vietnam Reading and Mathematics Assessment Study (meaning a standard at which they were deemed to have the reading and math skills needed for independent learning in Grade 6). These points have been described as bench marks in this report. Reading and mathematics 500 scores and sub-scores: A scale of achievement was used where a score of 500 was the mean for the country and the standard deviation was 100. These scores were produced for the overall scores in reading and mathematics and then for each sub-score: narrative, expository, documents, number, measurement, and space. Other measures Teacher tests in reading and mathematics were Teacher tests: There were two aims for the teacher tests. First they should developed. provide a good spread of teacher achievement in reading and mathematics. Second, they should have enough items in common with the pupil tests so that teacher and pupil scores and reading and mathematics skills could be brought onto one scale. Items were written and trialled. The resultant teacher tests had 45 items each. Questionnaires were developed by special Questionnaires: There were three questionnaires: pupil, teacher, and school committee. head. The policy questions developed by the Ministry of Education were taken and a list made of the information required for them. This list was then developed into questionnaire items by a questionnaire committee of fourteen persons, most of whom came from NIES. The questionnaires were finalised in English and the English version was translated into Vietnamese. It is normal to have a forward translation (in this case from English into Vietnamese) and also a back translation by another group of translators (in this case from Vietnamese into English) in order to be sure that the meaning of the translated version is the same as in English and that All measures were the difficulty of words used was about the same. Only the forward piloted. translation was made. Despite trialling the questionnaires in Thanh Hoa at the same time as the test items were trialled, a later back translation showed that there were some deviations in one or two questionnaire items from the original intention. The data emanating from some questions had to be dropped from the anticipated analyses. But, the data from all of the other questions could be used. The Conduct of the Study Data collection The number of schools in the sample was 3639 (of the 3660 schools some 14 Vietnam Reading and Mathematics Assessment Study had two groups of 20 pupils because they were very large and hence the reduction to 3639 schools) and it was estimated that each school should be visited for two days. The time allowed for the total data collection was four week(s) (from the first day of the training sessions for data collectors on March 19, 2001 to the last day of the arrival of the instruments by mail at the MOET on April 21). A data collection team headed by Dr. Do Cong Vinh of NIES was created. The preparation work was enormous. The pupil names in Grade 5 were collected by means of a Pupil Name Form from each of the schools selected in the sample. The sub-sample of 20 pupils per school was drawn at random by the local data collector following the instruction provided by the central team. Pupil booklets were prepared for each of the sampled pupils with the relevant ID information being written into the front page of the booklet. The booklet consisted of a Pupil Questionnaire followed by a mathematics test and then a reading test. In a similar fashion the names of the mathematics and reading teachers of Grade 5 pupils were known from the School Form and a sub-sample of two mathematics and reading teachers was drawn for each school. Again, Teacher booklets were prepared for each of the sampled teachers. A Data Collection Manual was prepared including information on every step to be taken by the data collectors. 4405 persons involved in In all there were 4,405 persons involved in the data collection. Each sample data collection. school needed one data collector. This person was selected from the local district educational officers; each district having a school in the sample selected a coordinator. Each province formed a small committee of three persons to organize the data collection within a province. The sampled schools were known about three months before the data had to be collected. The district and province in which the sampled schools were located were also known. Contact was made with the Provincial Directors of Education and discussions were held to arrange for the DEOs (District Education Officers) in which the schools were located to be available for the data collection. Data collectors were The packages were sent to the Provincial Education offices two weeks trained at national and before the date of the data collection. Three training courses were held, provincial levels. the first for those central officers (14-15/3/2001), who were to train the provincial officers in the second training course (27-29/3 for the northern provinces and 2-4/4 for the southern provinces). The third training course consisted of sixty-one different provincial training courses to train the data collect within each province. In all of the training courses the participants completed all instruments themselves and were also trained in what they had to do from the moment they arrived at schools for the testing to the point when they departed the 15 Vietnam Reading and Mathematics Assessment Study schools with the completed instruments. The data collector for a particular school was given the package for that school where he or she had to collect data. If a pupil or teacher was absent on the day of testing then no replacements were allowed, but a red line had to be drawn through the name. If a pupil or teacher was not present for any one testing session, then a comment to this effect had to be made on the Pupil and Teacher Name forms. First day in school devoted to organising On the first day, the data collector selected a random sample of 20 pupils per the testing. The second school and, at the same time, selected a random sample of two Grade 5 day was devoted to the teachers. The data collector also prepared the `Pupil Name Form', `School actual testing. Form', and `Teacher Name Form'. Arrangements were made for rooms to be available for the testing of both pupils and teachers. The data collector filled in the ID numbers on each of the booklets at the end of the first day. On the second day in the morning the pupil questionnaire, pupil reading test, pupil math test, and school head questionnaire were completed. In the afternoon the teacher questionnaire, teacher reading test, and teacher math test were completed. In three schools there were two pupils in each of them who were present for the questionnaire and math test but not for the reading test. In one school only a teacher was absent. All test instruments were returned to the district The data collectors handed in the completed packages for their schools to the offices and then to District Education Office, that in turn sent them to the National Centre. The Hanoi. informal feedback from the data collectors was that the pupils and teachers all enjoyed taking the tests. Two teachers complained that the Teacher questionnaire was too long but they had completed it. Data editing, entry, and cleaning Part of a Hanoi Educational Management College was used as the central place to which all completed instruments had to be returned. There had to be a lot of storage space so that the instruments from each school could be stored by province in such a way that they could be accessed very quickly when required. Data editing and entry was completed in 12 As soon as the data collection finished three teams (editing, entering, and weeks. cleaning) of fourteen persons were formed for the data-entry. They were trained in the use of the Windows version of the Data Entry Manager (WINDEM) computer software developed by the IEA Data-Processing Center in Hamburg, Germany (IEA, 2000). This software was adapted specifically for the entry of the Vietnam Grade 5 data and no problems were 16 Vietnam Reading and Mathematics Assessment Study encountered in the installation and use of this software. Cleaning of data Once the data-collection instruments were returned to the National Centre, completed in 3 months. they were checked (edited) to ensure that the instruments for each pupil, each teacher, and each school head were there and that there were no missing data for the factual questions on the questionnaires. There were some missing instruments and data and several school heads had to be contacted for the missing instruments and data to be provided. The data entry took 12 weeks. Just over three percent of the data were entered twice, the second time for verification purposes. No major problems were encountered and at the end of 14 weeks all data had been entered. However, near the beginning of the data entry phase a `virtual' workshop was held for three days where the satellite classroom in the World Bank was connected to the satellite classroom in the World Bank building in Paris. At the Hanoi end was the Vietnam team of data enterers and at the Paris end there was the IIEP team that was training the Vietnam data enterers. This type of teaching/learning proved to be very effective. At the same time as the data were entered, a further data cleaning team of six persons checked the data for accuracy and reasonableness. They conducted a series of consistency checks in order to correct some data and in other cases to contact schools in order to check some of the information that had been filled in on questionnaires and where the information looked suspect. The names of the peoples working on data editing, entry, and cleaning have been presented in Appendix 1.4. Further cleaning, file merging and weighting The data were sent to the IIEP team in Paris (Ken Ross, Mioko Saito, and Stephanie Leite) by Miyako Ikeda who had been working with the Vietnam data entry and checking team. Further consistency checks were undertaken and problems resolved as a result of continuous e-mailing between Paris and Ha Noi. The data files were merged to the pupil level and the sampling weighting undertaken to correct for disproportionality between strata. The overall final The numbers of pupils teachers and schools in the data file have been response rate was 99.15%. presented in Table 1.2 together with the response rates. It is to be noted that final response rates were 99.78 percent for school heads, 98.63 percent for 17 Vietnam Reading and Mathematics Assessment Study Table 1.2: The planned and achieved samples of schools and pupils Schools Pupils Teachers School heads Province Planned Achieved % % % response Planned Achieved response Planned Achieved response Planned Achieved % response Ha Noi 60 60 100 1200 1191 99.25 120 120 100 60 60 100 Hai Phong 60 60 100 1200 1193 99.42 120 119 99.17 60 60 100 Ha Tay 60 60 100 1200 1196 99.67 120 120 100 60 60 100 Hai Duong 60 60 100 1200 1197 99.75 120 120 100 60 60 100 Hung Yen 60 60 100 1200 1194 99.5 120 117 97.5 60 60 100 Ha Nam 60 60 100 1200 1200 100 120 120 100 60 60 100 Nam Dinh 60 60 100 1200 1200 100 120 120 100 60 60 100 Thai Binh 60 60 100 1200 1200 100 120 119 99.17 60 60 100 Ninh Binh 60 60 100 1200 1182 98.5 120 119 99.17 60 60 100 Ha Giang 59 58 98.31 1200 1139 94.92 118 108 91.53 59 59 100 Cao Bang 59 59 100 1200 1194 99.5 118 107 90.68 59 59 100 Lao Cai 60 58 96.67 1200 1141 95.08 120 112 93.33 60 58 96.67 Bac Kan 60 60 100 1200 1191 99.25 120 118 98.33 60 60 100 Lang Son 60 60 100 1200 1196 99.67 120 118 98.33 60 60 100 Tuyen Quang 60 60 100 1200 1194 99.5 120 118 98.33 60 60 100 Yen Bai 60 60 100 1200 1183 98.58 120 115 95.83 60 60 100 Thai Nguyen 60 60 100 1200 1188 99 120 118 98.33 60 60 100 Phu Tho 60 60 100 1200 1199 99.92 120 120 100 60 60 100 Vinh Phuc 60 60 100 1200 1200 100 120 120 100 60 60 100 Bac Giang 60 60 100 1200 1195 99.58 120 120 100 60 60 100 Bac Ninh 60 60 100 1200 1199 99.92 120 120 100 60 60 100 Quang Ninh 60 59 98.33 1200 1167 97.25 120 116 96.67 60 59 98.33 Lai Chau 58 58 100 1200 1190 99.17 116 111 95.69 58 58 100 Son La 60 60 100 1200 1162 96.83 120 117 97.5 60 59 98.33 Hoa Binh 60 60 100 1200 1194 99.5 120 119 99.17 60 60 100 Thanh Hoa 60 60 100 1200 1195 99.58 120 120 100 60 60 100 Nghe An 60 60 100 1200 1186 98.83 120 119 99.17 60 60 100 Ha Tinh 60 60 100 1200 1200 100 120 120 100 60 59 98.33 Quang Binh 60 60 100 1200 1197 99.75 120 119 99.17 60 60 100 Quang Tri 59 59 100 1200 1193 99.42 118 116 98.31 59 59 100 Thua Thien - Hue 60 60 100 1200 1199 99.92 120 119 99.17 60 60 100 Da Nang 53 53 100 1200 1195 99.58 106 106 100 53 53 100 Quang Nam 60 60 100 1200 1192 99.33 120 119 99.17 60 60 100 Quang Ngai 60 60 100 1200 1197 99.75 120 119 99.17 60 59 98.33 Binh Dinh 60 60 100 1200 1198 99.83 120 120 100 60 60 100 Phu Yen 60 60 100 1200 1195 99.58 120 120 100 60 60 100 Khanh Hoa 60 60 100 1200 1187 98.92 120 118 98.33 60 60 100 Kon Tum 56 56 100 1200 1191 99.25 112 108 96.43 56 56 100 Gia Lai 60 60 100 1200 1192 99.33 120 114 95 60 60 100 Dak Lak 60 60 100 1200 1191 99.25 120 119 99.17 60 60 100 Ho Chi Minh 60 60 100 1200 1194 99.5 120 120 100 60 60 100 Lam Dong 60 60 100 1200 1188 99 120 116 96.67 60 59 98.33 Ninh Thuan 58 58 100 1200 1195 99.58 116 114 98.28 58 58 100 Binh Phuoc 60 60 100 1200 1195 99.58 120 118 98.33 60 60 100 Tay Ninh 60 60 100 1200 1199 99.92 120 118 98.33 60 60 100 Binh Duong 59 59 100 1200 1197 99.75 118 117 99.15 59 59 100 Dong Nai 60 60 100 1200 1197 99.75 120 120 100 60 60 100 Binh Thuan 60 60 100 1200 1199 99.92 120 120 100 60 60 100 Ba Ria - Vung Tau 60 60 100 1200 1195 99.58 120 120 100 60 59 98.33 Long An 59 59 100 1200 1195 99.58 118 117 99.15 59 59 100 Dong Thap 60 60 100 1200 1196 99.67 120 119 99.17 60 60 100 An Giang 60 60 100 1200 1190 99.17 120 120 100 60 60 100 Tien Giang 60 60 100 1200 1195 99.58 120 120 100 60 60 100 Vinh Long 60 60 100 1200 1197 99.75 120 118 98.33 60 60 100 Ben Tre 60 60 100 1200 1193 99.42 120 120 100 60 60 100 Kien Giang 60 60 100 1200 1185 98.75 120 118 98.33 60 60 100 Can Tho 60 60 100 1200 1191 99.25 120 118 98.33 60 60 100 Tra Vinh 60 60 100 1200 1198 99.83 120 120 100 60 60 100 Soc Trang 59 59 100 1200 1193 99.42 118 118 100 59 59 100 Bac Lieu 60 60 100 1200 1190 99.17 120 120 100 60 60 100 Ca Mau 60 60 100 1200 1185 98.75 120 120 100 60 60 100 Total 3639 3635 99.89 73200 72660 99.26 7278 7178 98.63 3639 3631 99.78 18 Vietnam Reading and Mathematics Assessment Study teachers, and 99.26 percent for pupils from 99.89 percent of schools. Derived variables A series of derived variables were constructed and added to the file. These will become obvious as the reader progresses through the following chapters. A comment on the major breakdown variables used in tables in the report In the tables that follow in this report, the vast majority of them have been presented using region broken down by school location as the form of presentation. In order for readers to be fully aware of what these breakdowns are some information has been provided in this first chapter. Regions are the eight regions of Vietnam: Red River Delta, Northeast, Northwest, North Central, Central Coast, Central Highlands, Southeast, and Mekong Delta. Most ensuing tables in A question was asked of each School Head about the location of the school. this report have Where is your school located? examined the data by (Mark X next to the most appropriate answer) region and by school in a remote or isolated region ___ location. in a rural area ___ in or near a small town ___ in a town or city ___ The third and fourth answers were combined in order to have only three categories, isolated, rural, and urban. Table 1.4: Characteristics of school locations School location Province Urban Rural Isolated % SE % SE % SE Red River Delta 17.3 1.43 81.5 1.49 1.2 0.66 Northeast 16.1 1.36 67.7 1.57 16.3 1.32 Northwest 15.0 2.67 49.7 3.40 35.3 3.48 North Central 14.5 2.12 75.8 2.68 9.7 1.76 Central Coast 25.0 2.85 63.0 3.14 12.0 1.95 Central Highlands 27.3 4.22 51.0 4.48 21.7 3.73 Southeast 41.6 2.22 45.9 2.31 12.5 1.79 Mekong Delta 17.2 1.69 67.3 1.86 15.6 1.38 Vietnam 21.3 0.68 66.5 0.77 12.3 0.63 The percentages of pupils in the sample from the various school locations in the eight regions have been given in Table 1.3 below; Since the isolated schools come in for a good deal of comment, it is worth Table 1.3: School location by region Average School distance Speaking Vietnamese Ethnic Books at Total Number of Primary + location from minority home possessions at satellite secondary amenities (km) Never Sometimes Always home campuses school Mean SE % SE % SE % SE % Mean Mean SE Mean SE Mean SE % SE Isolated 9.4 0.43 11 1.0 24 1.4 65 1.9 41 2.1 8.3 0.49 8.0 0.09 3.9 0.14 18 1.7 Rural 3.9 0.08 3 0.2 5 0.3 92 0.4 10 0.5 11.6 0.28 9.3 0.03 2.1 0.04 4 0.4 Urban 1.6 0.05 1 0.2 2 0.4 97 0.5 7 0.7 22.5 0.94 11.7 0.07 0.8 0.05 2 0.6 All schools 4.0 0.07 3 0.2 7 0.3 90 0.4 13 0.4 13.5 0.30 9.7 0.03 2.1 0.03 5 0.4 19 Vietnam Reading and Mathematics Assessment Study Isolated schools had pupils describing more what kinds of schools these are. Various characteristics of from poor ethnic minority these schools have been presented in Table 1.4 below. families, coming from The average distance from amenities is the average number of kilometers to further distances, the nearest health clinic, the nearest road accessible by car, the nearest public sometimes not speaking library, the nearest bookstore, the nearest junior secondary school, and the Vietnamese out side of nearest market. It can be seen that the average distance to these amenities schools. These schools had from isolated schools was more than from rural schools which in turn was more satellite campuses. somewhat more than from urban schools. The standard deviation of kilometers was greater for isolated schools than for other schools. It should also be noted that there were many missing data for this first variable and hence the distances for the isolated areas was probably underestimated. The pupils were asked to what extent they spoke Vietnamese outside of school The replies were never, sometimes, and always. It can be seen that pupils in isolated areas tended to speak less Vietnamese. There were many more ethnic minority pupils in the isolated areas. The pupils in the isolated areas had the least number of books in the home, had the least number of possessions in the home, and were in schools with more satellite campuses (1= main school only, 2 = 1-2 campuses, 3 = 3-4 campuses, 4 = 5-6 campuses, 5 = 7-8 campuses, 6 = 9-10 campuses and 7 = more than ten campuses). Finally 18 percent of pupils in isolated areas were in schools designated as primary school plus secondary school whereas in then other areas the pupils tended to be more in primary only schools. Conclusion In this chapter a very brief description has been presented of the setting for the study and the procedures undertaken in order to construct the instruments and then conduct the sample survey in Vietnam. A much more detailed description of the procedures for the study has been given in Volume 3 of this series. In the following six chapters of this report emphasis has been placed on the educational policy implications of the results arising from the data analyses. In each of these chapters one of the main policy questions described in this first chapter has been addressed. In Chapter 2 the achievement levels of both pupils and teachers have been described. The reason for having the achievement results first is that in the ensuing chapters reference has often been made to the achievement levels and scores. In Chapter 3 the personal characteristics and the home backgrounds of the Grade 5 pupils have been described. The characteristics of Grade 5 teachers have been presented in Chapter 4 and in Chapter 5 the characteristics of school heads and schools have been described. How the schools met with the Ministry's own benchmark standards has been taken up in Chapter 6 while in Chapter 7 analyses of the extent to which inputs (human and material resources) to schools have been allocated in an equitable fashion among and within provinces have been presented. An effective schools' analysis has been described in Chapter 8. Various attempts analysing the relative `effects' of selected variables on achievement have been described in Chapter 9. Finally, in Chapter 10 an `Agenda for action' has been presented. This summarises 20 Vietnam Reading and Mathematics Assessment Study Appendix 1.1: Examples of policy questions (Page 1 out of 20 pages) Group Question Policy question Table Questionnaire number reference reference Pupils 1 What are the characteristics, including home background, of the Grade 5 pupils? What actions do these characteristics require the Ministry to take? Do these characteristics and the home background have an influence on achievement? 1.01 What is the age distribution of Grade 5 pupils? Are there 3.1 P02 D distribution patterns requiring corrective action and/or having an influence on teaching methods and/or the curriculum? 1.02 What is the sex distribution of Grade 5 pupils? Are there 3.1 P03 D imbalances in the enrolment of male and female pupils requiring corrective action? 1.03 What is the ethnic group of the children? 3.1 P05 D 1.04 How regularly do Grade 5 pupils eat meals? 3.1 P10 1.05 What percentage of Grade 5 pupils speaks the language of 3.2 P04 D the test at home? 1.06 What is the level of the parents' education of Grade 5 3.1 P11 D pupils? (Add) P12 D 1.07 How many books are there in pupils' homes? 3.1 P06 D P = Pupil questionnaire D = Derived variable to be computed (see Vietnam: recoding of variables) TQ = Teacher questionnaire R = Rasch score to be computed S = School Head questionnaire In some cases the implication of the question is spelled out but in other cases this has not been done because it is felt to be self-evident. 21 Vietnam Reading and Mathematics Assessment Study Appendix 1.1: Examples of policy questions (Page 1 out of 20 pages) Group Question Policy question Table Questionnaire number reference reference Pupils 1 What are the characteristics, including home background, of the Grade 5 pupils? What actions do these characteristics require the Ministry to take? Do these characteristics and the home background have an influence on achievement? 1.01 What is the age distribution of Grade 5 pupils? Are there 3.1 P02 D distribution patterns requiring corrective action and/or having an influence on teaching methods and/or the curriculum? 1.02 What is the sex distribution of Grade 5 pupils? Are there 3.1 P03 D imbalances in the enrolment of male and female pupils requiring corrective action? 1.03 What is the ethnic group of the children? 3.1 P05 D 1.04 How regularly do Grade 5 pupils eat meals? 3.1 P10 1.05 What percentage of Grade 5 pupils speaks the language of 3.2 P04 D the test at home? 1.06 What is the level of the parents' education of Grade 5 3.1 P11 D pupils? (Add) P12 D 1.07 How many books are there in pupils' homes? 3.1 P06 D P = Pupil questionnaire D = Derived variable to be computed (see Vietnam: recoding of variables) TQ = Teacher questionnaire R = Rasch score to be computed S = School Head questionnaire In some cases the implication of the question is spelled out but in other cases this has not been done because it is felt to be self-evident. 21 Vietnam Reading and Mathematics Assessment Study Appendix 1.2: Blueprint of Reading Test: The structure of the reading test (Dimensions and total questions) Levels Narrative Expository Documents Level 6 Linking ideas from different Linking ideas from different Use of embedded lists and even parts of text. Making parts of text. Making subtle advertisements where the inferences from text or beyond inferences from text or message is not explicitly stated text. beyond text. 5 2 2 9 Level 5 Seeking and confirming Seeking and confirming Linking more than one piece of information when reading information when reading information in different parts of backwards through text backwards through text a document 4 5 3 12 Level 4 Linking information portrayed Linking information Systematic search for in sequences of ideas and portrayed in sequences of information when reading content, when reading forward ideas and content, when forward. reading forward 8 10 2 20 Level 3 Recognising the meaning of a Recognising the meaning of Linking simple piece of single word and being able to a single word and being able information to item or express it as a synonym in to express it as a synonym in instruction order to answer the question order to answer the question 6 5 2 13 Level 2 Word/picture association Word/picture association Word/picture association involving positional or involving positional or involving positional or directional prepositions directional prepositions directional prepositions requiring the linkage of a requiring the linkage of a requiring the linkage of a picture to a position or a picture to a position or a picture to a position or a direction in order to answer direction in order to answer direction in order to answer the the question the question question 1 1 1 3 Level 1 -- -- Word/picture association involving nouns and/or adjectives requiring the simple linkage of a picture to a word in order to answer the question 3 3 24 23 13 60 22 Vietnam Reading and Mathematics Assessment Study Appendix 1.3: Blueprint of mathematics test: Description of the levels of mathematics competence (and domains and numbers of items) Level Number Measurement Space/data Level 6 Combining operations in Combining operations in order Linking data from tables and order to undertake to undertake computations graphs in order to undertake computations involving involving several steps using computations involving several steps using a mixture a mixture of operations and several steps and with a of operations and translation of units mixture of operations combinations of fractions, decimals and whole numbers 3 2 3 8 Level 5 Combining operations in Two or three step operations Combining operations in order order to link information as in Number, using to link information from tables from tables and charts when measurement units and and charts when performing performing calculations conversion calculations 4 4 4 12 Level 4 Extend and complete Convert measurement units To translate shapes and number patterns, when undertaking one step patterns computations 4 4 4 12 Level 3 Simple fractions Basic calculations with Identify data in tabular form simple measurement units 4 4 4 12 Level 2 Single operations using Recognise units of Linking of patterns and graphs two digit numbers measurement to single digits. Recognise and name basic shapes 4 4 2 10 Level 1 Number recognition, linking patterns to numbers. 6 0 0 6 Total 6 0 0 60 The same description of the numbers applies as that from beneath Appendix 1.2. 23 Vietnam Reading and Mathematics Assessment Study Appendix 1.4: Names and affiliations of persons involved in Study Team Name Affiliation Entry Do Cong Vinh National Institute of Educational Science Ton Quang Cuong Hanoi National University Bui Kien Trung Hanoi National University Editing Pham Lan Huong National Institute of Educational Science Mai Kim Oanh National Institute of Educational Science Ngo Thanh Tung National Institute of Educational Development Vu Hong Yen National Institute of Educational Science Ngo Van Trung National Institute of Educational Science Cleaning Le Quang Phan National Institute of Educational Science Tran Tich Lang National Institute of Educational Science Tran Thai Ha National Institute of Educational Development Ta Ngoc Thanh National Institute of Educational Science Miyako Ikeda The World Bank Members of the analysis group were: Miyako Ikeda, Le Quang Phan, Tran Thai Ha, Bui Kien Trung, and the head of the team, Dr. Cong Vinh. 24 Vietnam Reading and Mathematics Assessment Study Chapter 21 WHAT WERE THE LEVELS OF ACHIEVEMENT OF GRADE 5 PUPILS IN READING AND MATHEMATICS Introduction I n any system of education, probably the most important aspect is Tests were used as a "whether or not the pupils are learning", or stated in a slightly more measure of learning at nuanced way "to what extent the pupils have learned what they were the end of primary meant to learn". In later chapters of this report, an examination has been school made of the home backgrounds of pupils, the classrooms in which they were, the teachers they had, and finally of the conditions of the schools they attended. But, in the end, the important question is "how well did the pupils learn?". In this chapter, a proxy measure for learning has been examined - the pupils' achievement measured towards the end of their time in Grade 5. It is, in a way, the culmination of learning that has taken place up to the end of Grade 5 and is an indication of the state of learning at the end of primary education in Vietnam. In order to structure the chapter the following major questions have been posed and answered. 1. What did the tests measure and how is this reported? 2. What percentages of pupil reached the different levels of skills in reading and mathematics? 3. What percentages of pupil reached different benchmarks in reading and mathematics? 4. What were the test scores in reading and mathematics? 5. What were the differences in test scores in both reading and mathematics between boys and girls and between the different socio-economic groups? 1This chapter was written by Professor Patrick Griffin of Melbourne University 25 Vietnam Reading and Mathematics Assessment Study 6. Were the pupils 'elite' (upper 5 %) performances similar in different regions and economically advantaged areas? To what extent did the performance 'tails' (bottom 5%) differ across regions and economically advantaged areas? 7. What were the relationships between teacher and pupil performance on the tests of reading and mathematics? What did the tests measure? Levels of achievement The tests were developed by teams of specialists from the National Institute in a criterion of Educational Science (NIES) following the specifications set out in Chapter referenced framework 1. They ensured that the test items measured both the old and new provided a competency curriculum. They were designed to measure achievement in two subject basis for describing areas: reading comprehension and mathematics. There were three different learning outcomes. The approaches that were used in the measurement of achievement. These have tests were constructed been explained to some extent in the test development section in Chapter 1 by NIES and in detail in Volume 3. The Vietnamese mathematics curriculum is similar to curriculum that is common to many countries. This, however, is not the case with the reading curriculum. In language education the emphasis in on a local curriculum which conveys the culture as well as the skills in reading, writing, listening and speaking. In Vietnam, the language curriculum is integrated in its approach and in its role of cultural development. This project has assessed reading on its own and in a manner unfamiliar to the pupils and teachers. The usual integrated approach to reading, discussion and writing in language classes, was not included in the test used. It could not be done with such a large sample and the performance testing techniques that the assessment of speaking and writing would require.. Despite this the level of reading achievement was generally high. This is not the case with the mathematics curriculum, which is linked to international mathematics approaches to mathematics teaching and learning. International curriculum trends and assessment methods influence it. The mathematics curriculum has much that is common with most other countries. Changes in curriculum Other aspects of testing and of changes in curriculum were also taken into taking place in account. The objectives of the language and reading curriculum emphasise primary education reading for details and not processing or inference development. This is were taken into changing in the curriculum currently being introduced to primary schools. account in the tests. The test also contained two areas possibly not familiar to all students- expository and document texts. Document literacy in particular could be expected to be associated with lower performance, as it does not involve continuous prose. However both expository and document styles of text can be expected to be routinely encountered in secondary texts and were therefore legitimately included in the tests. The three different measures were used because they provided different uses and interpretation of the data. The first describes the skill development of the pupil. The second indicates how the pupil has been prepared to cope with the reading demands of two contexts 26 Vietnam Reading and Mathematics Assessment Study into which the pupil will progress- the Vietnamese community and lower secondary education. The third is a scaled score used to analyse and interpret differences in performances between groups. Each of the measures and their uses are described in more details and the details of how they were developed are provided in Volume 3. a) Skill levels. Six skill levels were identified in each of the subject areas under Benchmarking at two consideration. The levels are hierarchical. It was possible to calculate the levels of competence- percentages of pupils who had or had not reached each of the skill levels. These minimum for skill levels enable a direct interpretation of the pupils' performances in each of participation in society - the tests and to link the skills to possible intervention through curriculum and desired for development and teaching strategies. They provide a criterion referenced and independent learning in developmental framework for interpreting how pupils are progressing and lower secondary were what whey can and cannot do in terms of increasing levels of competence. Details of how these levels were defined have been included in Volume 3. developed b) Benchmarks: Two benchmark levels2 were also established. They were Minimum benchmark based on the pupil's ability to cope with reading and mathematics tasks defined for societal encountered in specific circumstances. The first benchmark was based on a involvement pupil's ability to use a set of reading and mathematics skills needed to function in Vietnamese society. Those below this benchmark were described as 'pre-functional'. A second benchmark was based on an estimation of a pupil's ability to cope with the reading and mathematics tasks in the next grade of education, Grade 6, which is the first year of secondary education. The two benchmarks helped to identify three groups of pupils. Those below the first benchmark would need considerable help to enable them to function and participate fully in Vietnamese society. Those above this benchmark but below the second would need assistance to help them cope with the reading and mathematics involved in secondary education. Pupils above the second benchmark were expected to be able to cope with the reading and mathematics involved in secondary education. i) Benchmark 1: A group of pupils were described as pre functional because they had not yet reached a benchmark demonstrating reading or mathematics required for everyday activities in Vietnamese society. The label used in the tables is 'Pre functional'; It does not mean that a pupil is illiterate or non-numerate. There are basic skills that these pupils can demonstrate, but the skill level is not yet deemed by experts to be at a sufficient level to enable the person to be an effective member of Vietnamese society. ii) A second group of pupils was identified as those who could demonstrate A desired benchmark of the kinds of skills needed to cope with life in Vietnam. They were found independent learning skills was developed 2 A detailed technical explanation of how and why this was done has been given in Volume 3 of this series. The experts made two ratings for each item in the tests. The first was the probability that a person who could adequately function in Vietnamese society could obtain the correct answer to each item. The second was the probability that a pupil who had adequate skills to cope with Grade 6 learning, could obtain the correct answer to each item. In each case the probabilities were summed using an Angoff approach to establish the cut-off points. 27 Vietnam Reading and Mathematics Assessment Study to be above this lower benchmark but had not yet reached the second benchmark. These pupils were designated as "functional" in terms of their capacity to participate in Vietnamese society. However it was deemed that this group would need some remedial assistance to be able to cope with the reading and mathematics required at Grade 6 level. iii) Benchmark 2: These pupils, whose performances above the second benchmark, were described as demonstrating the kinds of skills that were desirable in order to learn independently at the next level of schooling, without needing remedial assistance. The label used in the tables was 'independent'. Scaled test scores were c) Total scores and sub-scores. Pupils' scores on each test and sub domain of used for comparison of items were converted to a new, neutral common scale allowing estimates of sub groups. (with a relative educational importance of differences between groups. scale with a mean of 500 and mean of 500 and a standard deviation of 100. This is a frequently used standard deviation of conversion of scores. The scaled score is used to examine differences 100). A difference of 20 between sub groups in the sample. In education differences of 0.2 standard points represents an deviations are often considered to be large and educationally important. With important difference very large samples such as we have in this study, very small differences can be statistically significant and this is often mistaken for educational importance. Using the scale with a mean of 500 and a standard deviation of 100, gives a simple way of interpreting differences. Scores below 500 are below the national average and scores over 500 are above the national average. Differences between groups of more than 20 points deserve to be examined in detail for possible explanation as they may be of some educational importance. Details of how these scores were derived have been presented in Volume 3. Before presenting the results it is important to discuss the standard errors of sampling presented for each estimate in this report. In Table 2.1 it can be seen that 4.6 percent of pupils were at Level 1 in reading. The standard error of sampling (SE) for each estimate provides an indication of how accurate the result is. For example, the estimate of the proportion of pupils at Level 1 was 4.6 percent and the error estimate for this was 0.17 percent. These figures indicated that we could be 95 percent confident that the population percentage of pupils who reached Level 1 was within the following limits: 4.6 ą 2 (.17) percent. That is, somewhere between 4.26 percent and 4.94 percent. What percentages of pupils reached the different skill levels in reading and mathematics? The skill levels have been examined at a) the national level, b) the regional level, and c) the provincial level, as well as by d) school location. To summarise the results, Vietnamese pupils demonstrated higher levels of skill in mathematics than in reading. Both in reading and mathematics, pupils in urban areas tended to achieve higher skill levels than pupils in isolated areas. 28 Vietnam Reading and Mathematics Assessment Study In reading, more than 30 percent of pupils in isolated areas in Northeast, Northwest, Central Coast, the Central Highlands reached only Levels 1 and 2. In isolated and rural schools in the Mekong Delta, more than 30 percent of pupils did not reach Level 3. While these percentages represent relatively small numbers of pupils, it illustrates the problem faced by pupils in isolated areas, relative to pupils in more urban contexts. Details of these performances are discussed later in this chapter. a) At the national level From Table 2.1, it can be seen that, at the national level, pupils in Vietnam demonstrated quite high levels of skills in reading and mathematics but that there was still a considerable number of pupils with low levels in reading (Levels 1 and 2). Table 2.1 Percentages of pupils reaching the different skill levels in reading and mathematics? Reading Skill Levels Percent SE More pupils reached Level 1 Matches text at word or sentence level aided by pictures. Restricted to a limited 4.6 0.17 high levels of range of vocabulary linked to pictures Mathematics skill than Level 2 Locates text expressed in short repetitive sentences and can deal with text 14.4 0.28 reached high levels of unaided by pictures. Type of text is limited to short sentences and phrases with repetitive patterns. Reading skills, Level 3 Reads and understands longer passages. Can search backwards or forwards 23.1 0.34 although overall through text to for information. Understands paraphrasing. Expanding achievement levels on vocabulary enables understanding of sentences with some complex structure. the tests were high Level 4 Links information from different parts of the text. Selects and connects text to 20.2 0.27 derive and infer different possible meanings. It is possible that the Level 5 Links inferences and identifies an author's intention from information stated in 24.5 0.39 mathematics test was different ways, in different text types and in documents where the message is not explicit. relatively easy Level 6 Combines text with outside knowledge to infer various meanings, including 13.1 0.41 compared to the hidden meanings. Identifies an author's purposes, attitudes, values, beliefs, reading test. motives, unstated assumptions and arguments. Mathematics Skill Levels Percent SE Level 1 Reads, writes and compares natural; numbers, fractions and decimals. Uses single operations of +,-,x and : 0.2 0.02 on simple whole numbers; works with simple measures such as time; recognises simple 3D shapes. Converts fractions with denominator of 10 to decimals. Calculates with whole numbers using one operation Level 2 (x,-,+ or : ) in a one-step word problem; recognises 2D and 3D shapes. 3.5 0.13 Identifies place value; determines the value of a simple number sentence; understands equivalent fractions; Level 3 adds and subtracts simple fractions; carries out multiple operations in correct order; converts and estimates 11.5 0.27 common and familiar measurement units in solving problems. Reads, writes and compares larger numbers; solves problems involving calendars and currency, area and Level 4 volume; uses charts and tables for estimation; solves inequalities; transformations with 3D figures; 28.2 0.37 knowledge of angles in regular figures; understands simple transformations with 2D and 3D shapes. Calculates with multiple and varied operations; recognises rules and patterns in number sequences; calculates the perimeter and area of irregular shapes; measurement of irregular objects; recognised Level 5 transformed figures after reflection; solves problems with multiple operations involving measurement units, 29.7 0.41 percentage and averages. Problem solving with periods of time, length, area and volume; embedded and dependent number patterns; develops formulae; recognises 3D figures after rotation and reflection and embedded figures and right Level 6 angles in irregular shapes, data from graphs and tables; 7 0.6 29 Vietnam Reading and Mathematics Assessment Study Changes in the The percentage of pupils reaching the top three levels in reading was lower curriculum are taking than in mathematics. There are a number of reasons for this. The reading test place in primary included items from old and new curriculum and text genre that were related education. to the secondary school, but not currently included in the primary school curriculum. The gap between the upper primary and lower secondary school reading demands placed on pupils is larger than the gap between the upper primary and lower secondary mathematics demands. For instance, the objectives of the reading curriculum emphasise reading for detail and not on processing and inference development. While this is changing in the new curriculum, the development of teachers and pupils has not yet reached the levels that appear to have been achieved in mathematics where the transition to secondary curriculum is not as abrupt. There are many issues that need to be dealt with in the Vietnamese language curriculum in the transition to the new curriculum, including the tradition of emphasising detail in reading. Accordingly the test performance was analysed using items related to the old curriculum and items related to the new curriculum. Results of these analyses have also been presented in this chapter. The test results have been reported in developmental or criterion referenced terms. This is because criterion-referenced interpretations of performances have meaning for the kind of assistance needed by pupils, teachers and curriculum developers as the new curriculum is ushered into the system and the schools. This developmental assessment strategy illustrated in this chapter provides powerful diagnostic and substantive information about pupil performance and related curriculum. Policy suggestion 2.1: The MOET might wish to consider the introduction of curriculum and assessment frameworks and benchmarks written in criterion referenced (profile) format. This can be done as the new curriculum is introduced. This will encourage a competency-based developmental interpretation of pupil performances. Policy suggestion 2.2: The MOET might wish to use developmental assessment procedures to plan intervention programs for specific groups of pupils and teachers. b) At the regional level High levels in the Red The percentages of pupils who reached the various levels of skills in reading River region and low and mathematics have been presented in Table 2.2. It can be seen that there achievement in the were large differences between regions in terms of achievement of levels of Mekong Delta were skills. At Level 6, for instance, the highest level measured by the tests, the observed in both reading Mekong Delta had 7.5 percent of pupils at this level in mathematics and 13.0 and mathematics percent in reading. In contrast to this, the Red River Delta had 18.8 percent and 37.8 percent respectively. At the lower levels of reading skills, especially Levels 1 and 2 the Mekong Delta had a total of 6 percent compared with the Red River Delta's 2.1 percent. A similar ratio of skill level was evident in mathematics where the percentages were 30.4 and 10.5 respectively. It is 30 Vietnam Reading and Mathematics Assessment Study clear that there were important differences between the regions in terms of achievement of pupils. Table 2.2: Percentages and sampling errors of pupils at each skill level in math and reading by region Mathematics Skill Level Region Level 1 Level 2 Level 3 Level 4 Level 5 Level 6 % SE % SE % SE % SE % SE % SE Red River Delta 0.1 0.02 2 0.25 7.2 0.45 22.4 0.87 30.6 0.92 37.8 1.75 Northeast 0.2 0.04 4.4 0.34 12.2 0.52 24.9 0.84 26 1 32.3 1.34 Northwest 0.8 0.28 8.3 1.37 13.3 1.39 24 2.12 24.6 2.22 28.9 3.07 North Central 0.1 0.11 2.2 0.36 7.6 0.69 24.3 1.36 33.2 1.46 32.6 2.32 Central Coast 0 0.02 2.2 0.31 10.1 0.68 30.5 1.12 34.2 1.01 22.9 1.6 Central Highlands 0.3 0.15 2.9 0.54 9.2 1.21 24.5 1.84 32.6 2.18 30.5 3.1 Southeast 0.1 0.03 2.5 0.23 10.2 0.59 31.2 1.04 33.8 0.97 22.2 1.48 Mekong Delta 0.2 0.04 5.8 0.32 19.6 0.68 37.3 0.83 24.1 0.9 13 0.93 Vietnam 0.2 0.02 3.5 0.13 11.5 0.27 28.2 0.37 29.7 0.41 27 0.6 Reading Skill Level Region Level 1 Level 2 Level 3 Level 4 Level 5 Level 6 % SE % SE % SE % SE % SE % SE Red River Delta 1.9 0.2 8.6 0.5 19.7 0.75 20.9 0.69 30.1 0.98 18.8 1.24 Northeast 5.5 0.4 15 0.59 20.6 0.74 19.2 0.6 24.3 0.8 15.4 1 Northwest 8.9 1.3 16.9 1.39 23.3 1.72 18.6 1.65 23.4 2.22 9 1.44 North Central 4.1 0.59 11.9 0.93 22 1.11 20.5 0.83 28.6 1.65 13 1.43 Central Coast 4.4 0.66 15.7 0.78 25.4 0.87 21 0.7 22.6 1.04 11 1.23 Central Highlands 6.4 1.07 12.5 1.48 20.1 1.59 20.9 1.78 26.9 1.93 13.2 2.4 Southeast 2.6 0.29 11.7 0.66 24.9 0.94 22.6 0.77 25.1 0.77 13.1 1.06 Mekong Delta 7.4 0.36 23 0.74 27.6 0.56 18.1 0.59 16.6 0.65 7.5 0.67 Vietnam 4.6 0.17 14.4 0.28 23.1 0.34 20.2 0.27 24.5 0.39 13.1 0.41 c) At the provincial level The percentages of pupils at the different skill levels for each province have There were important been presented in Tables 2.3 and 2.4 for reading and mathematics. Even differences between within the top performing region (the Red River Delta) it can be seen that provinces in both there were large differences between Ha Noi and Ha Nam provinces. These reading and kinds of differences were even larger in the North East Region when Bac mathematics Ninh and Long San provinces were compared. achievement The province with the highest percentage of pupils at Level 6 was Quang Ninh. This was in both reading and mathematics. The province with the highest percentage of pupils at Level 1 in reading was Kon Tum. For math it was Hoa Binh. d) By school location The school location was coded as "isolated", "rural", or "urban" as described 31 in Chapter 1. From Table 2.5 it can be seen that most pupils at Level 1 were to be found in the isolated areas and that most pupils at Levels 6 were to be found in the urban areas. This was true for both reading and math. It was in the Mekong Delta as well as in the Northwest, Northeast, Central Coast, and the Central Highlands that there were unacceptably high percentages of pupils at Levels 1 and 2 in isolated schools. Policy suggestion 2.3: The MOET might wish to consider the introduction of performance monitoring programs at province and district level in which school, district and province level performances are monitored against competency levels and benchmarks. Policy suggestion 2.4: The MOET might wish to consider for district and provincial officers specific training in techniques for monitoring and intervention against competency levels and benchmarks. In Table 2.3 it can be seen that the distribution of pupils at the lower skill level was not even. In some provinces there were many pupils at low skill level. In other provinces there were few. Clearly this information can be used to design targeted intervention programs aimed at redressing highly specific skill deficits of groups on pupils. There were for instance more than 10 percent of pupils at level 1 in reading skills in Cao Bang, Tuyen Quang, Hoa Binh, Kon Tum, Tra Vinh, and Bac Lieu. Such prevalent low skill levels in reading at the end of primary school clearly require intervention, before these pupils enter the community as independent citizens or begin their lower secondary education and expect to be independent learners. These incidences of low skill levels were prevalent in four regions. The North West, North East, Central Highlands and the Mekong Delta. On the other hand there were provinces where more than 20 percent of pupils are at the highest level of reading assessed and more than 35 percent of pupils at the highest leels of mathamatics assessed.. These were found in Ha Noi, Hai Duong, Hung Yen, Thai Binh, Bac Ninh, Quang Ninh, Da Nang and Ho Chi Minh. All of them were in the highly industrialised urban regions of Vietnam. Vietnam Reading and Mathematics Assessment Study Table 2.3: Percentages and sampling errors of pupils at different skill levels in Reading. Reading Skill Levels (pupil) Region Province Level 1 Level 2 Level 3 Level 4 Level 5 Level 6 % SE % SE % SE % SE % SE % SE Ha Noi 0.9 0.38 5.4 1.31 15.1 1.99 18.9 1.73 33.5 2.48 26.2 2.83 Hai Phong 2.2 0.62 9.0 1.35 23.5 2.25 20.4 1.82 27.9 2.42 17.0 3.06 Ha Tay 2.7 0.69 11.5 1.66 21.0 2.56 21.8 2.34 27.2 3.10 15.8 3.29 Hai Duong 1.6 0.43 8.0 1.52 19.1 2.24 18.4 1.98 28.4 2.90 24.5 4.31 Red River Delta Hung Yen 2.1 0.72 7.1 1.40 16.0 2.42 21.4 1.95 32.9 3.11 20.4 3.52 Ha Nam 4.0 0.92 15.2 1.61 26.4 2.30 23.6 2.38 23.1 2.35 7.8 1.80 Nam Dinh 0.9 0.27 6.1 1.18 15.5 1.94 21.5 1.94 38.9 2.68 17.2 2.79 Thai Binh 0.6 0.25 3.8 0.83 16.3 2.13 20.6 2.13 34.2 2.71 24.6 3.87 Ninh Binh 4.6 1.08 16.6 1.81 32.4 1.93 24.1 1.74 16.7 1.78 5.6 1.55 Ha Giang 7.5 1.66 22.1 3.23 27.4 3.06 18.7 2.97 18.5 3.07 5.7 2.09 Cao Bang 14.4 3.16 22.4 3.28 23.4 3.11 16.2 2.61 14.9 3.10 8.7 3.04 Lao Cai 1.4 0.78 6.7 1.68 15.3 2.18 22.7 2.90 38.2 3.59 15.7 3.40 Bac Kan 8.2 2.02 21.5 2.77 26.3 2.67 15.5 2.25 18.7 3.09 9.9 3.07 Lang Son 11.0 2.20 26.2 2.87 22.2 2.32 18.2 2.45 16.9 2.98 5.6 1.46 Tuyen Quang 12.5 2.26 24.9 2.71 24.2 2.56 18.7 2.23 15.7 2.71 3.9 1.30 Northeast Yen Bai 11.4 2.27 20.9 2.40 23.1 2.24 17.8 2.15 14.0 2.22 12.9 3.51 Thai Nguyen 3.8 0.98 10.4 1.84 18.3 2.31 18.8 2.09 29.7 3.66 19.1 4.00 Phu Tho 2.4 0.59 11.3 1.81 19.5 2.35 22.4 2.17 28.3 2.66 16.2 3.61 Vinh Phuc 5.3 1.06 18.1 2.34 22.7 2.01 20.1 1.84 22.5 2.56 11.3 3.06 Bac Giang 4.5 1.08 16.2 2.03 26.1 2.05 22.2 1.67 22.2 2.39 8.8 2.18 Bac Ninh 0.5 0.24 4.0 1.09 12.7 2.01 16.4 2.14 31.5 3.12 34.9 4.76 Quang Ninh 0.3 0.20 4.0 1.05 11.0 2.16 14.9 2.06 32.8 3.88 37.0 5.71 Lai Chau 2.2 0.94 12.1 2.36 18.9 2.98 25.5 3.05 33.2 3.79 8.0 2.05 Northwest Son La 8.7 2.29 16.3 2.49 22.6 2.87 19.2 3.19 23.9 3.77 9.5 2.79 Hoa Binh 12.3 2.35 19.9 2.63 26.1 3.23 14.7 1.91 18.0 3.80 9.0 2.37 Thanh Hoa 4.2 0.72 14.6 2.25 23.2 2.32 18.7 1.74 28.7 3.20 10.6 2.53 Nghe An 6.7 1.72 12.1 1.75 22.7 2.62 18.8 1.89 26.7 3.35 13.1 3.50 Ha Tinh 1.2 0.44 8.3 1.60 20.2 2.18 23.6 2.05 27.5 2.28 19.2 3.68 North Central Quang Binh 0.8 0.43 8.3 2.13 16.5 2.14 24.2 2.31 36.0 3.27 14.2 3.23 Quang Tri 3.5 1.12 12.2 1.87 25.4 2.62 23.0 1.92 27.3 2.59 8.7 2.00 Thua Thien - Hue 2.1 0.92 8.7 1.60 20.0 2.18 23.8 2.10 30.2 2.46 15.2 3.45 Da Nang 0.8 0.34 5.7 0.88 15.4 1.79 21.3 1.89 32.9 1.98 24.1 3.23 Quang Nam 4.3 0.91 16.6 2.34 23.1 2.18 20.9 2.03 26.1 2.76 8.9 1.91 Quang Ngai 7.1 1.61 20.2 2.44 27.5 2.36 17.4 1.68 18.4 2.11 9.5 2.52 Central Coast Binh Dinh 4.1 2.13 14.6 2.07 26.9 2.55 21.1 2.23 19.8 2.81 13.6 4.20 Phu Yen 3.8 0.74 15.0 1.73 26.8 2.10 22.5 1.70 23.5 2.34 8.3 1.69 Khanh Hoa 4.0 0.81 16.6 1.86 28.7 1.92 23.9 1.54 20.3 1.82 6.5 1.27 Kon Tum 18.7 4.71 19.1 2.32 21.5 2.70 14.0 1.90 17.7 3.43 8.9 2.91 Central Highlands Gia Lai 7.8 2.14 13.5 2.31 18.7 2.63 18.9 2.75 27.1 3.53 13.9 3.41 Dak Lak 3.9 0.92 11.0 2.02 20.4 2.27 22.9 2.46 28.3 3.03 13.6 3.51 Ho Chi Minh 0.5 0.24 6.2 1.39 19.5 2.11 20.7 2.04 31.0 1.90 22.2 2.89 Lam Dong 3.7 1.18 11.8 1.75 23.4 2.29 21.7 1.99 24.8 2.75 14.7 3.19 Ninh Thuan 8.8 1.56 25.5 2.30 26.1 1.82 17.4 1.63 16.9 2.27 5.3 0.95 Binh Phuoc 4.7 1.01 19.9 1.75 31.0 1.74 23.3 1.90 15.6 1.58 5.6 1.82 Southeast Tay Ninh 4.8 0.95 18.9 2.34 28.5 2.23 22.8 2.53 19.7 2.44 5.4 1.48 Binh Duong 1.6 0.53 8.3 1.26 24.1 2.22 22.9 2.02 29.2 2.71 13.9 2.69 Dong Nai 2.4 0.78 9.1 1.42 25.9 1.65 26.2 1.63 25.8 1.82 10.5 1.27 Binh Thuan 4.0 1.17 18.4 1.71 31.6 2.19 23.0 1.66 16.9 1.95 6.2 1.64 Ba Ria - Vung Tau 1.0 0.39 7.9 1.20 26.9 2.17 24.6 1.61 28.8 2.15 10.8 2.02 Long An 3.9 1.09 17.0 2.12 28.0 1.89 26.5 2.29 20.3 2.32 4.3 1.25 Dong Thap 6.0 1.30 23.1 2.63 26.8 2.26 18.5 2.28 18.0 2.87 7.6 2.41 An Giang 9.0 1.67 24.3 2.87 26.7 2.50 18.6 2.23 13.9 2.19 7.6 2.38 Tien Giang 2.8 0.70 13.4 2.00 28.8 2.49 20.2 1.80 22.4 2.46 12.5 2.78 Vinh Long 4.1 0.91 18.7 1.93 23.5 1.82 20.4 1.73 21.9 2.27 11.5 3.00 Ben Tre 2.9 0.72 13.8 1.57 28.0 2.30 24.7 1.56 21.8 2.22 8.9 2.02 Mekong Delta Kien Giang 9.6 1.70 27.6 2.48 30.3 2.00 12.8 1.38 13.4 2.37 6.3 2.35 Can Tho 8.1 1.46 26.5 2.67 28.3 2.40 15.0 1.69 14.7 2.77 7.3 2.68 Tra Vinh 11.2 1.58 32.3 2.78 27.8 2.29 13.2 1.96 12.6 2.26 2.9 1.49 Soc Trang 13.1 2.36 29.0 2.67 28.6 2.30 16.3 1.94 11.0 2.65 2.0 0.85 Bac Lieu 11.9 1.63 28.2 2.65 26.0 2.35 11.4 1.58 10.8 2.49 11.8 5.02 Ca Mau 8.1 1.27 24.4 2.43 26.7 2.31 17.3 2.19 16.3 2.96 7.3 2.81 Vietnam 4.6 0.17 14.4 0.28 23.1 0.34 20.2 0.27 24.5 0.39 13.1 0.41 33 Vietnam Reading and Mathematics Assessment Study Table 2.4: Percentages and Sampling errors of pupils at different skill levels of Mathematics. Math Skill Levels (pupil) Region Province Level 1 Level 2 Level 3 Level 4 Level 5 Level 6 % SE % SE % SE % SE % SE % SE Ha Noi 0.0 0.04 0.5 0.37 3.3 0.77 14.8 2.01 31.1 2.42 50.3 3.83 Hai Phong 0.1 0.10 1.6 0.56 7.8 1.30 23.0 2.39 33.7 2.63 33.9 4.11 Ha Tay . . 2.9 0.85 9.8 1.46 27.6 3.00 24.8 2.42 34.9 5.01 Hai Duong 0.2 0.11 1.8 0.51 6.3 1.37 21.1 2.83 25.9 2.72 44.7 5.04 Red River Delta Hung Yen 0.1 0.14 1.5 0.51 5.6 1.19 20.9 3.05 28.6 2.84 43.3 5.07 Ha Nam 0.2 0.13 5.1 1.05 13.4 1.78 34.1 2.46 32.8 2.49 14.5 2.42 Nam Dinh . . 1.0 0.48 4.9 1.10 19.4 2.42 38.2 2.99 36.6 4.49 Thai Binh . . 0.6 0.32 3.2 0.84 17.6 2.37 32.3 2.89 46.3 4.63 Ninh Binh 0.4 0.18 5.8 1.35 16.6 2.02 30.8 2.44 30.2 3.04 16.2 3.02 Ha Giang 0.5 0.36 7.9 2.05 16.6 2.63 26.9 3.15 22.9 3.17 25.2 4.70 Cao Bang 0.6 0.49 4.9 1.42 13.9 2.63 28.0 3.42 28.2 3.78 24.3 4.39 Lao Cai 0.0 0.04 0.5 0.29 3.8 1.70 16.3 2.50 34.6 4.55 44.8 5.47 Bac Kan 0.2 0.18 6.1 1.60 15.2 2.64 31.7 3.59 24.3 3.30 22.6 4.90 Lang Son 0.3 0.20 10.7 2.08 25.8 2.88 32.0 2.85 21.6 3.38 9.7 2.85 Tuyen Quang 0.6 0.24 11.7 2.24 21.9 2.55 30.6 3.23 19.9 3.12 15.4 3.52 Northeast Yen Bai 0.5 0.22 10.3 2.20 17.5 2.49 32.8 3.16 18.7 2.53 20.3 4.36 Thai Nguyen . . 2.7 1.09 9.1 1.95 22.5 3.15 28.3 3.23 37.4 4.93 Phu Tho 0.2 0.14 2.3 0.92 8.7 1.72 22.8 2.83 28.6 2.85 37.5 4.57 Vinh Phuc 0.1 0.14 4.7 1.08 14.4 2.15 25.3 2.68 25.9 2.60 29.5 4.57 Bac Giang 0.1 0.06 2.6 0.84 13.1 1.62 33.6 2.61 29.3 2.26 21.4 3.34 Bac Ninh . . 0.2 0.12 2.3 0.82 11.8 2.21 24.4 3.00 61.4 4.81 Quang Ninh . . 0.2 0.18 3.6 0.86 12.0 2.24 25.9 3.47 58.3 5.43 Competency Lai Chau 0.1 0.13 1.8 1.00 7.2 1.93 23.3 3.93 33.8 3.91 33.8 5.03 Northwest Son La 1.1 0.63 7.0 1.89 12.7 2.41 23.9 3.56 23.8 3.50 31.5 4.97 levels of Hoa Binh 0.9 0.35 13.0 2.71 17.0 2.18 24.4 2.96 21.0 3.35 23.8 5.57 pupils were Thanh Hoa . . 2.2 0.53 9.5 1.75 26.8 2.46 35.8 2.74 25.7 4.07 Nghe An 0.4 0.36 3.3 1.11 7.0 1.47 23.5 3.08 29.7 3.13 36.2 4.64 low in some North Ha Tinh . . 1.1 0.44 5.7 1.31 21.8 2.79 30.0 2.27 41.4 4.30 provinces Central Quang Binh . . 1.3 0.47 6.9 1.63 16.6 2.41 34.8 3.51 40.5 5.04 Quang Tri 0.1 0.10 2.3 0.74 7.5 1.21 29.0 2.61 38.2 2.67 22.9 3.74 Thua Thien - Hue . . 0.6 0.48 5.4 1.36 24.2 2.88 34.9 3.09 34.9 4.56 Da Nang . . 0.4 0.20 3.5 0.92 18.5 2.08 37.3 2.62 40.3 3.87 Quang Nam . . 2.2 0.74 9.7 1.47 30.3 2.38 34.2 2.30 23.6 3.30 Central Quang Ngai 0.2 0.11 3.4 0.79 14.4 1.98 29.7 2.79 30.8 3.01 21.6 3.75 Coast Binh Dinh . . 2.1 0.84 8.4 1.75 31.2 3.09 33.2 3.17 25.1 4.67 Phu Yen . . 3.0 0.73 14.7 1.81 31.7 2.32 33.1 2.51 17.5 2.82 Khanh Hoa . . 1.2 0.38 8.2 1.20 36.7 2.67 39.1 2.56 14.8 2.01 Kon Tum 0.8 0.37 14.2 3.13 18.4 2.71 26.9 3.22 21.3 3.26 18.5 4.03 Central Highlands Gia Lai 0.1 0.09 2.5 0.80 9.7 2.11 23.3 2.72 35.4 3.30 29.0 4.44 Dak Lak 0.4 0.23 1.3 0.49 7.5 1.58 24.7 2.77 33.1 2.81 33.0 4.50 Ho Chi Minh 0.1 0.07 0.2 0.11 3.0 0.58 22.2 2.49 38.7 2.20 35.9 3.78 Lam Dong 0.1 0.10 1.0 0.34 9.4 1.75 24.0 2.55 33.8 2.77 31.6 4.49 Ninh Thuan 0.4 0.18 9.4 1.69 23.4 2.19 33.3 2.19 21.9 2.35 11.7 2.66 Binh Phuoc 0.3 0.18 5.4 0.93 18.0 1.96 38.8 2.32 28.8 2.64 8.7 2.02 Southeast Tay Ninh . . 2.5 0.66 14.9 2.21 40.0 2.85 31.8 2.89 10.7 2.28 Binh Duong . . 1.2 0.41 7.0 1.20 33.2 3.01 36.9 2.81 21.7 3.51 Dong Nai . . 3.1 0.85 10.6 1.54 36.0 2.26 35.2 2.33 15.1 1.91 Binh Thuan . . 6.0 1.31 19.2 2.16 40.1 2.42 23.4 1.78 11.2 2.33 Ba Ria - Vung Tau . . 2.1 0.73 7.5 1.17 31.3 2.50 37.6 2.17 21.5 2.81 Long An 0.1 0.09 2.8 0.69 13.4 1.81 38.6 2.77 31.5 3.11 13.6 2.69 Dong Thap 0.3 0.22 6.6 1.38 19.7 2.27 37.3 3.35 21.1 2.63 15.1 3.64 An Giang . . 2.9 0.82 18.3 2.52 38.1 3.16 27.4 3.58 13.4 3.35 Tien Giang . . 1.0 0.39 10.0 1.85 32.7 3.18 33.9 2.96 22.4 4.13 Vinh Long . . 2.6 0.64 13.6 1.62 35.4 2.49 31.6 2.56 16.8 2.99 Ben Tre . . 1.3 0.31 12.2 1.72 38.7 2.92 31.7 2.94 16.1 3.09 Mekong Kien Giang 0.7 0.32 11.9 1.77 26.4 2.92 35.7 2.67 16.6 2.59 8.8 3.17 Delta Can Tho 0.1 0.11 6.3 1.23 24.0 2.71 40.5 2.73 21.4 2.90 7.8 2.29 Tra Vinh 0.2 0.17 7.6 1.70 24.3 2.59 42.0 3.04 19.1 2.47 6.8 2.24 Soc Trang 0.4 0.18 12.7 2.01 27.8 2.38 40.3 2.68 15.4 2.68 3.5 1.44 Bac Lieu 0.3 0.17 12.2 1.88 26.0 2.69 33.1 3.06 11.1 1.82 17.3 5.22 Ca Mau . . 5.8 1.12 22.2 2.90 34.9 2.80 22.2 3.00 14.9 4.40 Vietnam 0.2 0.02 3.5 0.13 11.5 0.27 28.2 0.37 29.7 0.41 27.0 0.60 34 Vietnam Reading and Mathematics Assessment Study Table 2.5: Percentages and sampling errors of pupils at different skill levels in Reading and Mathematics by school location Reading Skill Levels (pupil) Math Skill Levels (pupil) Region Area Level 1 Level 2 Level 3 Level 4 Level 5 Level 6 Level 1 Level 2 Level 3 Level 4 Level 5 Level 6 % SE % SE % SE % SE % SE % SE % SE % SE % SE % SE % SE % SE Isolated 2.3 na* 12.6 na 26.8 na 19.2 na 34.0 na 5.1 na . . 1.5 na 9.3 na 35.5 na na na 32.5 na Rural 2.3 0.25 9.7 0.59 21.5 0.86 21.3 0.77 29.2 1.16 16.1 1.34 0.1 0.03 2.4 0.30 8.2 0.58 24.7 1.02 30.9 1.04 33.7 1.94 Red River Delta Urban 0.4 0.17 3.3 0.53 10.7 1.25 19.1 1.23 34.3 1.53 32.2 2.19 0.1 0.06 0.3 0.12 2.3 0.46 10.5 1.15 29.9 1.82 57.1 2.40 Total 1.9 0.20 8.6 0.50 19.7 0.75 20.9 0.69 30.1 0.98 18.8 1.24 0.1 0.02 2.0 0.25 7.2 0.45 22.4 0.87 30.6 0.92 37.8 1.75 Isolated 12.4 1.60 22.0 1.98 20.5 1.56 17.5 1.55 15.2 1.72 12.4 2.57 0.4 0.17 8.4 1.34 18.1 1.73 28.9 2.14 22.5 2.53 21.8 3.47 Rural 5.0 0.43 15.1 0.79 22.1 0.92 20.0 0.78 24.4 1.11 13.5 1.17 0.2 0.05 4.3 0.40 12.4 0.69 25.9 1.18 25.9 1.19 31.4 1.71 Northeast Urban 1.0 0.25 7.3 1.15 14.1 1.66 17.8 1.20 33.2 2.55 26.6 3.17 0.0 0.02 1.0 0.27 5.2 0.92 17.0 1.97 30.3 2.39 46.4 3.82 Total 5.5 0.40 15.0 0.59 20.6 0.74 19.2 0.60 24.3 0.80 15.4 1.00 0.2 0.04 4.4 0.34 12.2 0.52 24.9 0.84 26.0 1.00 32.3 1.34 Isolated 14.0 3.09 21.4 2.78 24.8 3.41 18.7 3.38 14.8 3.73 6.3 2.83 2.0 0.74 12.9 3.13 16.6 2.81 25.0 3.75 23.1 4.35 20.5 4.94 Rural 7.1 1.45 16.9 2.04 23.3 2.33 17.1 2.04 26.2 3.36 9.3 2.01 0.3 0.17 6.9 1.65 13.3 1.70 24.1 2.75 22.9 2.79 32.6 4.49 Northwest Urban 2.5 1.00 6.4 1.99 19.6 3.40 23.2 3.96 33.9 3.39 14.4 2.59 . . 2.6 1.09 5.7 1.31 21.2 3.99 34.3 3.59 36.2 4.96 Total 8.9 1.30 16.9 1.39 23.3 1.72 18.6 1.65 23.4 2.22 9.0 1.44 0.8 0.28 8.3 1.37 13.3 1.39 24.0 2.12 24.6 2.22 28.9 3.07 Isolated 8.0 2.39 16.3 3.24 19.9 3.34 16.1 2.39 27.4 5.00 12.4 4.62 . . 2.9 1.20 11.8 2.85 21.4 4.36 30.4 4.54 33.4 6.94 Rural 4.2 0.72 12.3 1.21 23.7 1.28 20.4 0.99 27.7 1.97 11.8 1.73 0.2 0.14 2.2 0.45 7.7 0.82 25.6 1.64 33.1 1.68 31.2 2.82 North Central Urban 1.1 0.37 7.0 1.86 14.5 1.78 23.5 1.59 33.8 2.46 20.1 3.46 0.0 0.04 1.5 0.75 4.2 1.61 18.9 2.22 35.6 2.90 39.9 4.55 Total 4.1 0.59 11.9 0.93 22.0 1.12 20.4 0.83 28.6 1.65 13.0 1.44 0.1 0.11 2.2 0.36 7.6 0.68 24.2 1.36 33.2 1.46 32.7 2.32 Isolated 13.8 4.42 23.8 3.09 24.7 2.63 15.8 2.70 15.4 2.69 6.5 3.05 . . 4.9 1.66 18.0 3.12 33.5 3.07 32.0 3.49 11.7 2.77 Rural 3.7 0.38 17.2 1.02 28.1 1.13 21.6 0.82 21.7 1.36 7.8 1.09 0.0 0.03 2.1 0.35 10.3 0.85 34.3 1.32 33.6 1.32 19.7 1.75 Central Coast Urban 1.5 0.43 7.8 1.05 18.9 1.83 22.1 1.60 28.4 1.92 21.3 3.40 . . 1.0 0.31 5.6 0.96 19.7 1.74 37.1 1.83 36.7 3.29 Total 4.4 0.67 15.6 0.78 25.4 0.88 21.0 0.70 22.6 1.05 11.0 1.23 0.0 0.02 2.2 0.30 10.0 0.69 30.5 1.13 34.3 1.01 23.0 1.60 Isolated 14.0 3.38 19.3 3.53 23.9 3.42 18.9 4.27 17.4 3.77 6.6 2.77 0.7 0.31 5.9 1.75 13.1 2.78 28.6 4.84 28.2 3.92 23.6 6.36 Rural 5.9 1.32 13.0 2.42 19.6 2.45 20.3 2.90 26.6 3.59 14.6 3.82 0.3 0.25 2.5 0.69 10.6 2.06 24.9 2.75 32.1 3.46 29.6 5.01 Central Highlands Urban 1.4 0.57 6.2 1.38 17.9 2.71 23.5 2.10 35.0 3.30 16.0 4.59 0.2 0.18 1.2 0.44 3.4 0.68 20.6 2.63 37.2 3.22 37.5 5.18 Total 6.4 1.07 12.5 1.48 20.1 1.59 20.9 1.78 26.9 1.93 13.2 2.40 0.3 0.15 2.9 0.54 9.2 1.21 24.5 1.84 32.6 2.18 30.5 3.10 Isolated 4.6 1.29 17.0 2.48 30.1 2.38 22.4 2.58 19.9 2.30 6.0 1.37 0.1 0.07 3.9 0.99 15.6 2.00 38.7 2.30 25.9 2.37 15.8 3.17 Rural 3.3 0.51 14.7 1.06 27.3 1.01 23.9 1.08 21.5 0.87 9.3 1.22 0.1 0.05 3.7 0.42 13.7 0.96 37.2 1.31 30.9 1.13 14.4 1.86 Southeast Urban 1.4 0.22 6.7 0.63 20.7 1.45 21.2 1.11 30.6 1.35 19.5 1.82 . . 0.9 0.23 4.7 0.60 22.2 1.44 39.2 1.71 33.0 2.21 Total 2.6 0.29 11.7 0.66 24.9 0.94 22.6 0.77 25.1 0.77 13.1 1.06 0.1 0.03 2.6 0.23 10.2 0.59 31.2 1.04 33.7 0.96 22.3 1.48 Isolated 7.6 1.08 25.5 1.68 30.4 2.11 16.4 1.17 14.2 1.89 6.0 1.59 0.4 0.18 7.0 1.11 21.0 1.63 40.6 2.50 22.0 2.92 9.0 2.15 Rural 8.3 0.49 24.5 0.88 28.6 0.64 18.1 0.66 15.1 0.70 5.5 0.74 0.2 0.04 6.3 0.35 21.4 0.80 38.2 1.00 23.6 0.91 10.5 0.91 Mekong Delta Urban 3.6 0.56 14.6 1.60 21.1 1.55 19.4 1.40 24.6 1.84 16.8 2.39 . . 2.9 0.67 11.3 1.36 31.1 2.42 28.0 1.71 26.7 3.26 Total 7.4 0.36 23.0 0.74 27.6 0.56 18.1 0.59 16.6 0.65 7.5 0.67 0.2 0.04 5.8 0.32 19.6 0.68 37.3 0.83 24.1 0.90 13.0 0.93 Isolated 9.7 0.73 21.1 0.93 25.5 0.92 17.9 0.80 17.6 1.04 8.2 0.96 0.4 0.09 6.4 0.56 17.0 0.82 32.6 1.38 25.2 1.36 18.5 1.69 Rural 4.7 0.21 15.4 0.37 24.5 0.39 20.5 0.33 23.8 0.47 11.2 0.45 0.1 0.03 3.6 0.15 12.4 0.33 29.8 0.47 29.1 0.51 25.0 0.75 Vietnam Urban 1.5 0.14 7.6 0.42 17.5 0.51 20.7 0.53 30.9 0.63 21.7 0.88 0.0 0.02 1.3 0.15 5.4 0.43 20.6 0.67 34.2 0.80 38.5 1.12 Total 4.6 0.17 14.4 0.28 23.1 0.34 20.2 0.27 24.5 0.39 13.1 0.41 0.2 0.02 3.5 0.13 11.5 0.27 28.2 0.37 29.7 0.41 27.1 0.60 * very few schools in the isolated category in the Red River Delta and hence errors are not applicable 35 Vietnam Reading and Mathematics Assessment Study What percentages of pupils reached the benchmarks levels in reading and mathematics? Benchmarks Isolated schools showed In addition to the distribution of skills within school location and region, it is consistently lower instructive to examine how well pupils are prepared at the end of primary school achievement levels in to enter the community as independent citizens or to begin their lower secondary mathematics and education and expect to be independent learners. It was also important to see reading.. how well this is achieved in different regions and provinces in Vietnam. Some provinces, In order to do this, the benchmarks were also examined at a) the national especially in the Mekong level, b) the regional level, and c) the provincial level, as well as by d) school Delta, exhibited high location. As a brief overview, the findings showed that there is a potentially proportions of pupils not serious issue emerging for reading. About 50 percent of pupils have been identified as not reaching the second benchmark (where they could be reaching the described as having the reading and mathematics skills to enable independent independent learning learning in Grade 6) despite the fact that they had reached the first benchmark benchmark of functional reading and mathematics levels. The figure for mathematics was 20 percent. Less than 40 percent of pupils were identified as having independent learning skills in isolated areas in the Northeast, Northwest, Central Coast, Central Highlands, and Southeast regions, and in isolated and rural areas in the Mekong Delta. Less than 30 percent of pupils in provinces in the Mekong Delta such as Kien Giang, Tra Vinh, Soc Trang, and Bac Lieu had reached the level of independence for reading. Again these areas are in isolated areas and the areas where higher proportions of pupils are prepared for independent learning in secondary school are largely concentrated in highly industrialised modern locations in Vietnam. a) At the national level From Table 2.6 it can be seen that 10.7 percent of pupils had not yet reached a level of functional benchmark in reading and 2.8 percent had not reached the functional benchmark in mathematics. The standard errors of sampling (SE) for these figures were 0.30 and 0.13. The sampling in this study was so designed that it was possible to estimate these percentages as actual numbers of pupils. The equivalent numbers for 10.7 percent in reading were 205,763 and for 2.8 percent in mathematics were 54,739. The results have been graphically presented in Figures 2.2 and 2.3 below. There is an obvious discrepancy in the how well pupils are prepared in mathematics and reading. Pupils are expected to have high levels of reading skills in order to enter society as an independent citizen and much higher levels of reading skills in order to enter secondary education as independent learners. The demand placed on them with respect to mathematics is not as high and this is consistent with world-wide experience of studies in literacy and numeracy. In Vietnam the importance of language in the culture and the efforts of schools to maintain that culture is indicative of the reading demands identified in this study. 36 Vietnam Reading and Mathematics Assessment Study Table 2.6: Percentages and sampling errors of pupils reaching functionality levels in reading and mathematics Read Math Functionality % SE % SE Independent Reached the level of reading and mathematics to enable independent learning in 51.3 0.58 79.9 0.41 Grade 6 Functional Reached the level for functional participation in Vietnamese society 38.0 0.45 17.3 0.36 Pre functional Not reached the level considered to be a minimum for functional purposes in 10.7 0.3 2.8 0.13 Vietnamese society Figure 2.1: Percentages of pupils at Figure 2.2: Percentage of pupils at different different functional levels of reading functional levels of mathematics b) At the regional level The differences between regions in terms of pupil benchmark levels in reading and mathematics have been presented in Table 2.7. Table 2.7: Percentages and sampling errors of pupils at each benchmark by region Reading Maths Region Pre functional Functional Independent Pre functional Functional Independent % SE % SE % SE % % SE % SE % SE % Red River Delta 5.0 0.37 31.6 1.10 63.4 1.35 95.0 1.7 0.24 11.2 0.67 87.1 0.83 98.3 Northeast 12.0 0.63 34.8 0.95 53.2 1.13 88.0 3.6 0.32 18.0 0.72 78.4 0.88 96.5 Northwest 16.6 1.92 38.6 2.26 44.9 2.79 83.5 7.8 1.42 19.3 1.82 72.9 2.72 92.2 North Central 8.8 0.95 35.7 1.52 55.5 2.09 91.2 1.8 0.40 12.0 1.00 86.3 1.22 98.2 Central Coast 10.9 0.91 41.2 1.23 48.0 1.65 89.1 1.6 0.24 15.5 0.85 82.9 0.96 98.4 Central Highlands 12.2 1.78 33.9 2.16 53.9 2.95 87.8 2.9 0.60 13.7 1.59 83.5 2.05 97.1 Southeast 7.0 0.56 39.9 1.34 53.1 1.51 93.0 1.9 0.21 15.9 0.78 82.2 0.85 98.1 Mekong Delta 17.6 0.66 46.3 0.81 36.1 1.06 82.4 4.6 0.30 28.6 0.86 66.8 0.93 95.4 Vietnam 10.7 0.30 38.0 0.45 51.3 0.58 89.4 2.8 0.13 17.3 0.36 79.9 0.41 97.2 37 Vietnam Reading and Mathematics Assessment Study Fewer pupils prepared It can be seen that the national benchmark levels were generally high. In for reading demands of reading the percentage judged to be functional was 89.3 (38.0 + 51.3) and in secondary school where mathematics the percentage was 97.2 (17.3 + 79.9). Eighty percent were text types is varied and deemed to have the reading and mathematics skills needed to enter secondary demanding. school and expect to be independent learners in mathematics. The figure for Mathematics reading was 51 percent identified as having the reading skills needed for expectations are not as independent learning in secondary school. The difference is not because the demanding. skills in reading are weaker, but perhaps because the change from primary to secondary curriculum is more abrupt in reading than it is in mathematics. Hence the preparation for secondary learning in mathematics is more aligned with the secondary curriculum than is the case with reading. It is also true that reading demands in society are more extensive than are mathematics demands and this is also reflected in the lower benchmark figures. Only 10.7 percent and 2.8 percent of pupils in reading and math respectively were found not to be able to function at a minimal level in Vietnamese society. The definition of functional used in this study was "the level of reading (or mathematics) comprehension needed to function in Vietnamese society effectively ". This did not mean that they were illiterate or non-numerate! The expert panels that worked on these definitions took into account the school curriculum, the performances on the test and the types of mathematics and reading that were assessed in the Grade 5 tests. These were compared to the kinds of reading that a citizen would need to be able to do in Vietnam to function in the community. A score was then set, using the Angoff procedure, which separated those considered to be functional from those who were pre functional level and who would struggle in literacy and numeracy tasks required on a day to day basis. A similar procedure was followed for mathematics in identifying the level of numeracy needed in Vietnamese society. The benchmark levels The point was made above that the test was assessing the old and new were the same for the old curricula and that expository and document texts were not emphasised in the and the new curriculum primary reading curriculum. From Table 2.8 it can be seen how the Angoff items in the test. procedure applied to the items representing the two parts of the test yielded percentage scores for the two benchmarks. As can be observed the expectations in terms of the benchmarks did not change. Accordingly in the remainder of this report the full test result has been used for reporting purposes. Table 2.8: Comparison of Benchmark Scores after Allowing for Old and New Curriculum Item set all all old new old new old new Benchmark Functional Independent Functional Independent raw score 36.70 20.90 25.00 32.00 15.90 20.80 9.30 12.70 Proportion 0.64 0.37 0.64 0.65 0.37 0.40 38 Vietnam Reading and Mathematics Assessment Study In the Mekong Delta region and in the Northeast, Northwest, Central Coast, and the Central Highlands more than 10 percent of pupils were still classified as being below this functional benchmark in reading. In reading, the situation for promotion to lower secondary school was serious. In the Mekong Delta only 36 percent of pupils were categorised as reaching the independent learning benchmark. In mathematics the figure was 67 percent. By the end of primary school it could be reasonably expected that all pupils would have sufficient reading skills to engage in everyday activities of the Vietnamese community. It should also be expected that at least 95 percent of pupils would have reached the level of competence at which they could be expected to learn independently in lower secondary school. Indeed, it was this figure that was set as the target for school principals. Hence is it reasonable to expect that the pupil achievement level would also be at that level for them to learn without remedial assistance in Grade 6, the first year of lower secondary. In several regions and provinces (see Table 2.9 and the Grade 5 Study of Reading and Mathematics: Provincial Tables, 2002) there was a discrepancy between the reported number of pupils transferring from Grade 5 to Grade 6 (i.e., starting lower secondary education) and the number of Grade 5 pupils reaching an "independent" learning level. The MOET and provincial educational authorities may wish to re-examine the design and conduct of the examination at the end of Grade 5 if this discrepancy is linked to the amount and type of assistance needed by pupils not reaching the second benchmark. This might be resolved by moving more to a competency based interpretation of test information so that teachers, pupils and parents would have better information on which to base the progression from primary to secondary education and provide targeted assistance to those pupils who need assistance in reading and mathematics. Moving to a competency based interpretation of assessments at the end of secondary school would assist all people involved in helping the pupil learn. Policy suggestion 2.5: The MOET might wish to consider accelerating the shift in the primary school reading curriculum to ensure that more pupils are adequately prepared for the wider range of genre encountered in secondary school learning materials. c) At the provincial level 39 Vietnam Reading and Mathematics Assessment Study The percentages of pupils reaching specific levels for each of the benchmarks within provinces have been shown in Table 2.9. Where pre- functional rates exceeded 20 percent, the figures have been presented in bold. Where the percentages deemed to be at a sufficient level to progress to Grade 6 were less than 30 percent, then the figures have also been presented in bold. There were large There were differences among regions in achievement and there were proportions of low also large differences among provinces. Even within the highest achieving pupils and performing region, there was considerable variation among provinces. small proportions of high performing pupils Ha Nam and Ninh Binh provinces within the Red River Delta region had concentrated in specific about 10 percent of pupils in reading who were pre functional and some provinces 50 to 60 percent of pupils had not yet reached the independent benchmark. The provinces of Ha Noi and Thai Binh had about two percent of pupils who were pre functional and only about 25 percent of pupils not yet independent. The situation was worse, again in reading, in the Mekong Delta region where overall performance levels tended to be low. In this region pre- functional rates rose to more than 20 percent in a number of provinces and the proportion of pupils with independent reading and mathematics learning skills for lower secondary school was low across all provinces in the region. The situation in mathematics was much better and it was only in the provinces of Hoa Binh, Kon Tum, Soc Trang, and Bac Lieu that there were over 10 percent of pupils categorised as pre-functional. In Vietnam 80 percent of pupils were categorised as possessing reading and mathematics skills for independent learning in Grade 6 but the percentages were somewhat low in several provinces in the Mekong Delta. There is clearly a very wide range of variation in the extent to which pupils are prepared to enter the community or to enter secondary education. Just as was the case with the competency levels, the benchmark data illustrates that there are several provinces where pupils have not attained a level that can assist them to enter society with reading and math skills needed by independent citizens. Once again these are concentrated in the same regions in the Northwest, North East, Central Highlands and in the Mekong Delta. It is evident that in these regions the capacity of people to participate in he community and to make informed decisions must be limited if the level of reading skills remains so low at the end of primary school. 40 Vietnam Reading and Mathematics Assessment Study Table 2.9: Means and sampling errors of Math and Reading scores and percentages of pupils categorised as Pre functional, functional, and able to progress to Grade 6 Reading Maths REGION PROVINCE Pre SE Functional SE Independent SE Pre SE Functional SE Independent SE Ha Noi 2.6 0.70 23.7 2.81 73.7 3.26 0.5 0.31 4.9 1.13 94.6 1.30 Hai Phong 5.5 1.12 35.4 2.92 59.2 3.58 1.5 0.56 12.2 1.89 86.4 2.21 Ha Tay 6.6 1.31 36.8 3.86 56.5 4.57 2.5 0.80 15.8 2.10 81.7 2.61 Hai Duong 4.7 1.03 29.8 3.35 65.5 3.99 1.4 0.46 10.3 2.01 88.3 2.21 Red River Delta Hung Yen 5.2 1.32 26.5 3.15 68.3 3.87 1.4 0.48 8.5 1.70 90.1 2.03 Ha Nam 10.5 1.66 41.2 2.63 48.3 3.48 3.6 0.83 20.7 2.31 75.7 2.86 Nam Dinh 2.7 0.70 25.9 2.71 71.4 3.07 1.0 0.47 7.6 1.37 91.4 1.67 Thai Binh 1.5 0.44 24.4 2.96 74.1 3.14 0.5 0.24 5.5 1.12 94.0 1.19 Ninh Binh 10.9 1.71 51.2 2.48 37.9 3.26 5.5 1.36 23.4 2.49 71.1 3.36 Ha Giang 15.6 2.86 46.8 3.84 37.6 4.74 6.0 1.59 24.2 3.57 69.9 4.62 Cao Bang 25.8 4.51 39.7 3.93 34.5 4.88 4.5 1.63 21.2 3.46 74.4 4.32 Lao Cai 4.4 1.63 25.2 3.34 70.4 4.10 0.6 0.29 5.5 2.14 93.9 2.38 Bac Kan 18.4 3.63 42.5 3.88 39.2 5.12 5.2 1.59 23.6 3.55 71.2 4.57 Lang Son 22.6 3.41 42.7 3.17 34.7 4.58 8.8 1.93 36.0 3.50 55.2 4.58 Tuyen Quang 24.7 3.30 42.3 3.48 33.0 4.60 9.8 1.98 33.4 3.48 56.8 4.52 Northeast Yen Bai 20.9 3.41 40.1 3.28 39.0 4.70 8.5 1.98 26.3 3.17 65.3 4.36 Thai Nguyen 7.4 1.77 30.3 3.54 62.3 4.61 1.6 0.77 14.6 2.84 83.8 3.26 Phu Tho 7.2 1.44 32.4 3.13 60.4 3.86 2.2 1.01 12.2 2.18 85.6 2.79 Vinh Phuc 12.3 1.80 39.4 3.20 48.3 4.36 3.6 0.94 20.1 2.87 76.3 3.45 Bac Giang 11.7 1.91 42.2 2.61 46.1 3.65 1.7 0.64 20.5 2.40 77.8 2.70 Bac Ninh 1.7 0.56 19.5 3.02 78.8 3.35 0.2 0.12 3.4 1.04 96.4 1.13 Quang Ninh 1.5 0.52 19.0 3.31 79.5 3.60 0.2 0.18 4.7 1.17 95.0 1.27 Lai Chau 6.6 2.01 33.3 4.20 60.1 4.90 1.7 1.09 11.3 2.60 87.0 3.29 Northwest Son La 15.8 3.38 38.3 4.19 46.0 5.42 6.5 2.13 18.7 3.22 74.8 4.20 Hoa Binh 22.2 3.58 41.4 3.76 36.3 5.08 12.2 2.73 23.7 2.96 64.1 4.80 Thanh Hoa 10.1 1.79 38.6 3.44 51.3 4.36 1.4 0.42 15.0 2.49 83.6 2.66 Nghe An 11.9 2.68 35.2 3.47 52.9 4.73 3.2 1.29 11.3 2.13 85.5 3.13 Ha Tinh 4.0 1.04 32.2 3.17 63.9 3.76 0.9 0.40 8.5 1.70 90.6 1.87 North Central Quang Binh 3.6 1.26 29.7 3.24 66.7 3.95 0.8 0.37 10.0 2.27 89.2 2.50 Quang Tri 7.9 1.80 41.6 3.22 50.6 3.74 1.8 0.68 12.0 1.80 86.2 2.24 Thua Thien-H 5.0 1.52 32.6 3.12 62.5 3.82 0.5 0.40 8.9 1.70 90.6 1.96 Da Nang 2.6 0.56 23.9 2.51 73.5 2.75 0.3 0.18 6.1 1.40 93.6 1.38 Quang Nam 11.3 1.93 39.2 2.97 49.6 4.06 1.7 0.56 15.1 1.99 83.2 2.31 Quang Ngai 16.2 2.43 44.8 3.20 38.9 4.18 3.0 0.65 20.6 2.55 76.4 2.93 Central Coast Binh Dinh 9.7 2.55 43.3 3.54 46.9 4.31 1.4 0.69 13.7 2.35 85.0 2.82 Phu Yen 10.4 1.54 42.0 2.68 47.6 3.40 2.1 0.53 21.3 2.32 76.6 2.65 Khanh Hoa 11.0 1.47 46.0 2.16 43.0 2.91 0.7 0.33 13.8 1.67 85.5 1.82 Kon Tum 27.3 5.08 36.6 4.10 36.0 4.88 13.2 3.05 25.7 3.53 61.2 5.55 Central Highlands Gia Lai 14.2 3.44 32.4 3.58 53.4 4.87 2.1 0.65 13.7 2.46 84.3 2.92 Dak Lak 8.8 1.80 34.2 3.12 57.1 4.06 1.6 0.69 11.7 2.12 86.7 2.53 Ho Chi Minh 2.7 1.06 31.4 2.84 65.8 3.18 0.2 0.16 5.7 1.08 94.1 1.12 Lam Dong 8.1 1.78 38.1 3.19 53.8 3.90 0.8 0.27 13.0 2.14 86.2 2.34 Ninh Thuan 19.7 2.70 46.5 2.67 33.8 3.46 7.8 1.60 32.5 2.74 59.7 3.54 Binh Phuoc 12.4 1.82 50.7 2.31 36.9 2.99 4.6 0.90 26.2 2.49 69.2 3.14 Southeast Tay Ninh 12.6 1.97 47.3 3.14 40.1 3.82 1.4 0.41 24.2 3.05 74.3 3.30 Binh Duong 4.3 0.99 37.2 2.94 58.5 3.41 0.6 0.24 13.2 1.97 86.2 2.09 Dong Nai 5.1 1.20 40.5 2.22 54.5 2.68 2.1 0.64 16.7 2.09 81.3 2.36 Binh Thuan 10.9 1.92 50.8 2.88 38.3 3.45 4.4 1.07 29.3 2.80 66.3 3.36 Ba Ria - Vung Tau 4.0 0.74 39.5 2.77 56.5 3.18 1.3 0.68 11.9 1.68 86.8 2.01 Long An 11.0 1.84 45.8 2.60 43.2 3.52 2.0 0.53 20.8 2.75 77.3 3.05 Dong Thap 15.1 2.13 46.9 3.32 38.0 4.38 6.0 1.34 28.3 2.77 65.7 3.63 An Giang 19.1 2.89 48.2 3.39 32.7 4.23 1.9 0.58 27.3 3.38 70.8 3.59 Tien Giang 7.6 1.44 43.0 3.41 49.4 4.17 0.7 0.24 16.8 2.65 82.6 2.79 Vinh Long 12.3 1.90 40.6 2.89 47.1 3.76 1.6 0.43 21.1 2.46 77.3 2.75 Ben Tre 8.1 1.38 44.4 2.85 47.6 3.39 0.8 0.24 18.7 2.50 80.5 2.53 Mekong Delta Kien Giang 22.7 3.14 49.8 2.90 27.5 4.09 9.6 1.73 37.5 3.34 52.9 3.98 Can Tho 21.2 2.77 46.2 3.31 32.6 4.44 4.5 0.93 34.6 3.20 60.9 3.66 Tra Vinh 26.9 2.91 49.8 2.67 23.4 3.65 6.1 1.55 33.5 3.16 60.4 4.14 Soc Trang 27.6 3.43 48.6 3.08 23.8 4.03 10.8 1.75 40.3 2.84 48.9 4.00 Bac Lieu 25.5 2.85 44.8 3.68 29.8 5.32 10.2 1.67 36.8 3.49 53.0 4.59 Ca Mau 19.0 2.47 46.3 3.10 34.7 4.43 4.5 0.83 30.8 3.48 64.7 4.05 Vietnam 10.7 0.30 38.0 0.45 51.3 0.58 2.8 0.13 17.3 0.36 79.9 0.41 41 Vietnam Reading and Mathematics Assessment Study Closer scrutiny of It can be seen in the data that there is a potentially serious issue emerging. progress in reading and About 50 percent of pupils have been categorised as not being of such a level mathematics is needed in reading as to be able to cope independently in Grade 6 despite the fact that using a system of they had attained functional reading levels. The figure for mathematics was national benchmark and 20 percent. In the Mekong Delta, more than 70 percent of pupils fell into this competency levels. category for reading and just over 60 percent for mathematics. Using the benchmark performances set by the experts panels in mathematics and reading all regions may need to examine the progress of pupils. In terms of functional reading and mathematics the situation was somewhat more optimistic. d) By school location There was a clear From the foot of Table 2.10, it can be seen that 19.5 percent of pupils in relationship between isolated areas were pre- functional in reading but this figure was only 5.7 school isolation and percent for mathematics. The equivalent figures for rural areas and urban achievement in reading areas were respectively 11.0 and 4.4 percent for reading and 2.9 and 1.0 and mathematics percent for mathematics. It is clear that isolation is related to achievement - especially in North West, and an examination of other factors that arise from isolation and are, in turn, Central Highlands and related to achievement have been examined in later chapters. In every region, Mekong regions the pupils in isolated schools had the lowest performance levels in reading and mathematics. Isolated schools also had low percentages of pupils with the reading and mathematics skills for independent learning in secondary school. In the isolated locations of some regions (notably the North West, Central Highlands and Mekong Delta regions) little more than a quarter of the pupils were deemed to have those skills. Figure 2.4: Relationship between school location and functionality level of achievement 42 Vietnam Reading and Mathematics Assessment Study The relationship between school location and achievement has been presented in Table 2.10. There was a clear relationship between isolation and pre functional levels. There was also a clear relationship between urban areas and the chances of having higher reading and mathematics skills before entering secondary education. However, it is worth recalling that functionality was also defined by specialists in an urban context and the classification may need to be interpreted with this in mind. It is possible that reading and mathematics demands differ in different contexts, but the overall lack of skills should still be regarded with concern. Table 2.10: Percentages and sampling errors of pupils at Functionality levels by school location Reading Mathematics REGION Location Pre Function Independent Pre Function Independent % SE % SE % SE % SE % SE % SE Isolated 6.7 1.45 41.7 11.10 51.6 10.87 0.7 0.96 20.8 8.00 78.4 8.48 Rural 5.7 0.46 34.5 1.23 59.8 1.51 2.0 0.29 12.8 0.80 85.2 1.00 Red River Delta Urban 1.5 0.34 17.4 1.64 81.2 1.78 0.4 0.16 3.2 0.61 96.4 0.67 Total 5.0 0.37 31.6 1.10 63.4 1.35 1.7 0.24 11.2 0.67 87.1 0.83 Isolated 22.9 2.40 37.6 2.64 39.5 3.80 7.2 1.30 26.6 2.25 66.2 3.03 Rural 11.4 0.78 36.8 1.14 51.8 1.43 3.3 0.36 18.4 0.97 78.3 1.18 Northeast Urban 3.3 0.60 23.7 2.17 73.0 2.58 0.9 0.23 7.9 1.31 91.2 1.43 Total 12.0 0.63 34.8 0.95 53.2 1.13 3.6 0.32 18.0 0.72 78.4 0.88 Isolated 24.3 4.45 42.6 4.40 33.1 5.66 13.1 3.39 23.0 3.55 63.9 5.26 Rural 14.6 2.33 38.6 3.15 46.9 3.87 6.0 1.56 19.7 2.36 74.4 3.37 Northwest Urban 4.9 1.91 29.1 4.45 66.0 5.35 1.8 0.62 9.0 1.99 89.2 2.44 Total 16.6 1.92 38.6 2.26 44.9 2.79 7.8 1.42 19.3 1.82 72.9 2.72 Isolated 16.3 3.95 33.4 4.75 50.3 6.66 1.9 0.98 19.0 4.70 79.2 4.99 Rural 8.8 1.19 37.8 1.89 53.3 2.65 1.8 0.51 12.0 1.17 86.2 1.46 North Central Urban 3.5 1.31 26.2 2.71 70.2 3.24 1.5 0.75 7.5 2.10 91.0 2.67 Total 8.8 0.95 35.7 1.52 55.5 2.09 1.8 0.40 12.0 1.00 86.3 1.22 Isolated 24.8 4.93 42.3 4.15 32.9 5.10 3.8 1.39 25.2 3.73 71.0 4.77 Rural 10.7 0.80 45.4 1.47 43.9 1.86 1.6 0.27 16.3 1.05 82.1 1.14 Central Coast Urban 4.7 0.75 29.8 2.49 65.4 3.07 0.7 0.22 8.9 1.22 90.5 1.32 Total 10.9 0.91 41.2 1.23 48.0 1.65 1.6 0.24 15.5 0.85 82.9 0.96 Isolated 23.0 4.86 41.4 4.28 35.7 6.40 5.6 1.63 17.1 3.27 77.3 4.37 Central Rural 12.2 2.57 32.8 3.34 55.0 4.46 2.6 0.88 16.1 2.69 81.4 3.31 Highlands Urban 3.6 0.96 30.1 4.05 66.4 4.55 1.2 0.49 6.4 1.23 92.4 1.56 Total 12.2 1.78 33.9 2.16 53.9 2.95 2.9 0.60 13.7 1.59 83.5 2.05 Isolated 11.1 2.33 49.3 2.80 39.6 3.52 3.2 0.83 24.1 2.99 72.7 3.50 Rural 9.1 0.95 43.9 1.54 47.1 1.86 2.7 0.37 21.4 1.28 75.9 1.44 Southeast Urban 3.5 0.43 32.8 1.88 63.7 2.07 0.6 0.19 7.5 0.80 91.9 0.91 Total 7.0 0.56 39.9 1.34 53.1 1.51 1.9 0.21 15.9 0.78 82.2 0.85 Isolated 19.8 1.83 48.7 2.72 31.6 3.30 6.3 1.06 30.6 2.18 63.1 2.68 Rural 19.1 0.87 48.6 0.92 32.3 1.13 4.8 0.32 31.1 0.99 64.1 1.05 Mekong Delta Urban 9.9 1.30 35.3 2.17 54.8 3.13 2.2 0.55 17.1 1.87 80.7 2.14 Total 17.6 0.66 46.3 0.81 36.1 1.06 4.6 0.30 28.6 0.86 66.8 0.93 Isolated 19.5 1.10 42.9 1.26 37.6 1.70 5.7 0.54 25.0 1.25 69.4 1.45 Rural 11.0 0.38 40.2 0.51 48.8 0.68 2.9 0.15 18.6 0.42 78.5 0.48 Vietnam Urban 4.4 0.33 28.6 0.62 67.0 0.80 1.0 0.13 8.5 0.52 90.4 0.59 Total 10.7 0.30 38.0 0.45 51.3 0.58 2.8 0.13 17.3 0.36 79.9 0.41 43 Vietnam Reading and Mathematics Assessment Study What were the test scores in reading and mathematics? Explanatory Analyses using the Neutral Scale The first two scores have demonstrated that there are serious differences in the distribution of skills within and between provinces and regions. The third measure has been developed to enable exploration and explanation of differences between and within the groups of pupils and teachers. It cannot be interpreted directly in terms of the skills or the preparation to enter independently the community or secondary education. It does however, enable many new analyses to e carried out in order to identify where important differences exist and to use other variables to explain those differences. The test scores were converted to a scale with a mean of 500 and a standard deviation of 100 as explained in Chapter 1. These 500 scores have been examined at a) the regional level, and b) the provincial level, and c) by school location. In summary, the situation is the same as for the competency and benchmark presentation: the lowest scores were found in the isolated school locations and the highest in the urban areas. The Mekong Delta pupils and Northwest pupils had the lowest scores. a) At the regional level If statistically significant differences were to be used in the comparison of means, then many of the differences would be significant. This would be because the sample is very large. It was felt that when mean scores were being compared it would be advisable to ensure that the difference of the means would constitute at least two items on the test. This is the equivalent of 0.20 of a standard deviation or 20 points in the 500 score scale. Thus differences have been reported in terms of 0.20 of a standard deviation. The first comparisons have been presented in Figure 2.5. The symbols and point to the regions with higher score levels and where the difference was more than 20 points on the mean 500 scale. The symbol indicates that the difference was less than 20 points on the mean 500 scale Reading Math Figure 2.5: Total scores in Reading and Mathematics by region 44 Vietnam Reading and Mathematics Assessment Study Reading down a column and across a row enables a comparison of every region with every other region in a series of pairwise comparisons. The symbol ' ' indicates that there is no educationally important difference between the provincial means in that column and row. The arrow head symbol ' ' points to the region with the higher score and hence the table shows that the regions at the top of the left hand column (eg Red River Delta) have a higher average than other regions. The symbol ' ' also points to the regions with the higher average. For example, the mean mathematics performances of pupils in the Mekong Delta region were lower than every other region. The mean mathematics performance of pupils in the North Central region was the same as Red River Delta, Central Highlands, and the Northeast but these had higher scores (more than 0.20 of the overall standard deviation) than Central Coast, Southeast, Northwest and the Mekong Delta. An important interpretation of these figures is based n the spread of the 'dots' (' ') throughout the figure. If the dots cover the entire figure it could be concluded that there was an equitable spread of achievement across regions. If the dots form a thin line across the diagonal, there would be evidence of serious discrepency between regions. b) At the provincial level In the Figure 2.6 (for reading) and Figure 2.7 (for mathematics), a similar series of pairwise comparisons has been made for provinces. It can be seen that there were no important differences in reading scores among the provinces from Quang Ninh to Hung In mathematics the provinces of Bac Ninh, Quang Ninh, Ha Noi and Thai Binh were not different from each other but all were higher (by at least 20 points on the 500 scale) than the other provinces. The symbols 5 , 3 and = in the figures below have the same meaning as they had in the regional results above. Overall the figure can show how equitable the system is in terms of achievement. If the yellow shaded section or middle band with the dots covered the whole chart, the system would be seen to be equitable. The thinner that shaded region in the middle band is, the more inequitable are the achievement levels. Achievement levels appeared to be slightly more equitable in Reading than in Mathematics. However, the narrowness of the middle band of comparisons indicated that, on the whole, the differences in achievement represented a high level of inequity of achievement in the school system. In fact the narrowing of the middle band in mathematics near the higher scoring end of the distribution is an indication of a bimodal distribution of provinces. There appeared to be a small group of high achieving provinces and a much larger group of lower performing provinces in mathematics. In reading this bimodal distribution was not evident, but the narrowing of the middle band at the top of the distribution suggested that there were few high achieving provinces and then a skewed distribution. 45 Vietnam Reading and Mathematics Assessment Study Figure 2.6: Pairwise comparison of provincial pupil reading scores 46 Vietnam Reading and Mathematics Assessment Study Figure 2.7: Pairwise comparison of provincial pupil mathematics score 47 Vietnam Reading and Mathematics Assessment Study c) By school location The '500' scores by school location have been presented in Figure 2.8 and Figure 2.9. Pupils in isolated areas tended to achieve lower scores than pupils in urban areas in any regions. Especially in the Northwest, Central Coast, Central Highlands, and the Mekong Delta, the mean scores of pupils in isolated were low. In the Mekong Delta, not only pupils in isolated areas but also pupils in rural areas had low scores in both reading and mathematics. School location within region was clearly related to reading and mathematics achievement Figure 2.8: Location within region and relationship to pupil reading achievement Figure 2.9: Location within region and relationship to pupil mathematics achievement 48 Vietnam Reading and Mathematics Assessment Study Sub domain scores The reading test consisted of three sub-domains: narrative, expository, and documents. The mathematics test consisted of the following three sub- domains: number, measurement, and space. As seen in Table 2.11, the correlations are relatively consistent within and Pupil performance on across reading and mathematics and this suggests that the six scores may be sub domains within each measuring a single underlying factor. A principal factor analysis was of the tests were highly conducted and this verified the proposition that a single achievement factor correlated underpinned the two tests and the six sub domain scores. The result of the principal factor analysis is shown in Table 2.12. This factor explained more than 65 percent of the test score variance. Table 2.11: Correlations between sub domain scores across reading and mathematics Number Measurement Space/Data Narrative Expository Document Number 1 Measurement 0.81 1.00 Space/Data 0.73 0.80 1.00 Narrative 0.64 0.62 0.58 1.00 Expository 0.65 0.63 0.59 0.79 1.00 Document 0.58 0.54 0.49 0.66 0.68 1.00 Table 2.12: Factor structure of the six sub domain scores. Factor Narrative .768 A single underlying Expository .785 factor of general Document .685 achievement was Number .878 observed Measurement .885 Space/Data .827 While it is important for curriculum specialists to understand the details of performance within each domain, the overall analysis of the data and explanation of differences in achievement for the purposes of intervention focuses on the overall achievement in reading and mathematics. The evidence of a single underlying factor strengthens the argument for the generalised interpretation of the benchmark performances. 49 Vietnam Reading and Mathematics Assessment Study Figure 2.10 Sub Domain Scores by Region for Mathematics and Reading The strong relationship between the sub domain scores has been shown in Figure 2.10. Such sub domain scores have been standardised. The mean500 score has been computed relative to the mean and standard deviation of the overall test resulting with a mean score relative to the overall test mean of 500. Sub tests can be compared on the overall test scale. The means of each sub test were 489, 491 and 469 for Narrative, Expository and Document literacy respectively and 500, 496 and 496 or number, measurement and space sub-domains. A strict interpretation should be based on the sub test relative standardised contribution to the overall reading performance. As such it enables a limited comparison across sub domain scores. As previous discussed, 0.2 standard deviations or 20 points has been used as a metric for educationally important differences. The vertical scale or the figure has been divided into units of 0.2 standard deviations or 20 points on the mean500 scale. The horizontal grid lines are therefore a guide to 'important' differences. It is clear that there are important differences between regions in terms of sub domain scores but no regions have important differences within the region between sub domains. This chart indicates that urgent help is needed in the Mekong Delta region in all domains. Document Literacy was the weakest area of reading, but this 50 Vietnam Reading and Mathematics Assessment Study is not surprising because it is new to the Vietnamese curriculum. Perhaps both curriculum developers and teachers may need professional development in this area of reading. In the Central Highlands there was a large difference between sub-domain scores in Mathematics with space/Data being the higher level performance area. Figure 2.11: Mean 500 scores of sub-domains As seen in Figure 2.11, in mathematics there was no difference among sub- There were no important domain scores. In reading, it would seem that Vietnamese pupils achieved or significant differences worse in 'documents' than in narrative and expository questions by more than between the sub domain 20 points, and it can be said that there is an educationally important scores in reading or difference. At the same time the progress in implementing the new mathematics curriculum is encouraged and curriculum specialists in reading may wish to consider the possibility of re-examining the kinds of reading texts in the Vietnamese textbooks and judge whether or not it would be desirable to introduce some more documents type of reading. What were the differences in achievement in both reading and mathematics between boys and girls and between the different socio-eeconomic groups? The MOET wanted to have information on pupil achievement between boys and girls and between the different socio-economic groups. In essence, in high-achieving regions such as the Red River Delta and the Southeast, girls achieved better than boys in reading. But, there were no differences between boys and girls in mathematics. In both reading and mathematics, pupils from higher socio-economic groups achieved higher scores than pupils from lower socio-economic groups. 51 Vietnam Reading and Mathematics Assessment Study Boys and girls At the national level, girls achieved better in reading. The mean reading score were between 503 and 509 for girls and between 492 and 497 for boys with 95 percent confidence. In mathematics, there were no differences between boys and girls. Gender differences in each region have been presented in Figures 2.12 and 2.13. After taking into account the 20 points on the 500 scale, there were no educationally important differences between boys' and girls' reading test scores in any region. There were no differences in mathematics test scores between boys and girls in any region. Figure 2.12: Boys and girls' mean reading scores by region There were consistent but unimportant differences in the achievements of boys and girls across all regions Figure 2.13: Boys and girls' mean mathematics scores by region 52 Vietnam Reading and Mathematics Assessment Study Socio-economic groups In Vietnam the socio-economic group definition was developed using a range of family background factors. In Table 3.6 in Chapter 3, the variables making up the construct of home background have been given. They were 'possessions in the home', 'parental education', 'pupil ethnic affiliation', 'pupil speaking Vietnamese outside school', and 'number of books in the home'. Each pupil received a weighted factor score for 'home background.' These were arranged in ascending order and split into four groups such that there were about equal number of pupils in each group (see Table 2.13). Table 2.13: Distribution of pupils across socio-economic (SES) levels SES level % SE 1 23.2 0.41 2 23.7 0.31 3 30.3 0.37 4 22.8 0.39 In the group of SES Level 1, 54 percent of pupils came from ethnic Socio economic levels minority groups. They had parents who had an average of 9.4 years of were defined by education, five books, and seven possessions' items at home. In this possessions in the home group, 14 percent of pupils never spoke Vietnamese and 30 percent of pupils sometimes spoke Vietnamese outside school. In the group of SES Level 2, one percent of pupils came from an ethnic minority. They had parents who had 12.8 years of education, five books, and nine possessions' items at home. In this group, all pupils always spoke Vietnamese outside school. In the group of SES Level 3, all pupils came from Kinh and always spoke Vietnamese. They had parents who had 16.4 years of education, 11 books, and 10 items at home. In the group of SES level 4, all pupils came from Kinh and always spoke Vietnamese. They had parents who had 21.9 years of education, 35 books, and 13 possessions' items at home. The mean scores in reading and mathematics by region and by socio- economic levels have been presented in Figure 2.14. In both reading and mathematics, pupils from higher socio-economic groups achieved higher scores than pupils from lower socio-economic groups in any regions. In both mathematics and reading the relationship between family background and achievement was clear. Pupils from poorer families had lower achievement. The link between isolation, poverty and achievement was also clear in reading but somewhat less so in mathematics. 53 Vietnam Reading and Mathematics Assessment Study Low socio- economic groups show persistent lower achievement levels, regardless of region. Figure 2.14: Pupil mean scores and sampling errors for regions by socio-economic levels There were clear differences between socio-economic groups and the differences are educationally important within and between regions. It was noticeable that the pupils in the lowest socio-economic group on average never reached the achievement levels of the pupils in the higher socio- economic group regardless of region. In other words poverty is not overcome in educational terms, no mater where the pupils live. The distribution of the upper and lower 5% of Policy suggestion 2.6: The MOET might wish to consider urgent action pupil performances was for schools in low performing regions where isolation and poverty linked to geographic combine to align with low competency levels. location. Were the pupils 'elite' (upper 5 %) performances similar in different regions and economically advantaged areas? To what extent did the performance 'tails' (bottom 5%) differ across regions and economically advantaged strata? Socio-economic level The upper five percent of pupils were identified by separating all pupils was also related to above the 95th percentile. Then the percentages of pupils who had scored distribution of upper and high enough to be in the national top five percent of pupils in each region lower performing groups by school location and in each region by socio-economic levels was of pupils tabulated. The results of this analysis have been presented in Table 2.14. What does the value of 1.2 percent for the isolated schools in the Red River Delta mean? It means that of all of the pupils in isolated schools in the Red 54 Vietnam Reading and Mathematics Assessment Study River Delta, 1.2 percent achieved high enough to be in the national upper five percent. It can also be seen that 6.2 percent of the rural school pupils and 14.4 percent of the urban school pupils in the Red River Delta were selected in the national upper 5 percent of pupils in reading achievement. This can be compared with pupils in isolated schools in the Mekong Delta, of whom only 0.2 percent achieved high enough to be in the national upper 5 percent of pupils, 2.1 percent of pupils in rural schools, and 5.3 percent of pupils in urban schools. It is clear that pupils in urban areas tended to be in the top 5 percent of pupils. It is also evident that there was a relationship between economic advantage and higher achievement. The trend across regions was consistent. As socio- economic status increased, the representation of the upper 5 percent of pupils in terms of reading achievement increased. In the Red River Delta, 3.2 percent of the pupils in the lowest socio-economic background (or from the poorer homes) were in the upper 5 percent of readers. This contrasted with 13.3 percent of pupils from the most advantaged home background. This pattern was repeated across all regions. In mathematics the pattern was the same. These are the same trends that were observed in the previous sections of this chapter. Pupils in urban areas achieved at a higher level that those in isolated areas regardless of the region or province in which they lived. Table 2.14: Percentages of pupils in the upper 5 percent in reading and mathematics achievement of all pupils by school location. Red River North Central Central Mekong Delta Northeast Northwest Central Coast Highlands Southeast Delta Vietnam Reading Location Isolated 1.2 3.9 1.6 5.1 2.9 1.2 1.6 0.2 2.2 Rural 6.2 4.6 2.8 4.2 2.4 5.0 3.1 2.1 4.0 Urban 14.4 12.4 4.5 7.2 5.4 6.3 8.2 5.3 8.6 SES Level 1 3.2 2.9 2.1 4.8 2.6 2.0 1.5 0.9 2.3 Level 2 3.6 3.0 2.6 3.8 2.0 3.9 2.5 1.8 2.8 Level 3 6.3 6.2 3.3 3.9 1.7 6.7 4.4 2.6 4.6 Level 4 13.3 13.6 6.6 6.5 7.1 7.0 10.8 6.4 9.9 Math Location Isolated 0.62 3.13 2.84 4.96 1.33 2.51 2.32 0.36 2.26 Rural 7.58 6.31 2.60 4.39 1.67 5.94 1.00 0.97 4.17 Urban 13.92 12.80 7.32 6.59 4.03 5.23 4.34 3.23 6.88 SES Level 1 2.7 2.4 2.5 6.4 0.6 1.8 0.8 0.7 2.0 Level 2 4.2 4.5 3.4 3.1 1.3 4.3 0.9 0.8 2.4 Level 3 7.7 8.6 4.4 4.0 1.8 8.2 2.1 0.8 4.8 Level 4 14.8 16.1 11.2 6.2 5.3 7.7 5.8 4.0 9.2 55 Vietnam Reading and Mathematics Assessment Study The combination of In Table 2.15 the percentages of pupils in the lowest 5 percent of all poverty and isolation Vietnamese pupils in reading and mathematics have been presented. In the was associated with low lowest 5 percent of all achievers, pupils from isolated schools dominated and pupil achievement urban schools had low representation. In terms of socio-economic level, pupils from poorer home background were over represented in the lower 5 percent of achievers in both reading and mathematics. Table 2.15: Percentages of pupils in the lowest 5 percent of all pupils in reading and mathematics by school location. Red River North Central Central Mekong Delta Northeast Northwest Central Coast Highlands Southeast Delta Vietnam Reading Location Isolated 2.5 12.3 14.1 8.0 13.9 14.0 4.6 7.6 9.7 Rural 2.3 5.0 7.1 4.2 3.7 5.9 3.3 8.3 4.7 Urban 0.4 1.1 2.5 1.1 1.5 1.5 1.4 3.6 1.5 SES Level 1 4.1 10.5 10.5 8.0 10.8 13.8 6.2 11.9 9.9 Level 2 3.2 3.7 4.3 4.7 3.5 3.1 2.4 6.7 4.3 Level 3 1.8 2.4 2.8 3.7 2.9 2.3 1.7 5.0 2.9 Level 4 0.4 0.9 0.7 1.6 1.4 2.0 0.7 2.9 1.2 Math Location Isolated 1.2 11.4 17.5 3.6 6.8 7.4 5.2 9.2 8.5 Rural 3.2 5.8 9.4 2.9 3.1 3.8 4.6 8.7 5.0 Urban 0.4 1.4 3.5 1.6 1.3 1.6 1.1 4.0 1.7 SES Level 1 7.2 11.0 13.4 6.2 7.0 8.8 6.7 12.9 10.0 Level 2 4.2 5.2 6.0 3.9 3.2 1.5 3.4 7.1 4.7 Level 3 2.4 2.4 4.4 1.6 2.3 2.0 2.2 5.0 2.7 Level 4 0.8 0.8 2.6 1.6 1.2 1.3 0.8 3.2 1.3 What was the relationship between teacher knowledge and pupil achievement? The within province relationships between mean teachers performance and mean pupil performances are also shown in the Figures 2.13 to 2.15. The province where both pupils achievement and teachers knowledge were very low is Lang Son. In Lang Son province teachers were weak in both mathematics and reading skills and this was associated with low pupil performances. In general, the teacher performances in both reading and mathematics tests were related (see Figure 2.14). The consistency with which low performing teachers were identified as related to location and to region in particular is disturbing. Pupils being taught by teachers of low skills in mathematics and reading have a serious handicap that needs to be overcome. 56 Vietnam Reading and Mathematics Assessment Study Teacher and pupil average performance s in reading were related at provincial level Figure 2.13: Relationship between teacher and pupil provincial mean reading scores. Teacher performanc es in reading and mathematic s were on average correlated at province level. Figure 2.14: Relationship between teacher and pupil provincial mean mathematics scores. Pupils from poorer families were taught by teachers with lower levels of mathematics and reading skill. (correlation 0.16 and 0.22 respectively). Analyses reported in Volume 3 show that the relationship was more complex. 57 Vietnam Reading and Mathematics Assessment Study Policy suggestion 2.7: The MOET might wish to consider urgent intervention plans for teachers and pupils with lower performance levels, especially in regions where the combined effects of isolation and poverty are linked with low teacher skills. Summary and conclusions This chapter has dealt with the reading and mathematics of Grade 5 pupils. First this chapter looked at the percentage of pupils at different skill levels. This was followed by a review of the percentages of pupils who could be said to be a) pre-functional, b) able to cope with daily demands of Vietnamese society, and c) of sufficient standard to be able to learn independently in Grade 6. Subsequently, a comparison was made between various sub-groups of pupils on their total scores scaled to a mean of 500 and a standard deviation of 100. The national upper five percent and the bottom five percent of pupils were analysed by school location and by socio-economic levels within each region. Finally, the relationship between teacher knowledge and pupil achievement was examined. Poverty, isolation and Of particular concern were the pupils in isolated schools with low socio- low teacher skills were economic background. They had low scores and were more likely to be pre prevalent among groups functional and not to possess reading and mathematics skills deemed of lower performing necessary for independent learning in Grade 6. The pupils from the lower pupils. Issues of equity may need special socio-economic levels also required special supports. They tended to score attention lower than pupils from the higher socio-economic levels in both reading and mathematics. Pupils from the lowest socio-economic level in the Mekong Delta scored lowest in both subjects. On the other hand, the analysis has shown that there were no significant sex differences in achievement. It would appear that in provinces with teachers of higher knowledge, pupil achievement was also higher. In provinces of lower pupil achievement, the mean teacher knowledge was lower. The major question for government at both the central and provincial level becomes what can be done about the low achieving pupils. In later chapters an examination will be made of the school conditions of the pupils in isolated schools and whether or not they need materials and other resources. 58 Vietnam Reading and Mathematics Assessment Study Chapter 3 WHAT WERE THE BASELINE DATA FOR SELECTED INPUT DATA ABOUT PUPILS IN GRADE 5 IN VIETNAM? Introduction T he aim of this chapter as well as of Chapters 4 and 5 is to present some examples of baseline data for inputs to primary schools in order to establish a descriptive account of the pupils, their teachers and their schools. This chapter is concerned with the pupils. The data are important for two reasons. The first is that they provide a 'context' for the analyses presented later in this report. For example when the differences in the number of resources among provinces and among school within provinces are presented later (see Chapter 7) it is important to link the differences to the actual levels of provision. It is in chapters 3, 4 and 5 that the levels have been presented. The second reason is that, over time, the levels and differences of some variables may well change. Thus when Vietnam, in the future, undertakes another survey of Grade 5 conditions of schooling and pupil achievement, it will be possible to examine the extent to which the 'context' variables have changed. High quality data about 'context' and the 'levels and distribution' provide educational planners and policy-makers with a sound means of mapping the general evolution of the school system and also offer tools for the identification of existing or emerging problems. Policy suggestion 3.1: The MOET might wish to undertake follow-up surveys of Grade 5 at five year intervals in order to establish, inter alia, changes in important educational indicators over time. 59 Vietnam Reading and Mathematics Assessment Study A note on the interpretation of the data analyses Data in this report have Three points should be stressed before the results are presented. The first is been presented by region that the variables/indicators in chapters 3, 4 and 5 are but a small subset of and by school location. the many indicators that could have been collected. The MOET will The same data for each provide a separate publication containing descriptive statistics of all province separately can variables in the study. be obtained from the MOET. The second point, dealt with to some extent in chapters 1 and 2, is that each statistic has been presented together with its standard error of sampling. It will be recalled from Chapter 1 that the sample was drawn in order to yield standard errors of sampling such that a sample estimate of a population percentage would have a very small standard error. The sampling Sampling errors for each procedures have been presented in detail in Chapter 3 of Volume 3 of this estimate in this report series. However, although the standard errors for Vietnam as a whole are have been calculated small the standard errors of sampling for sub-groups of pupils and schools and presented. will be larger. For many of the tables appearing in this report it is the schools in isolated areas, rural areas, and urban areas that have been presented within each region. Where a percentage or a mean has been presented for a sub-group of pupils (such as for pupils in schools in isolated areas within regions) then the standard error will be greater than for the whole sample. The general reason why sub-group estimates have larger standard errors is because the sample sizes for sub-groups are smaller than for the total sample. To illustrate, consider the first column of entries in Table 3.1. The average age of pupils in months at the time the data were collected has been presented separately for each school location in each region and for Vietnam overall. The standard error (SE) of each average has also been presented. For the first entry in Table 3.1, namely isolated schools in the Red River Delta, the average pupil age was 132.0 months and the standard error was 0.67 months. That is, there 19 chances in 20 that the average age of the population of Grade 5 pupils in isolated schools in the Red River Delta was 132.0 months ą 2(0.67). In other words it can be said that we can be 95 percent confident that the population value was between 130.66 and 133.34 months. For all of the Grade 5 pupils in Vietnam we can be 95 percent confident that the population value was between 133.9 and 134.3 months The unit of analysis The third point to note is that the values in all tables have been presented throughout this report in terms of pupils. Pupils were the unit of analysis even though some of the has been the pupil. variables described in Chapters 4 and 5 refer to teachers and schools. Where a percentage refers to a pupil characteristic then this means that it is the percentage of pupils having this characteristic. Where a percentage for a variable refers to teachers then the percentage should be interpreted as 'the stated percentages of pupils were in schools with teachers having the particular characteristic'. Similarly, a percentage for a variable that describes schools should be interpreted as 'the stated percentages of pupils 60 Vietnam Reading and Mathematics Assessment Study were in schools with the particular characteristic'. Where the mean of a variable has been presented for a teacher or school this means that the average pupil in the region or in Vietnam has a teacher or is in a school with the stated characteristic. This concept is important for planning decisions. Take an example of libraries in schools. There are schools of different sizes and it may well be that 50 percent of schools (the small schools) do not have school libraries whereas the large schools have school libraries. But the large schools have 80 percent of Grade 5 pupils in them. Thus the figure presented in the table would be that 80 percent of pupils are in schools with libraries. Since a Ministry of Education is responsible for all pupils then it has been deemed to be more important to know what the pupils as a group are exposed to rather than the actual number of schools. Finally, it should be pointed out that most of the tables, as already mentioned, have been presented in terms of in terms of pupils in schools in different locations in each region of Vietnam. However, many other tables were produced and sometimes use is made of the results from these other tables. They are all to be found in an MOET publication entitled 'Grade 5 Achievement Study: Provincial data' and available from the MOET in Hanoi. Specific policy questions related to pupil inputs This chapter has been divided into four sections as follows: (a) What were the home backgrounds from which the pupils came? (b) To what extent did the pupils receive homework and have help at home? (c) To what extent did pupils receive extra tuition? (d) What were the relationships of pupil background indicators to achievement in mathematics and reading? What were the home backgrounds from which the pupils came? The selected data about home backgrounds have been summarised in two tables - Tables 3.1 and 3.2. (a) Age of Grade 5 pupils Pupils enter school at the beginning of the year in which they become six 61 Vietnam Reading and Mathematics Assessment Study years of age. The school year begins in September. Thus the pupils in Grade 5 in April, 2001 (the date of testing) should have been aged between 124 and 136 months if they had never repeated a grade. The provinces in which the average age exceeded 136 months have been given below. For the most part these are rural and isolated areas. Age Province (months) Mean SE Son La 145.2 1.30 Ca Mau 143.2 0.86 Kon Tum 142.7 1.53 Lao Cai 142.7 1.33 In some provinces, the Cao Bang 142.6 1.31 average age of pupils Bac Lieu 142.1 0.83 was high. Soc Trang 141.7 0.85 Lai Chau 140.4 1.08 Kien Giang 140.3 0.68 Ha Giang 140.2 1.30 Tra Vinh 139.8 0.81 Lang Son 139.5 0.90 Bac Kan 139.2 1.32 Dong Thap 137.8 0.58 Binh Phuoc 137.8 0.61 Can Tho 137.6 0.81 Hoa Binh 137.5 0.83 Gia Lai 137.3 0.84 Dak Lak 136.9 0.71 Tuyen Quang 136.5 0.91 These pupils tended to From Table 3.1 it can be seen that the average age of pupils in isolated come from ethnic schools was much higher than the average ages of those in rural and urban minority homes. schools in all regions except the Red River Delta. In all other regions, the average age of those in isolated schools was higher than 136 months. The highest age in isolated schools was in the northwest region where the average age was 147 months. That is to say that this age was 13 months higher than the average for Vietnam as a whole. In general, there was not very much difference between the average ages for those in rural and urban areas, but this was not the case in the Northwest, Central Highlands, and Mekong Delta regions. 62 Vietnam Reading and Mathematics Assessment Study Table 3.1: The means and sampling errors for selected pupil background measures Age Sex Books at home Possessions Meal per day Parent education Ethnic group at home affiliation Region School location (months) (% female) (number) (number) (number) (years) (% ethnic) Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Isolated 132.0 0.67 53 6.1 9.6 4.11 9.8 1.1 2.8 0.17 16.1 3.29 11 11.2 Red River Rural 130.1 0.09 48 0.6 10.7 0.53 9.7 0.1 2.9 0.01 17.2 0.10 0 0.2 Delta Urban 128.7 0.12 52 1.4 29.6 2.38 12.6 0.1 2.9 0.02 22.0 0.27 1 0.5 Total 129.9 0.08 49 0.6 13.9 0.67 10.2 0.1 2.9 0.01 18.0 0.11 1 0.2 Isolated 141.1 0.90 48 1.3 6.7 0.71 8.0 0.2 2.7 0.03 11.6 0.49 73 3.2 Rural 133.8 0.21 48 0.5 8.7 0.45 9.2 0.0 2.8 0.01 15.6 0.14 29 1.4 Northeast Urban 130.7 0.26 50 1.3 19.5 1.95 11.2 0.1 2.8 0.02 20.1 0.29 16 1.9 Total 134.5 0.22 49 0.5 10.1 0.45 9.3 0.1 2.8 0.01 15.7 0.15 34 1 Isolated 147.2 1.31 43 2.2 6.8 1.20 7.2 0.2 2.6 0.04 9.5 0.61 93 2.5 Rural 139.5 0.72 46 1.3 6.5 0.67 8.2 0.2 2.6 0.04 12.8 0.35 82 2.6 Northwest Urban 132.6 0.77 45 3.1 15.0 1.86 10.5 0.4 2.8 0.03 18.9 0.88 33 6 Total 141.1 0.71 45 1.1 7.9 0.64 8.2 0.2 2.6 0.03 12.6 0.39 78 2.5 Isolated 140.4 1.16 49 2 12.2 2.91 7.6 0.3 2.6 0.06 12.8 0.55 66 6.5 North Rural 132.7 0.26 48 0.9 15.4 1.09 9.6 0.1 2.8 0.02 16.8 0.19 6 1.8 Central Urban 131.2 0.44 46 1.9 25.9 3.69 11.6 0.2 2.9 0.02 20.1 0.41 1 0.3 Total 133.2 0.28 48 0.7 16.6 1.09 9.7 0.1 2.8 0.02 16.9 0.19 11 1.8 Isolated 140.2 2.03 50 2 7.6 1.35 8.0 0.3 2.8 0.05 9.4 0.59 35 7.2 Central Rural 132.1 0.15 49 0.9 14.1 1.09 9.6 0.1 2.9 0.01 13.8 0.20 0 0.1 Coast Urban 131.4 0.27 48 1.3 22.2 1.78 11.6 0.2 2.9 0.02 18.2 0.43 2 0.9 Total 132.9 0.29 49 0.6 15.3 0.86 9.9 0.1 2.9 0.01 14.4 0.23 5 1.1 Isolated 144.2 1.16 44 2.5 8.5 2.43 6.9 0.3 2.8 0.04 11.2 0.78 50 6.1 Central Rural 137.2 0.64 46 1.6 13.1 1.79 8.9 0.2 2.9 0.02 13.8 0.38 26 4.2 Highlands Urban 133.1 0.60 45 2.2 15.4 2.09 11.2 0.2 2.9 0.02 17.8 0.61 8 2 Total 137.6 0.51 45 1.2 12.7 1.27 9.1 0.2 2.9 0.01 14.3 0.33 26 2.6 Isolated 136.7 0.42 47 1.5 10.5 1.25 8.9 0.2 2.7 0.04 11.3 0.47 13 3.1 Rural 135.4 0.20 49 0.6 12.6 0.94 9.8 0.1 2.8 0.01 12.1 0.16 7 1 Southeast Urban 131.7 0.20 47 1.1 23.1 1.60 12.0 0.1 2.8 0.01 17.0 0.36 7 1.6 Total 134.0 0.17 48 0.6 16.7 0.81 10.6 0.1 2.8 0.01 14.1 0.20 8 1 Isolated 140.4 0.48 46 0.9 7.1 0.74 8.3 0.1 2.3 0.03 11.4 0.24 4 1.4 Mekong Rural 138.0 0.24 49 0.6 10.4 0.51 8.7 0.1 2.4 0.01 12.2 0.12 6 0.6 Delta Urban 134.5 0.51 47 1.2 18.8 1.47 10.9 0.2 2.6 0.03 15.8 0.37 6 1.3 Total 137.7 0.17 48 0.5 11.3 0.48 9.0 0.1 2.4 0.01 12.7 0.12 6 0.5 Isolated 140.7 0.33 47 0.7 8.3 0.49 8.0 0.1 2.6 0.02 11.3 0.21 41 2.1 Rural 133.8 0.09 48 0.3 11.6 0.28 9.3 0.0 2.7 0.01 14.9 0.06 10 0.5 Vietnam Urban 131.6 0.13 48 0.6 22.5 0.94 11.7 0.1 2.8 0.01 18.5 0.15 7 0.7 Total 134.1 0.08 48 0.3 13.5 0.30 9.7 0.0 2.7 0.00 15.2 0.05 13 0.4 The high average ages in some areas was probably a result of some pupils either starting school late or repeating several grades or a combination of both. In general, these pupils tended to come from isolated areas and also from minority groups. They often had to travel a distance and had a problem with the Vietnamese language because they did not speak Vietnamese but rather a minority language of their ethnic group. These children were also from families where the parents had not received very much education. The non-Kinh pupils who survived the system as far as Grade 5 were, on average 63 Vietnam Reading and Mathematics Assessment Study 10 months older than the Kinh. The same was true of pupils who spoke Vietnamese only sometimes or never but these tended to be non-Kinh. In general, as will be seen clearly in later chapters, the older children tended to be in schools with fewer material and human resources. It is clear from the above that it is the ethnic minority pupils especially in the isolated areas in the Northeast, Central Highlands and Mekong Delta regions who need to start school earlier and repeat grades less. It was also noted that the schools where the older pupils were located were less well resourced. It could be argued that such schools should be better resourced in order to overcome the deficits of their intake of pupils. The school heads need to consider the possibility of introducing school intervention programs. Policy suggestion 3.2: The MOET may wish to consider increasing the material and human resources to isolated schools. Policy suggestion 3.3: The MOET may wish to consider introducing intervention programs in schools with associated radio programs in order to increase parental interests in their children's education. These programs should involve showing parents how to exhibit their interest in the children's education, have them send their daughters to schools, keep them there, reduce the children's time spent helping the family, ensuring that homework has priority, and trying to provide a private study corner. The radio program should emphasize similar behaviours. The Mekong Delta might be a place to start. (b) Gender distribution In Table 3.1 the percentage of girls in Grade 5 has been given for the different areas in the regions and for the country as a whole. It can be seen that only 48 percent of the pupils in Grade 5 were girls. In the isolated schools in the Northwest, Central Highlands, and the Mekong Delta the percentages were lower. The provinces where there were fewer than 45 percent girls in school at Grade 5 were: 48% of Grade 5 pupils Sex were girls in some Province (% female) provinces only 41-44% Mean SE were girls. There is a Lai Chau 41 2 problem in some areas Lao Cai 43 2.1 of enrolling and keeping Son La 44 1.9 Gia Lai 44 1.8 girls in school. Lai Chau is a poor province with 70 percent ethnic minority people. It is mountainous and there are transport difficulties. 64 Vietnam Reading and Mathematics Assessment Study (c) Books at home The number of books in a pupil's home can be regarded as a reading resource. Most pupils were from From other studies (see, for example, Elley, 1994)1 it has been found that, in homes with very few most countries of the world, the availability of books for children to read is books. highly conducive to better levels of reading achievement. The Grade 5 pupils were asked to indicate the approximate number of books in their home according to six categories: 1 = no books in the home; 2 = 1-10 books in the home; 3 = 11-50 books in the home; 4 = 51-100 books in the home; 5 = 101- 200 books in the home; and, 6 = more than 200 books in the home. The mid point of each value range was used to calculate the total number of books in the home. For example, the value 1 was recoded to zero books, the value 2 was recoded as five books, and so on. The value 6 was recoded, as an estimate, as 250 books. From Table 3.1, it can be seen that the average pupil in Grade 5 came from a home with 13.5 books. The highest figure was for pupils from urban schools in the Red River Delta where the average number of books was 29.6. Pupils from isolated and rural areas had fewer books at home than did pupils from urban areas. Kinh families had more books at home than non-Kinh families. Parents with more years of education had more books in their homes. The provinces with the fewest books at home were have been presented below. Province Books at home (number) Mean SE Lang Son 5.1 0.73 Ninh Binh 5.2 0.57 Bac Giang 6.3 0.53 Tuyen Quang 6.5 0.99 Lai Chau 6.5 0.92 Long An 6.6 0.86 Bac Kan 6.6 1.07 In general, there was a lack of books in the home. As will be seen later there is only equivocal evidence on whether the pupils borrow books from school libraries and take them home. How can children learn not only school subjects but other material if they cannot practice reading? How can citizens read the more complicated arguments about different subjects if they cannot practise reading? For these reasons the benchmark of functionality dealt with in Chapter 2 assumes greater importance. 1Elley, W. B. (1994) The IEA Study of Reading Literacy: Achievement and Instruction in Thirty- Two School Systems. Oxford: Pergamon 65 Vietnam Reading and Mathematics Assessment Study Policy suggestion 3.4: The MOET might wish to consider: a) getting up- to-date information on the availability of supplementary reading materials and library books; b) strengthening the current availability of the three different types of books in school libraries; c) the possibility of having book flood programs in selected schools; and d) encouraging the PTAs to organize the donation of the books to the libraries. (d) Possessions in the home The average pupils came Pupils were asked about the existence of 18 different possessions in their from a home with 10 homes. These possessions were bicycle, motorcycle, clock, study desk, study possessions out of 18. chair, study lamp, newspaper/magazine, books, bookcase, radio, TV set, Urban homes had more wardrobe, electric fan, electric or gas cooker, washing machine, telephone, possessions. air conditioner, and computer. The number of possessions that a pupil said was in his/her home were summed to make a total possessions score. This total was seen as a proxy variable for the wealth of the home The average numbers of possessions of pupils in different locations in each region were presented in Table 3.1 above. The average pupil was from a home with 9.7 possessions. In all regions except for the Red River Delta pupils in isolated areas were from homes with fewer possessions than pupils from rural areas and in turn those in rural areas had fewer possessions than pupils from urban areas. Pupils from schools in isolated areas in the Northwest and Central Highlands regions were from the poorest homes. It is in these areas that there are many pupils from ethnic minority homes. There were also pockets of areas in the Central Highlands and Northwest regions where there was no electricity and some of the possessions in the list in the questionnaire concerned items where electricity is needed. A program is underway to supply electricity to all areas of Vietnam. When this program has succeeded it will be possible to have a computer or two in community centres, if required, to allow the children and adults (including teachers) to access the Internet for much of their information. (e) Number of meals per day Pollitt (Pollitt, E. (1990) Malnutrition and infection in the classroom Paris: Most pupils had 2.7 UNESCO) has shown that poor nutrition results in a lack of concentration meals per day. In some and reduced perseverance in school. Regularity of meals was therefore seen rural provinces, they as a factor likely to influence achievement. It had been hoped to use a had fewer meals. question about the number of breakfasts, lunches and evening meals eaten in a week as has been used successfully in many countries at this age level. However, the Vietnamese questionnaire constructors were of the opinion that Vietnamese Grade 5 pupils would not be able to remember how many meals they had eaten last week. Rather a question was asked about how many meals they ate each day (1 = one meal per day, 2 = two meals per day, and 3 = three or more meals per day). This resulted in a scale from 1 to 3 only. From Table 3.1 it can be seen that the average Grade 5 pupil ate 2.7 meals per day. In 66 Vietnam Reading and Mathematics Assessment Study general, yet again, it was the pupils in the isolated schools who tended to have fewer meals per day. It was in the Mekong Delta that the pupils ate the least number of meals (2.3 and 2.4 meals in the isolated and rural area schools). The provinces with the fewest number of meals eaten by pupils per day were: Meal per day Province (number) Mean SE Ca Mau 2.2 0.04 Dong Thap 2.3 0.04 Bac Lieu 2.3 0.03 An Giang 2.3 0.04 Tra Vinh 2.3 0.03 Vinh Long 2.3 0.03 Kien Giang 2.4 0.03 Can Tho 2.4 0.03 Soc Trang 2.4 0.03 It was of interest to observe that there was no difference between the number of meals eaten per day by Kinh and non-Kinh pupils (2.7 and 2.7 respectively). Policy suggestion 3.5: The provincial offices might like to consider ways of educating parents in the importance of regular meals and nutrition. In order to provide incentives for this, specific provinces may wish to consider a school meals program when poorer children have an increase hours of instruction. (f) Parental education The level of education of parents has been shown many times over to be a major predictor of achievement in many countries. A question was asked about both the mother's and father's education. The coding was as follows and the values for the mother and the father were summed thus making a scale that ranged from 0 to 32. 0 = Did not go to school and had no adult education 5 = Completed primary education 9 = Completed lower secondary education 12 = Completed upper secondary or vocational school 16 = Completed university education 67 Vietnam Reading and Mathematics Assessment Study The average pupil had From Table 3.1, it can be seen that the average Grade 5 pupil in Vietnam had parents who had parents with 15 years of education indicating that between them they had completed primary and completed at least primary school and some of lower secondary school. The some secondary pupils whose parents had most education were to be found in the Red River schooling. Delta urban schools (22.0 years), Northeast urban schools (20.1 years), and North Central urban schools (20.1 years). Those pupils with the least education were to be found in isolated schools in the regions of Northwest (9.5 years) and Central Coast (9.4 years). It was the province of Tay Ninh in the Southeast region where the parents had the least number of years education (9.9 years). (g) Ethnic group affiliation In all there are 54 ethnic groups in Vietnam. Many have their own language but all are instructed in Vietnamese in school. The questionnaire construction group felt that either each group had to be treated separately in the data collection or that only two groups should be formed: Kinh, the major group accounting for 86 percent of all Vietnamese and non-Kinh, all other ethnic minorities as a group. Given that many of the ethnic minority groups account for less than 14 percent of the population, it was decided to use the notion of two groups. It can be seen from Table 3.1 that 14 percent of Grade 5 pupils belonged to ethnic minority groups. In some provinces, more The average percentage of ethnic minority pupils in isolated schools was 41. than half the pupils were In isolated schools in the Northwest region it was 93 percent. The provinces affiliated with ethnic with the highest percentage of ethnic minorities' Grade 5 pupils have been minorities. presented below. Ethnic groups are very diverse. The provinces where the Thai, Tay and Nung ethnic minorities are found are not very poor but nevertheless have quite a few ethnic minority pupils. On the other hand the Hmong and Dao are said to be poor. When viewing the figures presented under the heading of Ethnic minority affiliation it must be borne in mind that it is the ethnic minority pupils who are still in school at Grade 5 level that have been reported. What is not known is how many have dropped out or are repeating a grade. Ethnic minority pupils Ethnic group affiliation tended to be poor homes Province (% ethnic) with parents with little Mean SE education, had to spend Cao Bang 95 1.5 more time helping the Bac Kan 85 2.8 family, had to travel Son La 84 3.4 Lang Son 82 3.2 father to school, and Ha Giang 77 3.3 tended to repeat grades Hoa Binh 76 4.5 more. Lai Chau 70 4.4 Lao Cai 60 5.4 Tuyen Quang 54 4.5 Kon Tum 54 5.6 68 Vietnam Reading and Mathematics Assessment Study What is sure is that the home backgrounds and characteristics of non-Kinh pupils tended to be more disadvantaged than those of Kinh pupils. This can be clearly seen from Figure 3.1 below. Compared with the Kinh pupils, the non-Kinh pupils had fewer possessions at home, parents with fewer years of education, had to spend more time helping families, repeated grades more, had to spend more time to travel to school, and were older. Figure 3.1: Selected pupil characteristics of Kinh and non-Kinh pupils (h) Speak Vietnamese outside school Vietnamese is the language of instruction in school. However, there are certain ethnic groups that speak their own language at home. As already stated, there are 54 ethnic groups. The reading test and mathematics test were in Vietnamese. It was therefore expected that the extent to which pupils speak Vietnamese outside of school would influence their ability to answer the test questions. From Table 3.2 it can be seen that only 3.3 percent of pupils never spoke Vietnamese. In isolated areas however more than ten percent never spoke Vietnamese in the Northeast, North Central, Central coast and Central Highland regions. The situation was particularly bad in the Northwest where 26.8 percent of pupils in isolated areas never spoke Vietnamese, and in rural areas the figure was 24.1 percent. From Table 3.2, it can be seen that in the Northeast only 31 percent of pupils spoke Vietnamese all of the time whereas this was 89 percent for all of Vietnam. The provinces where relatively few pupils spoke Vietnamese were: Cao Bang, Son La, Bac Kan, Lang Son, Hoa Binh, Ha Giang, and Lai Chau had anywhere between 57 percent and 80 percent of children who never spoke Vietnamese. Lai Chau was a little different from the other provinces because the percentages were only 8 percent for never speaking, 50 percent 69 Vietnam Reading and Mathematics Assessment Study for sometimes and 41 percent for always speaking. This might well be because of the high percentage of tourists to that province. As was seen earlier, the percentage of Grade 5 pupils from ethnic minority groups in Hoa Binh and Ha Giang were 76 and 77 respectively. The percentages never speaking Vietnamese outside of the school were different, namely 33 and 16 respectively. It is unclear why this is so. Table 3.2: Percentages, means and sampling errors for selected pupil characteristics Speak Vietnamese Days Absent Private corner at Time working for Time to go to family per day school each day Grade repetition Region School Never (days) home Never location (mins) (mins.) % SE Mean SE % SE Mean SE Mean SE % SE Isolated 0.7 0.96 0.64 0.19 93 5.9 109.5 8.09 17.1 0.44 81 1.9 Rural 0.1 0.10 0.41 0.02 87 0.7 110.0 1.25 15.6 0.17 93 0.4 Red River Delta Urban . . 0.44 0.03 96 0.7 93.5 1.60 12.2 0.32 97 0.6 Total 0.1 0.09 0.41 0.02 89 0.5 107.1 1.03 15.0 0.14 94 0.3 Isolated 19.2 2.01 0.81 0.07 78 2.4 136.5 2.55 24.8 1.00 67 2.1 Rural 6.4 0.62 0.5 0.02 88 0.6 125.9 1.24 19.1 0.26 81 0.7 Northeast Urban 1.5 0.39 0.42 0.04 93 1.1 103.4 1.80 13.6 0.30 90 1.0 Total 7.7 0.53 0.54 0.02 87 0.7 124.0 1.00 19.1 0.22 80 0.5 Isolated 26.8 3.66 0.92 0.10 71 3.9 127.6 3.13 24.6 1.22 63 2.8 Rural 24.1 3.10 0.84 0.11 76 2.3 129.4 2.69 22.0 0.79 65 2.2 Northwest Urban 3.1 1.27 0.6 0.09 88 3.2 103.2 4.41 12.8 0.88 85 2.2 Total 21.9 2.20 0.83 0.06 76 1.9 124.8 1.94 21.6 0.58 68 1.8 Isolated 11.3 2.98 0.89 0.11 66 4.2 115.6 4.02 18.5 0.95 72 2.8 Rural 1.9 0.75 0.73 0.05 88 1.3 125.3 1.63 17.0 0.45 86 0.9 North Central Urban 0.2 0.10 0.56 0.09 94 1.3 98.1 4.21 12.4 0.56 91 1.8 Total 2.6 0.65 0.72 0.04 86 1.1 120.4 1.49 16.5 0.40 85 0.9 Isolated 13.3 4.87 0.97 0.13 61 4 111.4 4.53 19.1 1.19 78 3.6 Rural 0.0 0.02 0.63 0.03 83 1.1 113.7 1.32 14.9 0.27 86 0.6 Central Coast Urban 0.2 0.12 0.5 0.05 87 1.2 93.1 2.70 12.6 0.45 91 0.9 Total 1.7 0.61 0.64 0.03 82 1.1 108.3 1.25 14.8 0.25 86 0.6 Isolated 18.0 3.82 1.42 0.17 61 5 128.2 4.81 22.8 1.87 66 3.1 Rural 7.3 1.70 0.82 0.07 75 3 115.2 2.91 18.7 0.91 70 1.5 Central Highlands Urban 0.6 0.23 0.66 0.07 86 2.9 99.2 2.97 13.3 0.41 85 1.9 Total 7.8 1.19 0.91 0.06 75 1.9 113.6 2.22 18.1 0.66 73 1.4 Isolated 2.8 0.87 0.67 0.06 66 3.5 110.0 4.73 17.0 0.86 71 1.4 Rural 1.9 0.48 0.57 0.03 72 1.7 103.0 1.69 15.3 0.33 79 0.8 Southeast Urban 1.4 0.49 0.42 0.04 81 1.5 86.2 2.04 12.6 0.23 89 0.7 Total 1.8 0.31 0.52 0.02 75 1 96.9 1.20 14.4 0.20 82 0.6 Isolated 1.0 0.68 0.71 0.05 68 2.5 105.3 1.81 20.2 0.55 74 1.6 Rural 1.5 0.28 0.79 0.03 73 1.1 104.7 0.93 20.0 0.27 77 0.7 Mekong Delta Urban 1.0 0.51 0.74 0.07 80 2 94.5 2.49 14.4 0.41 85 1.0 Total 1.3 0.21 0.77 0.02 74 0.9 103.0 0.86 19.1 0.22 78 0.7 Isolated 10.9 1.00 0.84 0.03 69 1.4 118.0 1.52 20.9 0.36 71 0.8 Rural 2.7 0.21 0.61 0.01 82 0.5 114.3 0.59 17.4 0.13 83 0.3 Vietnam Urban 0.9 0.18 0.52 0.02 87 0.7 93.7 0.90 13.0 0.13 90 0.4 Total 3.3 0.19 0.62 0.01 82 0.4 110.4 0.44 16.9 0.10 83 0.3 Finally, it is important to examine the percentages of minority group pupils only who spoke Vietnamese outside of school to varying degrees. This has been shown in Table 3.3 70 Vietnam Reading and Mathematics Assessment Study Table 3.3: Percentages of ethnic minority pupils speaking Vietnamese Speak Vietnamese Region School location Never Sometimes Always % % % Isolated 6 0 94 Rural 23 6 71 Red River Delta Urban 0 0 100 Total 15 3 82 Isolated 26 55 19 Rural 23 54 23 Northeast Urban 9 24 67 Total 23 52 25 Isolated 29 60 11 Rural 30 60 11 Northwest Urban 10 54 36 Total 28 59 13 Isolated 17 73 10 Rural 32 49 19 North Central Urban 25 0 75 Total 23 62 14 Isolated 38 53 9 Rural 8 15 77 Central Coast Urban 14 52 34 Total 34 52 14 Isolated 36 49 15 Rural 29 49 22 Central Highlands Urban 8 64 28 Total 30 50 20 Isolated 21 53 26 Rural 26 49 25 Southeast Urban 18 36 45 Total 22 45 33 Isolated 24 43 33 Rural 24 53 23 Mekong Delta Urban 16 30 54 Total 23 48 30 It can be seen that larger percentages of ethnic minority pupils in the Central Highlands and Central Coast said that they never spoke Vietnamese compared with the ethnic minority pupils in other regions. Even in the rural areas in the Central Highlands 29 percent of ethnic minority pupils never spoke Vietnamese. 71 Vietnam Reading and Mathematics Assessment Study Not speaking Vietnamese is associated with low pupil achievement. In Chapter 2, the percentages of pupils in the three categories of 'pre-functional', 'functional', and 'independent' were presented. For those speaking Vietnamese outside of school with different frequencies, the results were as follows: Table 3.4: Percentages and sampling errors for functionality level by speaking Vietnamese at home Speak Vietnamese Pre-functional Functional Independent at home % SE % SE % SE Never 33.78 2.03 37.83 1.71 28.38 2.41 Sometimes 21.53 1.23 37.79 1.38 40.68 1.94 All of the time 8.96 0.28 38.06 0.50 52.99 0.64 Vietnam 10.65 0.30 38.03 0.45 51.32 0.58 Measures are needed to help those pupils who never speak Vietnamese Pupils never or only outside of the school to have the opportunity to speak more Vietnamese. sometimes speaking The library books mentioned earlier will help with the written word but Vietnamese outside of at the same time, these pupils should have more exposure to the spoken school tended to be more word. TV and films will help but more thought is required about this disadvantaged. problem. As seen in the graph below, even within the group of non-Kinh pupils, their home background characteristics differed according to whether they spoke Vietnamese often or not. There was a very strong correlation between "speak Vietnamese" and other economic and cultural factors. Therefore, the relation between "speak Vietnamese" and pupils' achievement is not so simple as to conclude that having non-Kinh pupils speaking more Vietnamese will improve non-Kinh pupils' achievement. As mentioned above, careful interventions are required. As seen in Figure 3.2, the characteristics of non-Kinh pupils who always spoke Vietnamese was closer to those of Kinh pupils, but they were still somewhat disadvantaged when compared with Kinh pupils. Table 3.5: Means and sampling errors for selected pupils characteristics by ethnic affiliation Age Books at home Grade repetition Pupil ethnic Speak (months) (number) (times) Vietnamese Mean SE Mean SE Mean SE Kinh Always 132.8 0.1 14.3 0.3 0.17 0.00 Never 143.4 0.5 7.0 0.8 0.48 0.02 Non-Kinh Sometimes 143.4 0.4 7.6 0.7 0.42 0.02 Always 138.2 0.4 11.8 0.7 0.31 0.02 72 Vietnam Reading and Mathematics Assessment Study Figure 3.2: Home background characteristics of Kinh and non-Kinh pupils speaking different amounts of Vietnamese Policy suggestion 3.6: The MOET might wish to conduct experiments on ways of having children speak more Vietnamese. These experiments might include a) enrolling these children in kindgarten and having special Vietnamese lessons; b) having intensive three months Vietnamese course for such children before they begin school; c) having a more flexible curriculum such that an increased amount of reading and speaking Vietnamese becomes cross-curricular: and d) introducing the concept of a flexible curriculum and cross-curricular activities in all teacher training courses and school management courses. (i) Days absent in previous month In some countries absenteeism is a problem. It is also likely that those pupils Pupil absenteeism from who are absent will learn less. The pupils were asked how many days they had school tended not to be been absent in the month before they were tested. From Table 3.2 it can be seen a big problem. that for the whole country only 0.6 days were lost due to absenteeism. Only in the rural North Central was it 2 days and in isolated Central Highlands it was 73 Vietnam Reading and Mathematics Assessment Study 1.4 days. The main reason given was illness. In the Central Highlands it was the province of Kon Tum where the greatest absenteeism was found. (j) Private corner at home Many pupils had a In many countries there is great variation in homes in the extent to which private corner to study pupils have a quiet place to study. A question was therefore asked about the at home. Those not pupil having a private corner or space in which to work at home. In a sense having a private corner this is a question that reflects the caring of the parents about their children's tended to have less learning. But it also depends on the parents having enough money to have a educated parents. home where there is enough space for such a private corner. Having a quiet place to study is normally strongly associated with pupil achievement. From Table 3.2 it can be seen that 82 percent of pupils had a private corner at home but this was 87 percent for urban pupils, 82 for rural pupils and only 69 percent for pupils in isolated areas. The pupils from isolated areas in the Central coast and Central Highlands regions were those who tended to be worse off for private corners. Private corner at home Province % SE Kon Tum 56 4.3 Kien Giang 61 5.2 Tay Ninh 63 3.5 Soc Trang 65 3.2 Dong Thap 66 3.6 Ninh Thuan 67 2.6 Binh Thuan 67 2.8 Long An 68 2.5 The provinces where the fewest pupils said that they had a private corner at home have been presented in the above table. It can be seen that four of the provinces were from the Mekong Delta (Kien Giang, Soc Trang, Dong Thap and LongAn). The worst province of all was Kon Tum in the Central Highlands. Table 3.6: Percentages and sampling errors for private corner by parental education Parents education Private corner (years) % SE 0---10 71 0.7 12--18 85 0.4 19+ 92 0.3 Total 82 0.4 74 Vietnam Reading and Mathematics Assessment Study The more years of education the parents had, the more likely it was that they provided a private corner for their children but they might well have had more money with which to provide the corner. (k) Time working for family per day It is common that children should help their parents in their work. It was expected that those children who had to help their parents a great deal each day would have less time for doing their homework and for studying. The minutes that pupils had to work each day to help the family have been given in Table 3.2. It can be seen that the average for Vietnam was 110 minutes or one hour and 50 minutes. Those pupils in urban areas only helped for one hour and 34 minutes whereas those in isolated areas helped for one hour and 58 minutes and two hours and 17minutes for those in Northeast. Nine out of the ten provinces where pupils worked the most were in the Northeast region (in general a poor farming area). Again there was a difference between Kinh and non-Kinh pupils in the number of minutes worked. Table 3.7: Means and sampling errors for time working for family by ethnic affiliation Time working for family (mins.) Pupil ethnic Mean SE Non-Kinh pupils spent Kinh 108 0.4 more time working for Non-Kinh 123 1.1 their families. Total 110 0.4 Neither the level of parental education nor the total possessions (a proxy for wealth) were related to the time the children spent working for the family. At the same time, this variable was not associated with achievement in school as measured by the tests used. (l) Time spent in travel to school each day Rather than ask pupils how they travelled to school each day it was felt that Most pupils took about what was more important was the amount of time it took them to travel to 15 minutes to go to school irrespective of the form of travel. Although there are many anecdotes in school, but in rural ministries of education about children having to travel a long way to school and areas this could be hence being too tired to learn properly, there is no evidence from many other somewhat longer. studies that travel to school has an important and negative influence on achievement. But the effect of travel on achievement was not known for Vietnam and hence a question was asked about the time taken to get to school each day. From Table 3.2 it can be seen that on average the time taken to go to school was just over a quarter of an hour but in isolated areas it was twenty minutes. In some areas it was 25 minutes. Ten percent of all pupils took longer than 30 minutes to go to school and the longest time was one hour and 40 minutes. The poorer the family the longer it took to go to school and longer 75 Vietnam Reading and Mathematics Assessment Study travel time was associated with lower achievement. However, when home background was partialled out, the partial correlation was small. Government is providing more infrastructure in order to facilitate travel and creating more schools. These measures should alleviate the situation. (m) Grade repetition Pupils from poor Pupils were asked how many times they had repeated a grade. The families and ethnic percentages never having repeated have been presented in Table 3.2. A minorities tended to lot of pupils in the Northeast, Northwest and Central Highlands in repeat a grade more. isolated and rural areas had relatively high repetition rates. Cao Bang and Lang Son, both in the northeast region, had the highest repetition rates of all provinces. A series of small analyses were run in order to identify with which other variables grade repetition could be related. More ethnic minority pupils repeated a grade than did Kinh pupils. The extent to which the pupils spoke Vietnamese outside the school was not related to grade repetition. Nor was the length of teacher education or the type of teacher training the teachers had received. The number of total possessions was a variable that was related to grade repetition - the fewer the number of possessions in the home from which the child came the more likely he or she was to repeat a grade. To what extent did the pupils receive homework and have help at home? Homework is a very important aspect of schooling. It is from home work that pupils can consolidate what they have learned through further practice but they can also learn to think more about the concepts that lie behind what they are doing. Walberg and Paik2 have shown that doing homework versus not doing homework is highly related to pupil achievement. (a) Getting homework Most pupils received A question was asked in the pupil questionnaire about the extent to which the homework regulary pupils received reading homework and also mathematics homework. The options were: 'do not get homework', 'occasionally', and 'regularly'.At the same time it should be recognised that 1.9 percent of pupils for reading and 0.6 percent for mathematics indicated that they never received homework from the teacher. The percentages of pupils recording that they received homework three or four times per week have been presented in Table 3.8. It can be seen that this was 76 percent for reading and 92 percent for mathematics. 2Walberg H W and Paik S. 2000 Effective Educational Practices. Educational Practices Series No. 3. Geneva: International Bureau of Education (23pp) 76 Vietnam Reading and Mathematics Assessment Study Table 3.8: Percentages and sampling errors for homework variables Get Read Get Math Someone Help with Someone pays homework homework Correct Read homework Correct Math homework reminds of homework attention to homework homework Region School location 3 or 4 times 3 or 4 times Often All Often All Never and do not per week per week Never have homework Always % SE % SE % SE % SE % SE % SE % SE % SE % SE Isolated 85 14.0 94 3.7 78 21.9 6 5.6 78 23.3 15 17.5 7 6.2 28 21.1 35 12.7 Red River Rural 85 1.0 95 0.5 63 1.3 20 1.2 67 1.4 23 1.3 6 0.5 29 1.2 53 1.4 Delta Urban 83 2.1 97 0.5 49 2.3 31 2.3 53 2.2 35 2.4 2 0.4 31 2.9 48 2.3 Total 84 0.9 95 0.4 60 1.1 22 1.0 65 1.1 25 1.2 6 0.4 30 1.2 52 1.2 Isolated 76 2.8 93 1.0 62 3.2 22 2.9 66 3.3 26 3.3 12 1.4 37 2.1 43 2.3 Rural 81 1.2 94 0.7 61 1.6 22 1.3 64 1.5 26 1.4 6 0.3 30 1.1 53 1.5 Northeast Urban 85 2.2 96 0.7 57 2.9 24 2.8 62 2.8 26 2.8 2 0.4 30 2.8 53 2.6 Total 81 1.0 94 0.6 60 1.3 22 1.1 64 1.2 26 1.1 6 0.4 31 0.9 51 1.1 Isolated 70 3.9 90 1.7 56 4.6 23 3.3 60 3.7 26 3.1 11 2.3 36 3.7 41 3.5 Rural 82 2.9 96 1.0 57 3.8 20 2.7 61 3.8 26 3.3 6 1.1 31 2.7 50 3.7 Northwest Urban 86 3.4 98 1.2 61 5.3 20 4.3 66 4.2 27 4.3 3 0.8 26 3.8 52 5.0 Total 78 2.0 94 0.9 57 2.3 21 1.7 62 2.4 26 2.1 7 1.0 32 2.0 47 2.5 Isolated 73 5.3 92 2.4 53 5.5 17 2.8 55 5.3 23 3.5 8 1.3 31 3.8 45 4.7 North Rural 82 1.4 93 0.9 61 1.8 20 1.5 65 1.7 24 1.6 5 0.6 28 1.5 57 2.2 Central Urban 84 3.0 94 2.2 50 4.3 20 2.4 55 4.0 24 2.8 2 0.4 26 2.0 50 4.2 Total 82 1.2 93 0.8 58 1.6 20 1.2 63 1.6 24 1.3 5 0.5 28 1.2 55 1.8 Isolated 65 4.6 83 2.5 44 4.1 27 3.7 47 4.7 33 4.2 13 2.7 41 4.0 41 4.3 Central Rural 68 1.9 89 1.0 49 1.9 24 1.6 56 1.9 29 1.7 7 0.9 37 1.5 48 1.8 Coast Urban 70 2.7 93 1.1 43 2.6 33 3.4 47 3.3 39 3.8 4 0.6 36 2.5 50 2.6 Total 68 1.5 89 0.8 47 1.5 27 1.4 53 1.6 32 1.6 7 0.6 37 1.2 48 1.4 Isolated 71 4.7 91 2.3 49 5.3 28 5.6 51 5.2 34 5.6 16 3.1 43 4.5 39 6.0 Central Rural 75 2.6 93 1.4 58 3.1 15 1.7 64 2.5 20 2.1 8 1.1 32 2.4 47 3.1 Highlands Urban 77 4.1 94 1.4 51 3.9 23 4.0 59 3.7 27 4.1 3 0.6 33 2.8 49 4.0 Total 75 1.9 93 1.0 54 2.2 20 2.1 60 2.1 25 2.2 8 0.9 35 1.6 46 2.4 Isolated 74 2.9 92 1.5 45 3.4 32 3.8 53 3.5 35 4.3 9 1.1 40 3.9 48 3.5 Rural 70 2.0 91 1.3 49 1.9 29 1.8 53 1.9 34 1.9 8 0.5 41 1.6 45 1.7 Southeast Urban 73 1.7 92 0.9 40 2.1 39 2.4 45 2.4 45 2.3 5 0.5 41 2.2 49 2.1 Total 72 1.3 91 0.7 45 1.3 33 1.2 50 1.4 39 1.3 7 0.3 41 1.1 47 1.2 Isolated 64 3.2 86 1.9 51 3.0 31 2.8 52 2.9 37 2.8 8 1.1 47 3.1 55 2.6 Mekong Rural 65 1.3 86 0.7 45 1.3 36 1.3 46 1.4 42 1.5 8 0.5 48 1.3 53 0.9 Delta Urban 63 2.7 88 1.5 41 2.0 36 2.8 44 2.1 43 2.5 6 0.9 43 2.6 52 2.6 Total 64 1.0 87 0.6 45 1.1 35 1.2 46 1.1 41 1.3 7 0.4 47 1.2 53 0.8 Isolated 71 1.5 90 0.9 53 1.4 26 1.2 56 1.3 31 1.2 10 0.6 40 1.1 46 1.4 Rural 76 0.5 92 0.3 56 0.6 25 0.6 59 0.7 29 0.7 7 0.2 35 0.5 52 0.5 Vietnam Urban 76 1.0 93 0.5 46 1.0 31 1.0 51 1.0 37 1.1 4 0.2 36 1.2 50 1.0 Total 76 0.4 92 0.3 53 0.5 26 0.5 57 0.6 31 0.6 6 0.1 36 0.5 51 0.5 Pupils in isolated areas tended to receive less homework except in the Southeast and Mekong Delta regions. As can be seen in the table below pupils in many of the Mekong Delta provinces (shaded in grey) had the least amount of regular homework given by teachers. Various checks were made to see if other variables were associated with a low percentage of pupils receiving homework 3-4 times per week. Possession of textbooks was not related and nor was the number of years of teacher education or the level of teacher training that the teachers had received. Somewhat surprisingly, the ethnic minority children did not receive less regular homework. 77 Vietnam report: Volum 2, Chapter 3 Pupil characteristics 78 They tended to receive Get Read homework Get Math homework more homework in Province 3 or 4 times per week Province 3 or 4 times per week mathematics than in % SE % SE reading. An Giang 48 3.9 Soc Trang 80 3.0 Can Tho 53 2.9 An Giang 82 2.2 Quang Ngai 56 3.8 Can Tho 83 2.0 Binh Thuan 57 3.5 Bac Lieu 83 2.5 Tay Ninh 59 4.2 Ca Mau 84 1.8 Soc Trang 60 4.0 Vinh Long 86 2.1 Long An 64 3.7 Quang Nam 87 1.9 Ninh Thuan 64 3.4 Binh Thuan 88 2.0 Bac Lieu 65 3.7 Tay Ninh 88 2.2 Dong Thap 66 4.1 Dong Thap 88 2.7 (b) Frequency of teacher correcting the homework The corollary of pupils receiving homework is that the teacher actually corrects it. Walberg has shown that the more that a teacher corrects the homework the better, on average, will be the achievement of the pupils. It is better to correct homework than not to correct it but that giving homework and not correcting it was better than not giving any homework. A question was asked of pupils: Does your teacher correct your homework in Vietnamese/mathematics? 1 = I am not given any homework 2 = My teacher never corrects my homework 3 = My teacher sometimes corrects my homework 4 = My teacher often corrects my homework 5 = My teacher corrects all of my homework It must be recalled that this question was asked of the pupil so that it is the pupils' perception of how the teacher corrected the homework. Although there was not much variation among the school locations and regions there was slightly more variation for reading than for math. What was somewhat shocking was that not all teachers corrected homework all of the time. In Table 3.9 the mean scores in math and reading have been presented for the various categories of the 'homework corrected' variable. Table 3.9: Means and sampling errors of achievement in reading and mathematics by homework corrected category Pupils receiving Homework corrected Reading 500 score Math 500 score homework and having it Mean SE Mean SE corrected perform better Not given homework 459.6 5.2 449.7 7.0 than those not receiving Never corrected 450.0 4.0 442.7 5.2 and not having Sometimes corrected 485.5 2.1 473.0 2.8 Often corrected 504.9 1.6 504.9 1.7 corrected. Always corrected 505.8 1.9 503.0 1.9 78 Vietnam Reading and Mathematics Assessment Study 79 Vietnam report: Volum 2, Chapter 3 Pupil characteristics For both subjects there was no difference in score between never being given homework and being given it but it never being corrected. There was a dramatic increase in score if the pupils were given homework and it was sometimes corrected. There was a further increase in score if the pupils were given homework and it was corrected often. There was no difference in pupil score between homework being corrected often and always. This is an important finding and shows the great importance of homework being given and corrected often. But, it has already been shown that it is the pupils in the isolated areas who received less homework. Policy suggestion 3.7: The MOET and the provincial authorities should re-stress in their guidelines the importance of giving and correcting homework. In the monitoring (inspection) system special checks should be made in the schools to ensure that this is happening. (c) Someone pays attention to school work It is commonplace in Vietnam that someone at home will look at the Most pupils had schoolwork to follow how the children are progressing. This behaviour is someone at home who known as 'paying attention'. The options in the question in the questionnaire "paid attention" to their were: 'never', 'sometimes', and 'most of the time'. From Table 3.8, it can be school work. seen that only seven percent said that no one paid attention. However, in isolated areas in the Northeast, Northwest and Central Highlands this figure was 10 percent or more. Pupils responding that they had less attention paid at home tended to have parents with fewer years of education and who met the teachers less frequently and also to come from poorer homes as measured by total possessions. (d) Help with homework outside school A further question was asked about the extent to which the pupils received 36% of pupils said that help with their homework from a person outside of school. The options were: they did not receive 'I do not get homework', 'never', 'sometimes', and 'most of the time'. From homework and did not Table 3.8, it can be seen that 36 percent of pupils said that either they did not receive help at home. receive homework or never received help at home. This percentage was And they tended to come higher in isolated areas than in rural and urban areas. The percentages are from poor homes and particularly high in the isolated areas in the Central Coast, Central Highlands, parents had little Southeast and Mekong Delta regions. education. The percentage of pupils responding 'never had homework' or 'never received help with homework' was related to the level of parental education and to the wealth of the home. Pupils whose parents had lower levels of education and who had fewer possessions at home tended to receive less help at home. But, those pupils whose parents have had very little education cannot be expected to give direct help to their children. 79 Vietnam Reading and Mathematics Assessment Study From experiences in other countries it might be assumed that parents showing interest in their children's homework might help increase achievement even though the parents cannot be of direct help to their children. This was suggested earlier in the chapter when it was suggested that there be a series of school intervention programs. (e) Someone checks that the homework has been done Only hafl of the pupils The pupils were asked to what extent there was someone that made sure that had someone at home they had done their homework. The options were the same as for the previous who check their question. It can be seen that about half of the pupils in Vietnam's Grade 5 did homework have been have someone at home who ensured that their homework had been done. done. Pupils in urban areas had someone who reminded them more than pupils in rural and isolated areas. The Southeast and Central Highlands had parents who reminded them less often than pupils in other regions. Again this variable was related to parents with fewer years of education, poorer families and meeting teachers less often. It is clear that pupils from poorer homes and where parents had less education were those who did not have the family support required for doing homework. To what extent did pupils receive extra tuition? Extra tuition is In most countries, many children take extra tuition. And in Asia these discouraged by the numbers are quite high (Bray, 19993). It was therefore of interest to discover government. the extent of this phenomenon in Vietnam. However, it is important to proceed carefully in Vietnam since the question of extra tuition is politically sensitive. This is because the government's position is to discourage the practice of extra tuition. At the same time, it is recognised that teacher salaries are low and hence many teachers wish to supplement their income and one way to do this is to give extra tuition. But, nearly 30% of It is also important to be careful about the pupils' responses since they may pupils said that they have misunderstood the question. Those pupils in full-day schooling have one took extra tuition.. session in the morning which is free, one on the afternoon for which the parents Nearly all came from must pay for and sometimes they have extra tuition in the evening. But these half-day schooling. pupils may have thought that they were having extra tuition in the afternoon session because their parents pay for it. The percentages of pupils not having extra tuition have been compared with the percentages of pupils in full day schooling. Nearly all pupils in full-day schools answered that they did not have any extra tuition. It can be assumed that most pupils understood the question well and that there was not a misunderstanding as had been feared. 3Bray, Mark (1999) The Shadow Education System: private tutoring and it implications for planners. Fundamentals of Educational Planning Series No. 61. IIEP: Paris 80 Vietnam Reading and Mathematics Assessment Study Table 3.10: Percentage and sampling errors for extra tuition by school shift No extra tuition Pupil shift % SE Full-day school 80 1.3 First shift only 55 0.9 Second shift only 67 1.6 Vietnam 62 0.6 Table 3.11a: Percentages, means and sampling errors of extra tuition variables (all pupils) Time spent on extra tuition (all pupils) Region School location None Viet Math Viet + Math % SE Mean SE Mean SE Mean SE Isolated 86 9.4 0.21 0.20 0.42 0.27 0.63 0.46 Red River Rural 76 1.7 0.68 0.05 0.71 0.06 1.39 0.11 Delta Urban 63 3.6 0.9 0.12 1.13 0.13 2.03 0.24 Total 73 1.6 0.72 0.05 0.78 0.05 1.49 0.10 Isolated 79 3.2 0.61 0.11 0.56 0.10 1.17 0.21 Northeast Rural 68 2.0 0.89 0.06 0.93 0.06 1.82 0.12 Urban 65 3.3 1.11 0.13 1.19 0.13 2.3 0.26 Total 69 1.6 0.88 0.05 0.91 0.05 1.79 0.10 Isolated 77 5.4 0.56 0.15 0.54 0.15 1.11 0.30 Northwest Rural 81 3.5 0.51 0.12 0.54 0.13 1.05 0.24 Urban 58 6.5 0.93 0.17 0.95 0.16 1.89 0.33 Total 76 2.7 0.59 0.08 0.6 0.08 1.2 0.16 Isolated 63 6.9 1.12 0.23 1.1 0.23 2.21 0.46 North Rural 57 3.2 1.21 0.10 1.31 0.11 2.52 0.21 Central Urban 55 5.3 1.43 0.21 1.52 0.20 2.95 0.41 Total 58 2.6 1.23 0.08 1.32 0.08 2.55 0.17 Isolated 45 6.5 1.66 0.23 1.75 0.24 3.41 0.47 Central Rural 51 3.3 1.5 0.12 1.59 0.13 3.09 0.24 Coast Urban 43 3.9 1.69 0.19 1.83 0.17 3.52 0.35 Total 49 2.3 1.57 0.09 1.67 0.09 3.24 0.18 Isolated 64 6.9 1.38 0.29 1.23 0.28 2.61 0.55 Central Rural 49 5.3 1.76 0.23 1.75 0.22 3.51 0.44 Highlands Urban 44 6.2 2.2 0.33 2.27 0.31 4.47 0.64 Total 51 3.3 1.8 0.15 1.78 0.15 3.58 0.29 Isolated 51 5.4 1.5 0.18 1.6 0.19 3.1 0.38 Southeast Rural 46 2.4 1.72 0.11 1.92 0.12 3.64 0.21 Urban 42 3.3 1.99 0.14 2.33 0.16 4.32 0.30 Total 45 2.0 1.8 0.09 2.05 0.10 3.86 0.18 Isolated 81 2.8 0.48 0.08 0.55 0.09 1.02 0.17 Mekong Rural 67 1.3 0.92 0.06 1.06 0.06 1.98 0.12 Delta Urban 53 3.5 1.39 0.14 1.55 0.14 2.93 0.28 Total 67 1.2 0.93 0.05 1.07 0.05 1.99 0.09 Isolated 69 1.9 0.92 0.06 0.93 0.07 1.85 0.13 Vietnam Rural 63 1.0 1.06 0.04 1.15 0.04 2.21 0.07 Urban 51 1.6 1.53 0.06 1.73 0.07 3.26 0.13 Total 62 0.6 1.15 0.03 1.25 0.03 2.39 0.05 81 Vietnam Reading and Mathematics Assessment Study Table 3.11b: Percentages, means and sampling errors of extra tuition variables (only pupils who had extra tuition) Pay for extra Who teaches tuition extra tuition Region School Same teacher location Viet Math Viet + Math % Yes as at school Mean SE Mean SE Mean SE % SE % SE Isolated 1.5 1.06 2.8 0.51 4.2 2.28 86 15.3 84 16.3 Red River Rural 2.8 0.10 2.9 0.11 5.7 0.20 92 1.9 97 0.5 Delta Urban 2.5 0.18 3.1 0.15 5.5 0.32 91 1.7 84 3.2 Total 2.7 0.10 2.9 0.09 5.6 0.18 92 1.5 93 0.8 Isolated 2.9 0.32 2.7 0.23 5.6 0.53 48 7.7 97 1 Rural 2.8 0.08 2.9 0.10 5.6 0.17 81 2.4 99 0.3 Northeast Urban 3.1 0.21 3.4 0.20 6.5 0.40 88 4.6 94 1.5 Total 2.8 0.07 3.0 0.09 5.8 0.14 79 2.1 98 0.4 Isolated 2.5 0.56 2.4 0.55 4.9 1.08 35 14.1 99 0.5 Rural 2.6 0.33 2.8 0.33 5.4 0.61 63 9.9 100 0.4 Northwest Urban 2.2 0.22 2.3 0.22 4.5 0.43 71 10.5 100 0.3 Total 2.5 0.22 2.5 0.22 5.0 0.42 57 7.3 100 0.3 Isolated 3.0 0.39 2.9 0.38 5.9 0.76 71 7.7 96 1.9 North Rural 2.9 0.13 3.1 0.13 5.9 0.25 89 2.4 98 0.7 Central Urban 3.2 0.29 3.4 0.29 6.5 0.56 96 1.3 92 1.9 Total 2.9 0.12 3.1 0.12 6.0 0.23 88 2.1 97 0.6 Isolated 3.0 0.25 3.2 0.23 6.2 0.46 81 9.9 99 1 Central Rural 3.1 0.11 3.3 0.12 6.4 0.23 91 2 97 0.7 Coast Urban 3.0 0.21 3.2 0.18 6.2 0.38 91 2.4 94 1.1 Total 3.1 0.09 3.2 0.09 6.3 0.18 90 1.7 96 0.5 Isolated 3.8 0.40 3.4 0.30 7.2 0.59 80 5.2 97 1.6 Central Rural 3.5 0.24 3.4 0.21 6.9 0.43 91 2.7 97 0.8 Highlands Urban 3.9 0.30 4.0 0.27 7.9 0.54 91 2.3 97 1.2 Total 3.7 0.18 3.6 0.14 7.3 0.30 89 1.8 97 0.6 Isolated 3.1 0.20 3.3 0.18 6.4 0.37 87 4.1 99 0.5 Rural 3.2 0.13 3.6 0.13 6.7 0.23 91 1.7 98 0.4 Southeast Urban 3.4 0.13 4.0 0.14 7.5 0.27 94 1.1 91 1.3 Total 3.3 0.09 3.7 0.10 7.0 0.17 92 1 95 0.6 Isolated 2.5 0.21 2.9 0.24 5.3 0.34 61 5.8 97 1.1 Mekong Rural 2.8 0.12 3.2 0.12 6.0 0.22 77 2.1 97 0.4 Delta Urban 3.0 0.18 3.3 0.17 6.3 0.34 89 2.3 97 0.6 Total 2.8 0.10 3.2 0.10 6.0 0.18 79 1.7 97 0.3 Isolated 3.0 0.11 3.0 0.12 6.0 0.22 71 3.1 98 0.5 Rural 2.9 0.05 3.2 0.05 6.1 0.09 87 0.8 97 0.2 Vietnam Urban 3.2 0.07 3.6 0.07 6.7 0.13 92 0.9 92 0.5 Total 3.0 0.04 3.3 0.04 6.2 0.07 86 0.7 96 0.2 82 Vietnam Reading and Mathematics Assessment Study (a) Number of hours of extra tuition per school week Since this study is focussed on reading in Vietnamese and also mathematics, Most of those taking extra it was thought important to find out about extra tuition in these two subject tuition took one hour in areas. The results have been presented in Table 3.11a and Table 3.11b. All mathematics or one hour pupils, both pupils who had extra tuition and who did not have extra tuition in Vietnamese. have been taken into account in Table 3.11a. In Table 3.11b, the calculations have been based on only those pupils who said that they had extra tuition. It can be seen that 62 percent of pupils stated that they had had no extra tuition. However, this means that 38 percent of pupils did have some extra tuition. More pupils in urban areas had extra tuition than in the other areas. The average number of hours per week spent on extra tuition in each subject area for all pupils (including those not taking extra tuition, was about 1.2 hours. The average number of extra tuition hours for only those who had extra tuitions was about 3 hours for Vietnamese and 3.3 hours for mathematics. Pupils in isolated schools tended to have fewer hours of extra tuition per week compared with pupils in rural and urban schools. The number of hours of extra tuition was not related to achievement scores. Presumably this is because it was the parents of the weaker pupils who sent them to extra tuition in order to catch up. (b) Do parents pay for extra tuition? Those pupils who attended extra tuition were asked if they paid for it. In other Most extra tuition was words had their parents given them money to give to the teacher? The paid for. percentage of pupils responding 'yes' was 86. It must be remembered that this is 86 percent of those saying that they took extra tuition. Pupils in isolated areas were less likely to have to pay. A higher percentage of pupils from richer homes went to extra tuition than from poorer homes. It was noted that in Hai Duong province the pupils had fewer hours of extra tuition but it was there that 99 percent paid for extra tuition. (c) Who teaches the pupils for extra tuition? The only options offered to the pupils were: 'I have the same teacher as at And, most pupils had school' and 'I have a teacher from different school'. The responses have been their own teachers as presented in Table 3.11b. It can be seen that in general the pupils had the their tutors. same teachers as at school for the extra tuition. Given that the issue of extra tuition is a sensitive one, it is suggested that the MOET select one or two provinces and carry out a much more detailed examination of who is doing extra tuition, why, for how long, how, and if it has an effect. Who pays how much for the extra tuition and whether extra tuition can be seen as a positive and needed supplement to ordinary school. Policy suggestion 3.8: The MOET might wish to consider mounting a separate detailed study of extra tuition in selected provinces in order to obtain a much more detailed account of what is happening, why, how, and if it can used to improve pupil attainment. 83 Vietnam Reading and Mathematics Assessment Study What were the relationships of pupil background indicators to achievement in mathematics and reading? It is instructive to examine the pristine relationships of the variables reviewed in this chapter with pupil achievement before moving to the teacher variables. Since many variables are inter-related the disentangling of the different effects will be taken up later in Chapter 8. However the initial simple correlations are of interest and have been presented in Table 3.12. Table 3.12: Simple correlations of selected pupil variables with achievement Nearly all of the pupil Pupil characteristics Math achievement Reading achievement variables examined in Age -0.17 -0.19 this chapter were Sex -0.01 0.06 correlated significantly Book at home 0.14 0.14 with mathematics and Possession in home 0.28 0.29 reading achievement. Meals 0.15 0.15 The exceptions were Travel time to school -0.11 -0.12 'time working for family' Speak Vietnamese 0.12 0.15 and 'help with homework.' Day absent -0.11 -0.11 Private corner at home 0.15 0.14 Time working for family -0.02 -0.03 Help with homework 0.04 0.03 Grade repetition -0.15 -0.17 Homework general 0.10 0.10 Math homework 0.15 0.15 Reading homework 0.17 0.16 Parent education 0.28 0.28 Major ethnic group 0.10 0.14 With such a large sample the standard error of sampling will be very small. In general any correlation greater than .026 will be significant at the 95 percent level of confidence. But, this would explain very little variance. For the time being any correlation greater than 0.12 will be taken as worthy of comment. The three homework variables in Table 3.12 are derived variables. General home work consists of 'parents ensuring that homework is done', 'someone at home paying attention to school work', and 'someone at home helping with homework'. The loadings on the construct were .803, .633, and .738 respectively. Math homework is a composite of 'getting math homework from the teacher' and the 'teacher correcting the math homework'. The loadings were .804 for each variable. The reading homework is the same as 84 Vietnam Reading and Mathematics Assessment Study for math homework but in this case Vietnamese homework. In this case the loadings for each variable on the construct were .842 From Table 3.12. it can be seen that, applying the cut-off point of .12, there was no relationship between the sex of the pupil and achievement, no relationship for days absent, and no relationship for travel time to school. The largest correlations were for possessions in the home (a proxy measure for wealth of home), level of parental education, age, a private corner at home and meals per day. In short, those pupils from wealthier homes with better educated parents, a private corner at home, more meals per day, and who were younger had better achievement in both reading and math. From the tables presented earlier in this chapter it has been seen that the results are much the same but that the kinds of 'poor' home backgrounds referred to are more likely to be found in isolated areas in certain regions of Vietnam. Finally, because a home background construct will be needed in order to examine the effect of teacher and school variables after the home effect has been 'taken out', a factor was developed consisting of the variables given below in Table 3.13. A principal components analysis was undertaken of the home variables that were not connected with school. The final list of variables used in the component retained together with the factor loading of each variable has been presented in Table 3.13. Table 3.13: Factor loadings of variables in 'Home background construct' Speak Vietnamese outside school 0.797 It was possible to create a home background Belonging to ethnic minority -0.794 factor which was Number of books in the home 0.467 relatively highly correlated with pupils' Years parental education 0.591 achievement. Wealth of home (total possessions) 0.662 The 'belonging to an ethnic minority' has a negative loading on the factor because the variable was coded 1=belonging to ethnic minority group and 2= belonging to main Kinh group. It will be seen that the main variables were speaking Vietnamese outside the home and belonging to the mainstream ethnic group. The wealth of the home was the next most important variable and parental education and the number of books in the home had the least weight in the home background factor. This home background factor/construct correlated 0.27 with mathematics achievement and 0.30 with reading achievement. Conclusion This chapter has dealt with the characteristics of pupils in Grade 5. For the most part the results showed that there was considerable variation in the 85 Vietnam Reading and Mathematics Assessment Study characteristics of pupils among the isolated, rural and urban areas of the country. Those regions with more provinces having isolated areas tended to be those where the pupils were somewhat disadvantaged compared with their peers in other parts of the country. The isolated areas had Grade 5 pupils who were older, sometimes a year older than the average for Vietnam, had fewer girls in school, had more ethnic minority children, and had pupils from poorer families where the parents had fewer years of education than parents in other areas. Such children were also from homes with fewer books. Indeed the average number of books at home for all of Vietnam was very low. These children also spoke less Vietnamese outside of school, tended to repeat a grade more than those from other areas. The parents of these children tended to pay less attention to their school work and homework. To compound matters, these disadvantaged pupils did not receive extra help from school to help overcome their disdvantagedness. Rather they were also the pupils who received less homework from school and had their homework corrected less frequently. A number of policy suggestions were made throughout the chapter. In many cases they were specific about what to do in which provinces. These mostly concerned the authorities instituting so-called school intervention programs in order that the school can influence the behaviour of parents in taking an interest in their children's education and homework as well as having the pupils come to school on time (not a year late) and not dropping out. It was suggested that the Vietnam authorities should study what other South-East Asian countries have done in way of school intervention programs and then develop and assess, before full implementation, the best kinds of intervention programs for Vietnam. Suggestions were made for school meal programs in selected areas as the system moves to full-day schooling. Book flood programs should be considered for schools in isolated areas. Other suggestions concerned strengthening the current school and classroom library program in order to for children to have sufficient leisure time reading materials, ensuring that teachers give and correct homework. Extra tuition was a form of extra schooling that was taken by more than thirty percent of pupils. Although teachers are discouraged from extra teaching by the State it nevertheless happens to a great extent. It was suggested that the MOET undertake a detailed study of this phenomenon in order to improve it and make it beneficial for as many pupils as possible. Finally, it was seen that teachers giving homework and then correcting it often or always was beneficial for pupil achievement. It was worrisome that some teachers were not perceived by pupils to be giving homework or correcting it. Steps should be taken by the MOET to emphasise homework even more in their teacher pre- and in-service training courses. Those responsible for monitoring what happens in schools should have the task of ensuring that homework is given and corrected on a very regular basis. 86 Vietnam Reading and Mathematics Assessment Study Chapter 4 WHATWERETHECHARACTERISTICSOFGRADE 5 TEACHERS AND THEIR CLASSROOMS? Introduction T eachers are viewed as a key part of the educational process. They also account for nearly 90 percent of the educational budget in many countries. They are selected for teacher training, undergo the teacher training, are allocated to schools where they gain experience and undergo in-service training. What then were the characteristics of Grade 5 teachers in Vietnam and how were the teachers distributed across schools? The same reasons apply for measuring the characteristics of teachers and their classrooms as for measuring the pupils' backgrounds and characteristics as were reported in Chapter 3. It will be recalled from Chapter 1 that only two teachers per school were drawn at random. However, only just over half the pupils were linked to teachers on the data file. But, when the differences were examined between pupils who were linked to teachers and those who were not linked to teachers there were no differences between the mean values and percentages for the two groups. In this chapter it must be recalled that where teacher characteristics have been related to pupil achievement it was only 51 percent of pupils that were used in the calculations. Finally, all teacher data have been disaggregated to the pupil level. This means that in tables in this chapter when a percentage is used it means that x percent of pupils had teachers with the particular characteristic. A "mean" value means that the average pupil had a teacher with that characteristic. What were the background characteristics of teachers of Grade 5 pupils? The first teacher background data to be examined have been presented in Table 4.1. 87 Vietnam Reading and Mathematics Assessment Study (a) Age of teachers The average pupil As can be seen from the figure at the foot of Table 4.1, the average pupil in Vietnam had a teacher who had a teacher aged 33.7 years. The teachers in isolated areas were, on average, was 34 years old, but younger than those from rural areas who in turn were younger than those from in isolated areas urban areas. teachers were younger. (b) Sex of teachers 73% of pupil had Twenty-seven percent of pupils had male teachers and 73 percent had female female teachers. teachers. There were more female teachers in urban areas than elsewhere. In isolated areas in the Northwest, Southeast and Mekong Delta there were more male teachers. Table 4.1: Means, percentages and sampling errors of selected teacher background characteristics Ethnic Group Coming from Academic Professional Teacher Perception of education education experience Excellent effectiveness Region School Age (years) Sex (female) (Kinh) local area teacher location (years) (years) (years) of INSET Mean SE % SE % SE % SE Mean SE Mean SE Mean SE % SE Mean SE Isolated 31.4 1.40 93 15.7 100 0.0 43 16.9 12.0 0.00 2.0 0.11 9.5 2.98 100 0.0 3.37 0.30 Red River Rural 32.7 0.28 91 1.1 100 0.3 54 2.0 12.3 0.05 2.3 0.02 11.9 0.30 83 1.6 3.11 0.03 Delta Urban 37.2 0.67 97 1.2 99 1.1 55 4.5 12.4 0.20 2.5 0.06 15.3 0.67 85 2.9 3.09 0.08 Total 33.4 0.25 92 0.9 100 0.3 54 1.7 12.3 0.06 2.3 0.02 12.4 0.26 84 1.3 3.11 0.03 Isolated 30.8 0.45 76 3.1 49 3.9 38 3.8 11.7 0.08 1.9 0.03 9.5 0.44 51 3.9 3.16 0.04 Rural 34.4 0.29 88 1.3 78 1.6 47 1.8 11.7 0.05 2.1 0.02 13.1 0.28 73 2.0 3.12 0.02 Northeast Urban 39.1 0.53 98 1.3 88 2.5 55 3.9 11.8 0.17 2.3 0.05 16.8 0.55 81 2.6 3.08 0.03 Total 34.4 0.26 87 1.1 74 1.3 46 1.5 11.7 0.04 2.1 0.02 13.0 0.24 70 1.8 3.12 0.02 Isolated 28.5 0.84 68 5.4 54 7.3 28 6.2 10.7 0.20 1.7 0.04 7.8 0.80 41 6.2 3.25 0.08 Rural 33.2 0.64 82 3.2 71 4.5 36 5.0 11.3 0.11 2.0 0.04 12.6 0.71 81 3.4 3.12 0.04 Northwest Urban 36.7 0.71 100 0.4 89 5.1 61 7.0 11.6 0.27 2.1 0.13 15.5 0.83 83 6.5 3.31 0.09 Total 32.0 0.54 79 2.3 67 3.3 37 3.6 11.1 0.10 1.9 0.03 11.4 0.52 66 3.5 3.20 0.04 Isolated 31.2 0.82 56 6.0 80 6.3 20 4.9 11.9 0.10 2.1 0.06 9.2 1.09 60 5.6 3.21 0.11 North Rural 32.8 0.49 80 2.1 99 0.6 42 2.5 12.0 0.06 2.1 0.02 11.5 0.44 87 1.8 3.20 0.03 Central Urban 39.3 0.83 91 2.3 100 0.0 55 7.7 12.8 0.23 2.5 0.11 16.3 0.76 92 2.3 3.09 0.07 Total 33.4 0.45 78 1.9 97 1.0 41 2.6 12.1 0.06 2.2 0.02 11.9 0.40 85 1.7 3.19 0.03 Isolated 31.1 0.76 52 8.0 100 0.0 37 6.3 11.9 0.21 2.1 0.07 10.7 1.02 57 9.1 3.02 0.11 Central Rural 35.1 0.38 62 2.6 100 0.3 63 3.0 12.2 0.05 2.2 0.03 13.8 0.39 74 3.4 3.15 0.05 Coast Urban 39.9 0.71 87 2.4 100 0.0 51 4.0 12.6 0.13 2.4 0.05 17.5 0.77 77 4.3 3.05 0.07 Total 35.5 0.38 66 2.0 100 0.2 57 2.3 12.2 0.06 2.2 0.02 14.1 0.36 72 2.6 3.11 0.04 Isolated 29.6 0.84 63 7.9 94 2.8 32 8.8 11.5 0.18 1.9 0.06 6.6 0.85 63 8.2 3.07 0.09 Central Rural 30.9 0.48 74 4.8 94 2.5 44 5.5 11.9 0.10 1.8 0.06 9.1 0.60 77 4.6 3.20 0.07 Highlands Urban 37.0 1.15 100 0.4 98 1.7 56 7.6 11.9 0.13 1.9 0.10 15.0 1.20 88 5.6 2.93 0.10 Total 31.9 0.46 76 3.2 95 1.5 43 4.2 11.8 0.08 1.8 0.04 9.8 0.49 76 3.0 3.11 0.05 Isolated 30.3 0.71 54 6.3 95 1.7 66 4.6 11.6 0.11 1.8 0.04 9.0 0.71 51 7.4 3.00 0.10 Rural 32.1 0.34 69 2.3 97 0.7 60 2.5 11.9 0.05 1.9 0.03 9.9 0.31 60 3.1 2.99 0.04 Southeast Urban 37.5 0.49 75 3.2 99 0.4 46 3.6 12.1 0.07 1.9 0.04 15.5 0.55 72 4.0 2.88 0.06 Total 33.7 0.29 69 1.9 98 0.5 56 2.0 11.9 0.04 1.9 0.02 11.7 0.29 63 2.3 2.95 0.03 Isolated 30.6 0.40 34 4.5 95 2.5 52 4.2 11.6 0.12 1.9 0.05 9.4 0.41 45 4.7 3.32 0.06 Mekong Rural 33.3 0.28 42 1.9 98 0.6 67 1.8 11.7 0.06 1.8 0.02 12.3 0.27 56 2.3 3.28 0.03 Delta Urban 36.6 0.72 73 3.6 96 1.5 61 4.2 11.9 0.11 2.0 0.06 14.5 0.80 65 4.7 3.26 0.05 Total 33.3 0.22 45 1.7 97 0.6 64 1.5 11.7 0.05 1.9 0.02 12.1 0.23 55 1.7 3.28 0.02 Isolated 30.5 0.22 57 2.6 80 1.6 42 2.1 11.6 0.05 1.9 0.02 9.1 0.22 53 2.5 3.17 0.03 Rural 33.2 0.14 73 0.7 95 0.4 54 0.9 11.9 0.03 2.1 0.01 12.0 0.13 73 0.9 3.16 0.01 Vietnam Urban 37.9 0.24 86 1.2 97 0.6 54 1.7 12.2 0.05 2.2 0.02 15.8 0.23 78 1.6 3.05 0.02 Total 33.7 0.12 73 0.5 93 0.3 52 0.7 11.9 0.02 2.1 0.01 12.2 0.12 71 0.6 3.14 0.01 88 Vietnam Reading and Mathematics Assessment Study One question of interest was if the female teachers had higher achievement scores in reading and mathematics than male teachers. Theoretically, this should not be so. The achievement differences by gender have been presented in the following table. The PMA500 and PRD500 are Pupil mathematics and reading scores and have been scaled to a mean of 500 and a standard deviation of 100. The teacher scores (Tmath500 and Tread500) were also scaled to a mean of 500 and a standard deviation of 100. The pupil and the teacher scores are not comparable. It should be mentioned until 1983 most primary school teachers had had ten years In general, female of education and thereafter the requirement was 12 years. For all of Vietnam the teachers had higher female teachers had slightly, but significantly, higher scores in reading than their achievement scores than male counterparts. This was particularly the case in rural areas. male teachers. Table 4.2: Differences in achievement for teachers and pupils Pupil Teacher School Teacher location sex PRD 500 PMA 500 Tread 500 Tmath 500 Mean SE Mean SE Mean SE Mean SE Male 452.4 6.1 455.8 5.7 469.2 7.7 469.5 8.3 Isolated Female 467.7 4.7 473.2 5.3 484.9 4.5 461.4 5.6 Total 461.2 4.3 465.7 4.4 478.2 4.2 464.9 5.3 Male 472.6 3.6 470.7 3.1 490.5 3.8 497.4 3.7 Rural Female 503.7 2.0 505.2 2.1 506.7 2.4 504.2 2.4 Total 495.2 1.8 495.8 1.6 502.3 2.0 502.4 1.8 Male 516.2 6.4 506.8 7.5 506.0 9.2 518.6 7.2 Urban Female 538.8 3.2 538.2 3.8 510.5 3.6 520.4 4.4 Total 535.6 2.8 533.8 3.5 509.8 3.4 520.2 4.0 Male 472.0 3.1 470.6 2.6 487.1 3.2 493.0 3.5 Vietnam Female 507.3 1.6 508.7 1.8 505.0 1.6 502.9 2.1 Total 497.7 1.5 498.4 1.4 500.1 1.4 500.2 1.7 Figure 4.1: Teacher and pupil achievement by sex and by school location 89 Vietnam Reading and Mathematics Assessment Study Both boys and girls had Male teachers were less likely to be selected as an excellent teacher, had higher achievement fewer possession at home, fewer years of academic education, and fewer under female than male years of professional training. A test was made to see if boys learned better teachers. from male teachers and girls from female teachers but the test showed that, in general, both boys and girls had higher scores under the guidance of female than under male teachers. (c) Ethnic group of teachers Ninety-three percent of Grade 5 pupils had teachers who belonged to the Kinh ethnic group, the majority ethnic group in Vietnam. In other words only seven percent of pupils had teachers who were from ethnic minority groups, even though 14 percent of the pupils were from ethnic minorities. In isolated areas there were more non-Kinh teachers and this was particularly true in the Northeast and Northwest provinces. Indeed, in Cao Bang province only one percent of pupils had teachers who were Kinh. The Kinh teachers had higher achievement scores, especially in mathematics, than the non-Kinh teachers. (see Figure 4.2 below) (d) Teachers from local area The MOET has a policy Fifty-two percent of pupils had teachers who said that they were from the of trying to send good local area where they taught. However, this figure was only 42 percent in teachers to isolated isolated areas. The percentages of teachers from the local area in isolated areas. areas in the Northeast, Northwest and North Central regions were particularly low. It is unclear why this is so but it is possible to think of such reasons as children from these regions being from homes with lower parental education, teaching not being financially attractive profession, and the existence of the attitude 'if you get an education then use it to get out of poor conditions, not to stay there'. However, the MOET has a policy of trying to place better teachers in isolated areas. To what extent it is possible to ask Kinh teachers to learn the specific language of the minority area to which they are sent is not clear. Figure 4.2: Teacher achievement by ethnicity Kinh teachers had higher achievement than non-Kinh teachers. 90 Vietnam Reading and Mathematics Assessment Study A check was made on whether the pupils of local area and non-local area There was no difference teachers scored differently. No differences were found in either Reading or in achievement between Math. There was also no difference in the extent to which local and non-local local and non-local teachers met parents. teachers. (e) Academic education of teachers A question was asked of teachers about the general academic education they The average pupil had a had received. This consisted of three categories: primary education = 1, lower teacher who had nearly secondary education = 2, and upper secondary education = 3. This was 12 years of academic recoded as 5 years for primary education, 9 years for lower secondary, and 12 education years for upper secondary education. The overall mean for Vietnam Grade 5 teachers was 11.9 years meaning that the average pupil in Vietnam had a teacher who had 11.9 years of academic education. Again in isolated areas Pupils with teachers this was half a year less. Only in three regions (Red River Delta, North with more academic Central and Central Coast) did the average pupil have a teacher with more education achieved well. than 12 years of education. It should be noted that pupils with teachers with more than twelve years of education scored higher (approximately one fifth of a standard deviation) than pupils with teachers with less than twelve years of education. (f) Professional training The question asked, together with the multiplying factor given in parentheses at the beginning of each type of teacher training, was: In which type of training were you trained to be a teacher? (0) I have not been trained to be a teacher (1) Below Teacher Training High School level (1.5) Teacher Training High School Level (9+3) (2) Teacher Training High School Level (12+2) (3) Teacher Training College level (12+3) (4) Teacher Training University Level (12+4) The multiplying factor was said to be the equivalent of the actual number of The number of years years of teacher training. In general, pupils in isolated areas had teachers with of teacher varied the least number of years of teacher training. Teachers in rural areas tended from 1.7 in isolated to have more years and teachers in urban areas the most years of teacher areas up to 2.2 in training. In all cases the teachers coming from the local area had fewer years urban areas. of teacher training than those from outside the area. The lowest and highest province values for teacher training were Soc Trang The correlation (in the Mekong Delta) with 1.43 years and Da Nang with 2.56 years. The between the number correlations of the number of years of teacher training and pupil achievement of years of teacher were 0.13 and 0.11 respectively with math and reading. These results point training and pupil to the need to pay more attention to the levels of professional training of achievement was teachers when allocating local and non-local teachers to schools. about 0.12. 91 Vietnam Reading and Mathematics Assessment Study Table 4.3: Means and standard errors for teacher professional training by school location Years teacher professional training School location Teacher Mean SE Isolated Non-local 1.99 .022 Local 1.81 .027 Rural Non-local 2.12 .017 Local 1.98 .012 Urban Non-local 2.19 .035 Local 2.19 .029 Vietnam Non-local 2.11 .013 Local 2 .011 (g) A further measure of teacher quality Another way of examining teacher quality is to examine their subject-matter knowledge. The mean scores of teachers on the mathematics and reading tests have been presented in Table 4.4 Teachers in urban areas The scores have been scaled to the pupil scores of a mean of 500 and a had higher achievement standard deviation of 100. For Vietnam as a whole it can be seen that pupils than those in rural areas in isolated areas had teachers whose subject matter knowledge was about half who had much higher of a pupil standard deviation less than teachers in rural and urban areas. achievement than those Although the difference between the mathematics scores of teachers in rural in isolated areas. and isolated areas was statistically significant and large, the difference between the urban and rural was not so great. It was the pupils in the Northwest region who had teachers with the least subject-matter knowledge. Care must be taken when comparing the differences between schools in different areas within provinces. This is because the sampling errors for some means are quite large. Figures 2.9 and 2.10 in chapter 2 showed that, at the provincial level, there was a high correlation between the teacher and pupil scores in both reading and mathematics. In Table 4.4 it can also be seen that the standard deviation of achievement scores was slightly larger than in other areas. The requirements for teacher training changed in the period of 1981 to 1983. All teachers after 1983 should have trained at least under the 12+2 system. The previous system had fewer years of education and it might be that the achievement of teachers under the old system (and therefore now over 40 years of age) would have lower achievement in both reading and mathematics. This was checked and the results have been shown in the following table. 92 Vietnam Reading and Mathematics Assessment Study Table 4.4: Teacher achievement in reading and mathematics in different areas and regions Tread500 Tmath500 Region School location Mean SD SE Mean SD SE Isolated 591.6 89.0 45.4 508.4 16.0 12.0 Red River Rural 530.5 99.8 4.9 538.1 89.0 4.2 Delta Urban 533.7 84.2 10.7 554.8 76.0 8.9 Total 532.1 97.5 4.7 540.4 86.5 3.7 Isolated 485.2 112.7 10.6 458.8 107.3 10.0 Rural 497.4 102.2 4.1 489.6 102.0 4.2 Northeast Urban 496.6 103.4 12.0 491.2 109.1 11.6 Total 494.9 104.6 3.7 483.9 104.9 4.2 Isolated 450.1 112.2 15.0 430.7 116.5 17.1 Rural 476.0 89.3 10.6 458.2 94.7 11.3 Northwest Urban 473.7 107.9 20.5 486.5 91.0 18.8 Total 465.7 102.3 8.1 452.1 104.9 9.7 Isolated 516.8 75.4 11.1 500.8 97.2 18.6 North Rural 522.8 97.3 6.9 519.9 89.9 6.6 Central Urban 527.8 84.6 12.9 539.2 86.1 11.3 Total 522.7 93.4 5.3 520.2 90.8 5.7 Isolated 502.9 76.9 10.8 515.7 91.7 14.3 Central Rural 498.6 91.0 6.9 531.9 89.8 6.6 Coast Urban 517.9 84.6 7.3 534.2 87.3 7.5 Total 503.4 88.0 5.3 530.1 89.7 5.1 Isolated 467.7 105.3 18.6 464.0 93.3 17.7 Central Rural 495.9 93.6 11.3 515.5 92.2 12.8 Highlands Urban 496.3 81.2 10.1 514.5 84.5 11.5 Total 488.6 95.1 7.8 501.7 93.6 8.9 Isolated 477.1 93.3 12.6 453.6 99.9 12.0 Rural 488.1 95.6 6.3 464.1 97.4 6.8 Southeast Urban 510.8 89.8 7.1 524.9 91.1 7.5 Total 494.4 94.1 4.2 484.2 100.4 5.2 Isolated 446.3 95.3 9.0 449.8 103.7 11.5 Mekong Rural 471.9 101.2 4.3 469.5 98.5 4.0 Delta Urban 488.5 97.2 7.0 487.8 92.5 9.5 Total 470.1 100.4 3.9 469.0 99.0 4.0 Isolated 478.2 101.9 4.2 464.9 104.3 5.3 Rural 502.1 100.9 2.0 502.0 98.8 1.8 Vietnam Urban 509.5 92.6 3.4 519.7 93.6 3.9 Total 500.0 100.0 1.4 499.9 100.0 1.7 Table 4.5: Teacher characteristics by age Teacher age Years of teacher training Teacher reading score Teacher math score under 40 2.06 502.8 500.4 over 40 2.01 489.0 498.6 93 Vietnam Reading and Mathematics Assessment Study It can be seen that the older teachers had less teacher training, that the reading score of the older teachers were somewhat lower than the younger teachers, but that in mathematics there was no difference between the two groups. The lower scoring teachers (less than one standard deviation from the mean) were identified their characteristics have been shown in the table below. Table 4.6: Characteristics of low scoring teachers Teacher School location Non-Kinh Teacher Years Years Years teaching academic teacher Isolated Rural Urban teacher age experience education training % % % % Mean Mean Mean Mean with low score in reading 20.8 64.7 14.5 15 34.1 12.5 11.6 1.9 with low score in math 24.4 62.9 12.8 18.1 33.2 11.7 11.6 1.9 Total of low and high teachers 14.7 66.6 18.6 7.1 33.7 12.2 11.9 2.1 It can be seen that low score teachers came proportionally more from isolated areas, but that there were still some of them in rural and urban areas. More were non-Kinh teachers, but the low scoring teachers had about the same mean age as all teachers, the same number years of teaching experience, the same number of years of academic education, and slightly fewer years of teacher training. In this sense, it cannot be said that the lower scoring teachers were all trained under old 9+3 system. (h)Teacher years teaching experience Teachers with more The teachers were asked how many years they had been teaching years of teaching altogether. From Table 4.1, it can be seen that pupils in isolated areas experience tended to tended to have teachers with fewer years of teaching experience than those have pupils with higher in rural areas and in turn this was less than teachers in urban areas. achievement scores. However, from Figure 4.3 below it can be seen that there was a slight trend for teachers with more years of experience to have pupils with slightly higher achievement scores. If first year teachers are omitted, the trend linking years of teaching experience and pupil achievement is linear and positive for both reading and mathematics. Figure 4.3: Relationship between years of teaching experience and pupil reading and mathematics achievement 94 Vietnam Reading and Mathematics Assessment Study From the table below it can be seen that, as expected, teachers from the local area in isolated and rural areas had more years of teaching experience. Table 4.7: Years teacher teaching by school location and teacher from local/non-local School location Teacher Years teacher teaching Teachers from the Mean SE local area in Isolated Non-local 7.44 0.26 isolated and rural Local 11.33 0.38 areas had more Rural Non-local 10.32 0.20 years of teaching Local 13.32 0.17 experiences than teachers from Urban Non-local 15.68 0.31 non-local areas . Local 15.8 0.32 (i) Excellent teacher In Vietnam there is the practice of selecting or nominating teachers as excellent teachers. They are, so to say, exemplary teachers. There are various levels of excellent teachers. There is the school level, the district level, the provincial level and the national level. The first designation takes place at the school level. A school committee selects the excellent teachers at the school level based on classroom observation (teachers by teachers within the school) and pupil achievement. The teachers selected as excellent teachers at the district are selected from the school excellent teachers every two years. The provincial level excellent teachers are selected from the district excellent teachers once every three years. The national excellent teachers are selected from the provincial excellent teachers once every five years. It is also from the provincial level excellent teachers that the inspectors are selected. There were so few teachers in the sample who were excellent teachers at the national level that it was decided to omit the national excellent teachers data. In Table 4.1 the figures presented were the percentage of teachers nominated as excellent teachers irrespective of level. The percentages at the school and district levels are quite high which makes one wonder exactly which criteria are used to select those to be designated as excellent. There is a higher percentage of teachers in urban areas designated as excellent than in rural and with one or two exceptions the figure is higher for rural than for isolated areas. The percentages of excellent teachers by levels have been presented in Table 4.8 below. As can be seen from Table 4.8, a much higher percentage of teachers were There were more designated as excellent at the school level, many fewer at the district level, excellent teachers at the and even fewer at the provincial level. At the provincial level, most of school level than at the excellent teachers came from urban areas. In theory, the excellent teachers district level. There were should produce higher achievement in their pupils than those not designated even fewer at the as excellent. The results have been produced in Figure 4.4. provincial level, and they tended come from urban areas. 95 Vietnam Reading and Mathematics Assessment Study Table 4.8: Percentages of teachers designated as excellent teachers at different levels Region School Classified as 'excellent teacher' location School District Province % SE % SE % SE Red River Isolated 55 14.7 7 16 34 35 Delta Rural 51 2.3 31 1.7 4 0.9 Urban 42 4.7 40 4.5 9 2.3 Total 50 2.1 32 1.7 5 0.9 Northeast Isolated 35 3.3 15 2.7 3 1.1 Rural 47 1.9 25 1.7 2 0.6 Urban 38 3.2 38 4.1 8 1.8 Total 43 1.6 25 1.4 3 0.6 Northwest Isolated 30 5.6 9 3.2 1 0.6 Rural 62 4.5 14 2.7 1 1.1 Urban 45 7.7 29 8.2 16 5.3 Total 47 3.8 15 2 3 1.1 North Isolated 41 7.4 16 4.3 0 0.4 Central Rural 47 2.6 41 3 3 0.8 Urban 38 6.3 43 6.2 10 2.7 Total 45 2.2 39 2.6 3 0.7 Central Isolated 41 7.2 24 6.7 3 2.4 Coast Rural 50 3.8 21 3 3 0.8 Urban 54 4.7 16 2.7 9 2.1 Total 50 2.7 20 2.2 4 0.8 Central Isolated 37 9.2 18 7.6 . . Highlands Rural 41 5.3 39 6 1 1 Urban 42 7.8 38 7.9 16 5.2 Total 40 3.6 33 4 4 1.3 Southeast Isolated 48 7.4 3 1.5 . . Rural 46 3.4 13 1.8 1 0.5 Urban 49 4.2 21 2.9 6 1.6 Total 48 2.2 15 1.5 3 0.6 Mekong Isolated 38 4.2 10 2.2 2 0.9 Delta Rural 39 2.3 14 1.2 3 0.6 Urban 38 4.6 25 4.4 10 3.4 Total 39 1.8 15 1.1 4 0.6 Vietnam Isolated 39 2.4 13 1.6 2 0.7 Rural 46 1.1 26 0.9 3 0.3 Urban 43 2 30 1.8 9 1 Total 45 0.8 25 0.8 4 0.3 . = no data in this cell. Pupils of excellent The blue columns represent pupil reading scores and the red columns pupil teachers at the district math scores. Along the horizontal axis there are first those teachers not and province levels designated as excellent at the school level, then those designated as achieved more than excellent at the school level. This continues up to the provincial level. It pupils under non- excellent teachers. can be seen that at each level except at the school level the teachers designated as excellent had pupils with higher scores than teachers not designated as excellent. Although there was no difference at the school level, there was quite a difference at the district level and a great difference at the province level. 96 Vietnam Reading and Mathematics Assessment Study Figure 4.4: Pupil reading and mathematics achievement by levels of excellent teachers The teachers designated as excellent also had, in general, higher scores on the teacher math and reading tests than did the non-excellent teachers, as seen in Figure 4.5 below. As seen in Figure 4.5 above, excellent teachers at district and province level had higher achievement scores than non-excellent teachers. Although not shown in any table in this report, it is also a fact that excellent teachers tended to have slightly more years of general education and teacher training, to be more satisfied with their living conditions, and to have more possessions at home. (j) Perception of the effectiveness of in-service training In the final column in Table 4.1 the mean scores have been presented for the variable concerning the extent to which teachers believed that the in-service training courses they had attended had contributed to improving the quality of the teacher's teaching. The coding was as follows: 1 = Did not participate or not effective 2 = A little effective 3 = Effective 4 = Very effective The mean scores varied between 2.88 and 3.37. Teachers in isolated areas tended to perceive the in-service courses as more effective than teachers in 97 Vietnam Reading and Mathematics Assessment Study Figure 4.5: Pupil reading and mathematics achievement by levels of excellent teachers other areas. Is this simply because they never had much training and therefore any training is regarded as useful? The details about teachers' perception on in-service training have been presented below. Most teachers thought Ninety percent of teachers indicated that they perceived the regular in-service the regular in-service courses to be effective or very effective. Unfortunately, no detailed courses were effective. information was requested about the strengths and weaknesses of regular in- service courses. In-service teacher training is not easy to organize and run successfully. Therefore, one must be cautious in assuming, from the above figures, that all is well. In-service courses are often perceived to be of use by teachers because they have enjoyed their attendance at them. Benjamin Bloom of Chicago University, best known for his Taxonomy of Educational Objectives, used to say that in-service training courses were only worth twice their length. By this he meant that if the course lasted one day the teachers would then return to their schools and put into practice what they had been taught for two days. If the course lasted a week then the attempted implementation lasted not more than two weeks. Some research has shown that at least half the teachers in school should go to the in-service training if it is to be successful. A critical mass of teachers was required otherwise the two out of six teachers, say, would not be listened to by the others and all would return to the status quo very quickly. Twenty-four percent of pupils had teachers who rated their in-service courses as 'very effective' and 66 percent of pupils had teachers who perceived the courses as 'effective.' (k) Marital status of teachers 98 Vietnam Reading and Mathematics Assessment Study Table 4.9: Perception of effectiveness of regular in-service training by region and school location Perception of effectiveness of regular INSET Region School Did not attend or location Not effective A little effective Effective Very effective % SE % SE % SE % SE Isolated . . . . 63 30.0 37 30.0 Red River Rural 0 0.2 10 1.2 68 2.0 21 2.0 Delta Urban 1 0.7 13 3.8 62 5.4 24 5.1 Total 1 0.2 10 1.3 67 1.7 22 1.7 Isolated . . 6 1.6 73 3.9 21 3.9 Northeast Rural 0 0.3 5 0.8 75 1.9 19 1.7 Urban . . 5 1.8 83 3.0 12 2.5 Total 0 0.2 5 0.7 76 1.5 18 1.5 Isolated . . 4 2.8 67 6.1 29 6.5 Northwest Rural . . 3 1.9 83 4.0 15 3.5 Urban . . . . 69 9.0 31 9.0 Total . . 3 1.3 75 3.4 23 3.2 Isolated . . 6 3.6 67 9.0 27 9.3 North CentralRural 0 0.1 8 2.0 64 3.1 28 2.5 Urban . . 10 4.4 71 6.2 19 4.8 Total 0 0.1 8 1.6 65 2.5 27 2.2 Isolated 3 2.1 18 6.1 54 6.3 25 5.6 Central Rural 1 0.5 13 1.9 58 3.3 29 3.2 Coast Urban 1 0.8 19 3.3 55 4.3 25 4.4 Total 1 0.5 15 1.7 57 2.2 27 2.3 Isolated . . 10 4.1 72 7.3 17 6.9 Central Rural . . 5 3.1 71 6.5 25 5.8 Highlands Urban 2 2.2 18 8.2 65 7.9 15 4.1 Total 0 0.5 9 2.5 70 3.9 21 3.6 Isolated 1 0.8 18 5.0 61 7.0 20 6.4 Southeast Rural 2 0.8 15 2.0 65 3.0 18 2.5 Urban 2 1.6 25 3.7 57 3.7 16 2.3 Total 2 0.7 19 1.7 62 2.2 17 1.9 Isolated 1 1.1 3 1.2 59 4.8 37 5.0 Mekong Rural 0 0.2 5 1.1 61 2.3 34 2.3 Delta Urban 0 0.2 4 1.6 66 4.3 30 4.2 Total 0 0.2 5 0.8 61 1.8 34 1.9 Isolated 1 0.4 8 1.1 65 2.7 27 2.7 Vietnam Rural 0 0.1 8 0.6 66 1.1 25 0.9 Urban 1 0.4 14 1.4 65 1.8 20 1.6 Total 1 0.1 9 0.5 66 0.8 24 0.7 .= no data in this cell The MOET was concerned to discover how many of the teachers were 80% of pupils had married and how many children they had, especially in the remote areas. The teachers who were information on these variables has been reported in Table 4.10. It can be seen married and with 1.4 that for Vietnam as a whole 81 percent of pupils had teachers who were married and with an average of 1.4 children. children. Only 73 percent of pupils in isolated areas had female teachers who were But, only 73 of pupils in married compared with 88 percent in urban areas. As seen in Table 4.1, the isolated areas had teachers in isolated areas tended to be younger. Often the female teachers did female teachers who not come from the local area to which they had been posted to teach. Of those were married compared married 86 percent were local teachers and 76 percent of those married were with 86% in urban from a different area. areas. 99 Vietnam Reading and Mathematics Assessment Study Table 4.10: Marital status of female teachers in isolated, rural and urban areas Married Number of Children Region School location % SE Mean SE Isolated 100 0.0 1.13 0.33 Rural 93 0.9 1.47 0.04 Red River Delta Urban 95 1.7 1.60 0.07 Total 93 0.8 1.49 0.03 Isolated 83 3.4 1.29 0.09 Rural 91 1.0 1.56 0.04 Northeast Urban 96 1.3 1.79 0.07 Total 90 0.9 1.54 0.03 Isolated 65 5.9 1.25 0.17 Rural 91 2.6 1.65 0.10 Northwest Urban 98 1.5 1.66 0.09 Total 82 3.3 1.51 0.08 Isolated 70 5.5 1.32 0.15 Rural 81 2.2 1.42 0.06 North Central Urban 96 1.5 1.99 0.08 Total 82 1.9 1.49 0.05 Isolated 71 5.4 1.34 0.15 Rural 79 2.2 1.56 0.07 Central Coast Urban 88 2.4 1.92 0.11 Total 80 1.6 1.61 0.05 Isolated 73 5.8 1.17 0.19 Rural 73 3.9 1.48 0.15 Central Highlands Urban 91 3.3 2.11 0.16 Total 77 2.5 1.54 0.10 Isolated 67 4.3 1.05 0.15 Rural 65 3.0 1.03 0.06 Southeast Urban 76 3.5 1.32 0.08 Total 69 2.0 1.14 0.04 Isolated 73 3.6 1.01 0.10 Rural 72 1.8 1.14 0.05 Mekong Delta Urban 75 3.3 1.29 0.09 Total 72 1.5 1.14 0.04 Isolated 74 1.5 1.18 0.05 Rural 81 0.8 1.37 0.02 Vietnam Urban 86 1.3 1.61 0.03 Total 81 0.6 1.39 0.01 The provinces with the lowest percentage of married teachers have been presented below. All are in the Mekong Delta and Southeast regions. Most unmarried Province Married teachers were in Mekong % SE Delta and Southeast Tay Ninh 62 5.6 regions. Tien Giang 64 5.0 Ca Mau 64 4.5 Vinh Long 65 5.3 Ho Chi Minh 66 5.8 But not all of the provinces with the least number of married teachers were the provinces with the youngest teachers. The mean ages of the provinces with the youngest teachers have been given below. Only the province of Ca Mau is in the Mekong Delta. 100 Vietnam Reading and Mathematics Assessment Study Age (years) Province Mean SE Ca Mau 30.49 0.70 Binh Phuoc 30.73 0.70 Kon Tum 30.81 0.85 Hung Yen 30.84 0.70 Son La 31.01 1.03 Finally, an analysis was undertaken to discover if there was any difference between the pupils' scores of those with married and those with unmarried teachers. The results of this analysis have been given below. Table 4.11: Pupil and teacher achievement by school location and teacher married/unmarried Pupil achievement Teacher achievement School location Teacher Reading Math Reading Math Mean SE Mean SE Mean SE Mean SE Isolated Unmarried 459.2 7.53 468.3 7.03 481.8 7.92 475.0 8.90 Married 461.6 4.36 464.4 4.53 477.0 5.31 461.3 5.72 Rural Unmarried 482.1 3.21 481.2 3.31 500.6 5.14 493.2 3.59 Married 498.2 1.97 499.1 1.78 502.5 2.00 504.0 2.00 Urban Unmarried 527.1 5.18 521.4 7.25 513.0 9.63 521.9 8.79 Married 536.9 3.18 536.1 3.69 508.9 3.79 519.4 4.21 Vietnam Unmarried 483.5 2.81 484.0 2.96 498.5 3.83 493.5 3.20 Married 500.9 1.69 501.7 1.60 500.3 1.43 501.4 1.91 It can be seen that there was no difference in scores between pupils with There was not difference unmarried and married teachers in isolated areas but that in rural and urban in pupils' achievement if areas there were some, but not very large, differences. There were not teachers were married differences teachers' achievement according to whether they were married or or not. unmarried. A question was also asked about the number of children the teachers had. The Teachers in isolated Vietnam government has so far encouraged married couples to have a areas had 1.2 children maximum of two children. From Table 4.10 it can be seen that the average but in urban areas 1.6 number of children was 1.4. In isolated areas the average was 1.2, in rural children. areas it was 1.4 and in urban areas it was 1.6. It was in the urban Central Highlands where teachers were most productive (2.1).The least productive were the teachers in isolated schools in the Mekong Delta. For comparison, the figures have been taken from an existing publication. Region Children per woman in 1999 Red River Delta 2 Northeast 2.3 Northwest 3.6 North Central 2.8 Central Coast 2.5 Central Highlands 3.9 Southeast 1.9 Mekong Delta 2.1 Urban 1.7 Rural 2.6 Vietnam 2.3 Source: "Report: on results of population projections Vietnam, 1999-2024" (Statistical Publishing House, 2000) 101 Vietnam Reading and Mathematics Assessment Study Thus far, descriptions of the characteristics of teachers have been presented. In particular differences among regions and also between schools in isolated, rural and urban areas have been examined. As in most countries in the world, it was the urban areas where there were the 'better' teachers. In isolated areas of Vietnam, there were more male teachers, more younger and more inexperienced teachers, teachers with fewer years of general education and also teacher training (as well as having lower scores on the teacher math and reading tests), and fewer teachers designated as an excellent teacher. This implies that when allocating teachers to schools the relevant authorities did not take into account having equity in the quality of teachers across schools. There was inequity in At present, the way in which a teacher obtains a post in a school is that upon distribution of teachers graduating from a provincial teacher training college, he or she applies to a to school. Urban schools district office (or sometimes a provincial office) for a job. After an interview had the "best" teachers and an examination, if the candidate is successful, he or she is allocated to a in terms of education, school in that district (or province). As found in the data, this can lead to teacher training, and inequity in distribution of teachers across schools within a province. There subject matter would also appear to be inequity between provinces. This process should be knowledge. Isolated reviewed with the aim of improving equity in teacher resources at school. schools had "worst" One practice to consider is a point-gaining system used in several countries. teachers. If a teacher wishes to teach in an urban area or in a town or city, then he or she must gain points to be able to do so. The more that they teache in isolated and rural areas the more points they will gain in order to be able to be posted to an urban area. There were two other aspects of schooling in Vietnam that stood out in what has just been presented. The first concerned the non-difference between school level excellent teachers and those who were not excellent teachers at the school level. But, on the other hand the excellent teachers at the district and provincial levels had much better results from their pupils than the non- excellent teachers at these two levels. The question must be asked about what is going wrong at the designation stage at the school and national levels. The MOET should look into this matter in some detail and suggest revisions to the excellent teacher system. The primary teacher assessment criteria need to be reviewed. Finally, it was felt that the data on the teachers' perception of the effectiveness of the current in-service training programs were inadequate and that a special study should be conducted on the quality of in-service training programs. 102 Vietnam Reading and Mathematics Assessment Study Policy suggestion 4.1: The MOET and provincial authorities might wish to review the system of how teachers are allocated to schools and then revise the procedures in order to ensure more equity of the quality of the teaching force in all schools. For a short period of time the MOET might even consider having better quality teachers in the isolated schools in order to compensate for the relatively poor teachers they have had over the last so many years. It is also suggested that a rotation scheme be tried out whereby good teachers from good schools much spend some time in schools in isolated areas. The point-gaining system for teachers to move from isolated to urban areas was mentioned. However, it is recognised that it is very difficult to devise good incentives for teachers to work in isolated areas. Policy suggestion 4.2: The MOET might wish to consider devising various means of increasing the subject matter knowledge of existing and future teachers of reading and mathematics. This may involve updating the existing teacher training course content to take account of the new curriculum and ensuring a national level of subject matter knowledge for the teachers to become certified. For existing teachers it may involve special in-service courses or special programs targeted at specific groups of schools. What were the conditions of the Grade 5 classrooms? It is important to examine the conditions of the classrooms in which the teachers work. Were there sufficient sitting places for the pupils? Had they enough classroom library books? Were all schools equipped with the necessary classroom furniture? Did the teachers have the necessary teaching materials? And so on! If the classrooms were overcrowded, poorly furnished and poorly equipped then it would be difficult for the teachers to teach well. Information has been presented in Table 4.12 about the number of books in the classroom library, the number of sitting places available and whether or not the desks were moveable. (a) Number of books in classroom libraries (book corners) The number of books on a bookshelf or in a classroom corner is an important aspect of all classrooms in primary education throughout the world. Elley (1994) has shown how in nearly all classrooms in the world the more pupils are exposed to books the better will be their reading comprehension. It is important that a classroom bookshelf or corner have at least twice as many 103 Vietnam Reading and Mathematics Assessment Study Table 4.12: Means, percentages, and sampling errors for books in classrooms and space available Books in book Sitting places available Desk / tables Region School corner (number) (seats per pupil) movable location Mean SE Mean SE % SE Isolated 9.7 9.82 1.02 0.06 40 21.9 Red River Rural 16.7 2.37 1.04 0.01 54 2.8 Delta Urban 22.7 3.68 1.03 0.02 55 4.7 Total 17.6 2.05 1.04 0.01 54 2.4 Isolated 3.9 2.58 1.12 0.05 43 4.6 Rural 6.5 1.20 1.04 0.01 42 2.4 Northeast Urban 1.4 0.76 1.03 0.01 33 4.6 Total 5.2 0.95 1.06 0.01 41 1.7 Isolated 1.1 0.54 1.01 0.03 43 7.0 Rural 3.1 2.31 1.00 0.01 46 6.4 Northwest Urban 1.5 1.05 1.05 0.04 57 9.9 Total 2.0 1.06 1.01 0.02 47 4.3 Isolated 1.4 0.99 1.12 0.05 40 8.6 North Rural 11.3 3.80 1.03 0.01 40 3.5 Central Urban 9.1 3.80 1.07 0.02 42 7.8 Total 9.8 2.85 1.05 0.01 40 3.1 Isolated 5.4 3.23 1.02 0.01 40 8.0 Central Rural 5.8 3.50 1.03 0.01 24 2.6 Coast Urban 6.3 2.08 1.03 0.01 28 4.7 Total 5.8 2.24 1.03 0.01 27 2.1 Isolated 0.6 0.64 1.07 0.03 49 11.0 Central Rural 1.1 0.80 1.09 0.02 39 6.7 Highlands Urban 1.1 0.69 1.04 0.02 28 7.2 Total 1.0 0.48 1.07 0.01 39 4.7 Isolated 1.8 0.96 1.16 0.05 23 5.2 Rural 2.7 0.74 1.17 0.02 25 3.1 Southeast Urban 18.3 5.33 1.09 0.02 33 4.2 Total 8.2 2.06 1.14 0.01 28 2.1 Isolated 3.2 2.88 1.16 0.03 30 4.3 Mekong Rural 2.5 0.50 1.09 0.01 36 2.1 Delta Urban 7.1 1.75 1.06 0.02 43 5.0 Total 3.3 0.66 1.10 0.01 36 1.9 Isolated 2.8 0.92 1.11 0.02 37 2.3 Rural 7.9 0.94 1.06 0.00 40 1.1 Vietnam Urban 11.3 1.65 1.06 0.01 39 1.9 Total 7.8 0.70 1.07 0.00 39 0.9 books as pupils in the classroom. Pupils must be able to borrow the books and take them home to read. Not only should there be a good number of books but the books should represent a variety of genres. For example, at least the books could include some involving narrative prose, some expository prose and some document reading. 104 Vietnam Reading and Mathematics Assessment Study The average numbers of books in schools in different areas in the different There is virtually no regions have been presented in the first data column of Table 4.12. The first classroom library books remarkable fact about Table 4.12 is that there were hardly any books in the for pupils to read. classrooms at all. Even pupils in the Red River Delta region, which was the best endowed for books, only had an average of 17.6 books per classroom. For the whole of Vietnam fifth graders the figure was 7.8 books. In isolated schools in the Central Highlands, pupils were in schools where there was only an average of half a book per classroom. The provinces where there was less than one book per classroom have been listed below. Province Mean SE Kien Giang 0.0 0.00 Phu Yen 0.0 0.00 Ha Giang 0.0 0.03 Ca Mau 0.1 0.09 Lang Son 0.3 0.20 Quang Ninh 0.4 0.43 Soc Trang 0.6 0.50 Binh Duong 0.6 0.27 Kon Tum 0.7 0.66 Gia Lai 0.9 0.55 Dong Thap 0.9 0.54 Main school only classrooms had more books than classrooms in schools with satellites. Satellite schools were not well served with books. Table 4.13: Mean number of classroom books by main schools only/ schools with satellite Books Satellite campus schools Mean SE had fewer classroom Main schools only 14.0 1.7 library books than main Schools with satellite 4.8 0.5 schools only. Teachers with more general education and more years of teacher training tended to have more books in their classrooms. Where schools tended to have more school resources so there were classroom But, schools with more corners with more books. Where there was a school library, there were also resources in general, classrooms in the school with more books. The numbers of books per also had more classroom classroom per classroom library per school have been presented in the table library books. below according to whether the school had a library and allowed pupils to borrow books to take home. All of the above tables simply describe a very inadequate situation. (b) Sitting places available Where there are insufficient places for pupils to sit in the classrooms, life can There was adequate become unbearable for the pupils and the teachers. A question was asked of supply of sitting places in classrooms. 105 Vietnam Reading and Mathematics Assessment Study Table 4.14: Mean number of classroom books by school by school library system Book Mean SE No library 3.6 0.8 Cannot borrow 6.9 0.8 Can borrow 10.8 1.4 Total 7.8 0.7 teachers as to how many sitting places were available in their classrooms. This figure was then divided by the number of pupils in the classroom. Where the index is greater than 1.0 there are sufficient sitting places. Where the index is less than 1.0, then there are insufficient places. From Table 4.8, it can be seen that there was an adequate provision of sitting places in the classrooms. Only in Lai Chau province was the figure was less than 1.00. (c) Desks/tables moveable? 40% of pupils were in It is considered to be desirable to have desks or tables that can be moved in classrooms where desks order to allow a more flexible grouping of pupils in classrooms for different could be moved for activities. From Table 4.10 it can be seen that on average about 40 percent of flexible grouping pupils were in classrooms where the desks could be moved. There were very purposes. few differences among the different locations within each region but there were quite large differences among the regions. The provinces with the lowest percentages of desks that were moveable were: Province Desks / table movable % SE Dong Thap 7 3.7 Dong Nai 9 3.4 Binh Dinh 15 4.7 Ben Tre 17 4.4 Ba Ria-Vung Tau 18 5.6 Tay Ninh 18 4.8 Binh Phuoc 18 4.7 Kien Giang 20 5.1 It is interesting to note that it was not the poorly achieving provinces that had the fewest number of moveable desks. In fact, several of the poorer achieving provinces had nearly all desks that are moveable. This is an example of the government supply of moveable desks to the most needy first having been a success. (d) Classroom furniture and supplies According to the government, each classroom should have fresh air, be cool in summer, be warm in winter, and be safe for teachers and pupils. Each classroom should have the following materials; a desk and chair (which 106 Vietnam Reading and Mathematics Assessment Study should of appropriate size for pupils) for all pupils, one desk and chair for each teacher, a blackboard for writing, a platform (the place for teachers to stand and teach), fan and lamp for schools which have in the areas available of electricity, and a cupboard if schools can afford. Although the government stated that each school should have sufficient teaching materials and that teachers have a responsibility to use the teaching material effectively, the government did not state the details about which teaching materials each classroom should have and who is responsible for providing them. In the five-province study (Griffin, Patrick (1999) Vietnam Primary School Monitoring Report. Mimeographed report. World Bank in Vietnam), one of the variables on which huge variance was found among schools was classroom furniture that formed part of the School Resources construct. Teachers were asked to respond to the following question. The percentages of pupils in classrooms with each of these pieces of equipment/furniture have also been presented for all of Vietnam. Table 4.15: Percentages and standard errors for the availability of classroom furniture % SE There was a lack of A usable writing board 94 0.5 some items of furniture Chalk 97 0.3 in many schools. A wall chart of any kind 43 1.1 A cupboard or locker 7 0.5 One or more bookshelves 4 0.4 A classroom bookcase 13 0.6 A teacher table 99 0.2 A teacher chair 99 0.2 A fan 35 0.8 A lamp sufficient to light the classroom 55 0.9 Which of the following are available in your classroom or teaching area? (Please tick the appropriate box for each resource.) It was regrettable to note that six percent of pupils were in classrooms without a blackboard. In a few provinces (Ha Giang, Hoa Binh, Nghe An, and Ca Mau) only 85 percent of pupils were in classrooms with a blackboard. Only 42 percent of pupils were in classrooms with a wall chart of any kind. Very small percentages were in classrooms with a cupboard or locker, one or more bookshelves, and a classroom bookcase. A fan or lamp will depend in many cases on electricity being available. In 2001 it can be seen that there were still some 15 percent of pupils in schools without electricity. The high percentages of pupils in non-electricity schools were to be found in isolated schools in the Northeast, Northwest, 107 Vietnam Reading and Mathematics Assessment Study North Central, Central Highlands, Southeast and Mekong Delta regions. It is said that by 2003 all schools will have electricity. Table 4.16: Percentage and standard errors of pupils in classrooms without electricity In 2001, 15% of pupils Do not have electricity Region School location were in schools without % SE electricity, but by 2003 Isolated .0 .0 Rural 5.4 0.96 all schools should have Red River Delta Urban 9.0 3.42 electricity. Total 5.9 1.03 Isolated 54.3 5.17 Rural 19.0 1.94 Northeast Urban 3.2 1.77 Total 22.2 1.49 Isolated 69.0 6.92 Rural 41.9 5.33 Northwest Urban 4.4 4.43 Total 45.9 3.69 Isolated 51.3 10.15 Rural 5.5 1.72 North Central Urban 0. 0. Total 9.2 1.86 Isolated 24.8 7.53 Rural 12.8 2.47 Central Coast Urban 2.7 1.58 Total 11.7 1.83 Isolated 57.1 9.92 Rural 29.7 6.13 Central Highlands Urban 5.2 3.81 Total 29.0 4.11 Isolated 20.1 4.68 Rural 8.9 2.05 Southeast Urban 6.0 2.54 Total 9.1 1.46 Isolated 21.4 4.31 Rural 23.4 1.87 Mekong Delta Urban 5.3 2.28 Total 20.0 1.52 Isolated 38.5 2.62 Rural 13.8 0.83 Vietnam Urban 5.0 1.01 Total 15.0 0.74 The paucity of items of furniture is such that the MOET may well wish to conduct an audit, especially in isolated areas, in order to discover the extent of the problem. The number of classrooms requiring the furniture can be calculated. For supplies the following question was asked: 108 Vietnam Reading and Mathematics Assessment Study Table 4.17: Percentages and standard errors for the availability of teaching materials When teaching Vietnamese or mathematics to which of the following do you have access in your school? (Please tick the appropriate box for each resource.) % SE A Vietnam dictionary 63 0.9 Geometrical instrument 96 0.3 Teacher's guide (Vietnamese) 95 0.4 Teacher's guide (Mathematics) 95 0.4 Reference materiel 85 0.6 Pictures teaching illustrations 93 0.4 Again, it can be seen that there were some pupils (5%) in classrooms where the teachers professed not to have access to the teacher guides for Vietnamese and mathematics. Fifteen percent were in classrooms where teachers did not have access to reference materials. Whereas it is the government that provides the teacher guides it is the school that is responsible for providing a dictionary and reference materials. It was in the isolated areas that there was a paucity of a dictionary and reference materials. The Northwest region was particularly badly off. In general there was a high correlation between access to teaching materials and total school resources. That is to say that schools with fewer school resources also tended to be poor in terms of what teachers had access to for teaching materials. In Table 4.18 it can be see that for the total of furniture (maximum 10) and the total of teaching materials (maximum 6) that there was large variation among provinces. It can be seen that the national average was 5.3 for teaching materials. The lowest region was the Northeast with 3.8 items. The correlations between teaching materials and pupil math and reading achievement were 0.16 and 0.16 respectively. The average number of furniture items was 5.5 out of 10. This is a little low and has been commented on above. The correlations between furniture and pupil math and reading achievement were 0.23 and 0.23 respectively. The provisions of teaching materials and classroom furniture using standardized scores for both sets of materials have been presented in Figure 4.6. It can be seen that, in general, regions with more furniture also tend to have more teaching materials and equally that regions without much furniture also have fewer teaching materials. 109 Vietnam Reading and Mathematics Assessment Study Table 4.18: Mean number of items of teaching materials and furniture Teaching materials index Classroom furniture index Region School (Max=6) (Max=10) location Mean SE Mean SE Isolated 4.3 0.62 7.1 2.14 Red River Rural 5.6 0.03 6.2 0.07 Delta Urban 5.6 0.09 7.7 0.14 Total 5.6 0.04 6.5 0.07 Isolated 4.6 0.11 4.5 0.10 Rural 5.2 0.05 5.1 0.06 Northeast Urban 5.6 0.08 6.3 0.12 Total 5.1 0.04 5.1 0.05 Isolated 3.8 0.16 4.1 0.13 Rural 4.7 0.15 4.7 0.09 Northwest Urban 5.3 0.16 6.0 0.29 Total 4.5 0.10 4.7 0.11 Isolated 4.5 0.30 4.4 0.10 Rural 5.4 0.08 5.7 0.13 North Central Urban 5.7 0.11 6.8 0.14 Total 5.4 0.07 5.7 0.10 Isolated 4.9 0.19 4.6 0.23 Rural 5.4 0.06 5.3 0.07 Central Coast Urban 5.8 0.06 6.4 0.16 Total 5.4 0.06 5.5 0.07 Isolated 4.2 0.35 4.4 0.21 Central Rural 4.8 0.15 4.6 0.14 Highlands Urban 5.5 0.10 5.3 0.26 Total 4.8 0.12 4.7 0.10 Isolated 5.1 0.13 5.0 0.28 Rural 5.3 0.08 5.4 0.09 Southeast Urban 5.6 0.09 6.6 0.10 Total 5.4 0.05 5.8 0.07 Isolated 5.2 0.12 4.6 0.08 Rural 5.2 0.05 4.7 0.04 Mekong Delta Urban 5.5 0.10 5.6 0.14 Total 5.3 0.04 4.8 0.04 Isolated 4.7 0.06 4.6 0.07 Rural 5.3 0.02 5.4 0.03 Vietnam Urban 5.6 0.03 6.5 0.06 Total 5.3 0.02 5.5 0.03 It is of interest to note that the Mekong Delta had an national average number of teaching materials but very few items of furniture. Furthermore the Central Coast region had more than national average teaching materials, but only average amount of furniture. It is also true that the more satellite campuses a school had, the less furniture and fewer teaching materials it had as can be seen from the table below. Is this because resources are allocated to schools rather than to school campuses? 110 Vietnam Reading and Mathematics Assessment Study Schools in the Northwest, Central Highlands, and Mekong Delta were worst-off for furniture and teaching materials. Figure 4.6: Classroom resources by region Table 4.19: Means and standard errors for classroom furniture and teaching materials by the number of satellite campuses Number of satellites Classroom furniture Teaching Materials Satellite campus schools Mean SE Mean SE were particularly badly- 0 6.1 0.06 5.5 0.03 off. 1 to 2 5.4 0.05 5.3 0.03 3 to 4 4.9 0.05 5.1 0.05 5 to 6 4.7 0.09 4.7 0.08 7 to 8 4.5 0.09 4.6 0.13 9 to 10 4.1 0.15 4.5 0.20 11+ 4.2 0.11 4.3 0.21 Vietnam 5.5 0.03 5.3 0.02 However, it is the large differences among regions as well as the dearth of provision in some areas that stand out and the MOET must correct this inequity among regions. Policy suggestion 4.3: The MOET might consider a review in annual school census of amount of classroom furniture and teaching materials in order to identify shortfall. 111 Vietnam Reading and Mathematics Assessment Study How much time did teachers devote to their work? Of interest is the actual amount of time that teachers devoted to their school work and even other work. The results have been presented in Table 4.20 for the numbers of hours of actual teaching, the numbers of hours preparing lessons and marking homework, and finally time spent on any other work in order to earn money. Table 4.20: Means and sampling errors of teacher loads, meeting parents, and earning extra money Hrs. per week preparing lessons and making Frequency of meeting Hrs. extra teaching and Region School Hrs. per week teaching parents other activities location homework Mean SE Mean SE Mean SE Mean SE Isolated 13.9 1.22 17.2 1.82 2 0.00 1.9 2.14 Red River Rural 18.0 0.14 19.8 0.26 2.02 0.01 1.3 0.11 Delta Urban 17.5 0.36 19.7 0.85 1.99 0.04 1.6 0.25 Total 17.9 0.13 19.7 0.25 2.02 0.01 1.4 0.09 Isolated 16.1 0.16 21.3 0.42 2.03 0.03 1.2 0.17 Rural 16.3 0.11 21.1 0.22 2.06 0.01 1.3 0.08 Northeast Urban 17.0 0.23 21.2 0.48 2.08 0.05 0.9 0.13 Total 16.3 0.08 21.1 0.18 2.05 0.01 1.2 0.07 Isolated 15.7 0.23 21.5 0.75 2.05 0.05 1.1 0.19 Rural 15.5 0.17 21.2 0.60 2.11 0.06 1.5 0.21 Northwest Urban 15.1 0.33 22.7 1.05 2.04 0.16 1.0 0.19 Total 15.5 0.13 21.6 0.40 2.08 0.04 1.3 0.12 Isolated 16.4 0.36 21.4 1.23 2.03 0.05 1.2 0.24 North Rural 16.8 0.12 21.9 0.46 2.11 0.03 1.3 0.14 Central Urban 17.3 0.41 20.8 0.60 2.07 0.04 1.8 0.32 Total 16.8 0.12 21.7 0.37 2.09 0.02 1.3 0.12 Isolated 15.7 0.26 22.4 0.90 2.02 0.08 1.1 0.29 Rural 16.7 0.19 21.9 0.47 2.11 0.03 1.7 0.13 Central Coat Urban 16.1 0.29 21.7 0.64 2.07 0.03 1.8 0.19 Total 16.4 0.15 22.0 0.36 2.09 0.02 1.6 0.11 Isolated 16.2 0.30 21.6 1.10 2.13 0.10 2.1 0.35 Central Rural 16.3 0.27 21.7 0.79 2.08 0.04 1.9 0.28 Highlands Urban 16.6 0.42 21.8 0.80 2.08 0.09 1.8 0.38 Total 16.3 0.19 21.7 0.53 2.09 0.04 1.9 0.20 Isolated 16.4 0.31 19.5 0.67 2.08 0.06 2.0 0.32 Rural 16.1 0.12 18.1 0.45 1.99 0.03 1.7 0.13 Southeast Urban 17.0 0.26 18.3 0.55 2.1 0.03 1.9 0.13 Total 16.5 0.12 18.4 0.27 2.04 0.02 1.8 0.09 Isolated 16.1 0.13 20.5 0.63 2 0.05 1.5 0.18 Mekong Rural 16.4 0.13 19.1 0.28 2 0.02 1.9 0.08 Delta Urban 16.3 0.28 19.4 0.61 2.02 0.04 1.7 0.14 Total 16.3 0.10 19.4 0.23 2 0.02 1.8 0.06 Isolated 16.1 0.08 20.9 0.27 2.04 0.02 1.5 0.09 Rural 16.7 0.05 20.3 0.13 2.05 0.01 1.5 0.05 Vietnam Urban 16.8 0.12 20.0 0.23 2.06 0.02 1.6 0.07 Total 16.7 0.04 20.4 0.11 2.05 0.01 1.5 0.04 112 Vietnam Reading and Mathematics Assessment Study (a) The teaching load What was the teacher's load in Vietnam? The results of the analyses have The average pupil was been presented in Table 4.6. It can be seen that the average Grade 5 pupil was taught by a teacher taught by a teacher teaching 16.7 hours per week. But this was 16.8 in urban teaching 16.7 hours per areas and 16.1 in isolated areas. The province with the most hours of week. But ,in full day teaching was Nam Dinh (21 hours) and that with the least hours of teaching schools, this was 18.7 was Binh Duong (13 hours). However, in Nam Dinh, 95 percent of pupils hours. were following the full-day schooling whereas in Bing Duong this was only 11 percent of pupils. Indeed, in general, the number of hours taught was related to the kind of days schooling being taught - full-day or shift (half day). Teachers teaching a full- day taught 18.7 hours, and those teaching first or second shift only 16.2 and 16.0 hours respectively. The percentages of pupils following a 'full-day schooling' were 4 in isolated areas, 20 in rural areas and 31 in urban areas. It is said that the ethnic minority children study fewer hours and it was certainly the case that teachers of Kinh pupils reported that they had teachers teaching 16.8 hours per week on average and those of non-Kinh only 16.1 hours per week. What is of interest is the generally low number of hours. From the OECD data base it was possible to calculate the number of hours per week teaching time in the OECD countries' primary schools. The following is the list: Australia 22.2 Vietnamese primary Austria 18.0 teachers taught Belgium (Fl.) 22.3 fewer hours per Belgium (Fr.) 21.3 week on average Czech R. 16.2 than teachers in Denmark 15.3 Finland1 17.3 many other France 26.0 countries. Germany 20.1 Greece 19.5 Hungary 21.0 Iceland 16.6 Ireland 25.0 Italy 21.2 Japan 18.1 Korea, Republic of 22.6 Netherlands 23.3 New Zealand 25.0 Norway 18.8 Portugal 24.0 Scotland 25.0 Spain 23.8 Switzerland 23.0 Turkey 16.8 United States m 113 Vietnam Reading and Mathematics Assessment Study The MOET set 20 hours as the maximum teaching per week, but did not set the minimum teacher hours. Given the low number of hours taught - compared with other countries, the MOET might wish to reconsider what the minimum and maximum contact hours per week should be. (b) Hours preparing lessons and marking homework The average teacher Lessons have to be prepared and this is very time-consuming especially for load was 37 hours per the beginning teachers. Homework, as already seen in chapter 3, is a very week. important part of the learning process and the more it is marked the better will the pupils achieve. The average pupil had a teacher who said that she prepared lessons and marked home work for 20.4 hours per week. This is more hours than they spent actually teaching and one wonders about the veracity of the teachers' replies. If the replies are regarded as valid then the average teacher load (teaching hours plus hours for preparing lessons and marking homework) in Grade 5 was 37.1 hours per week. Most OECD countries were around 40 hours (Table D3.2 on page 227 in Education at a Glance 2001, Paris: OECD). (c) Frequency of meeting parents The MOET encourages teachers to meet parents at least three times per school year. The first time, at the beginning of the first semester, is to inform parents of the class schedule. The second time is at the end of the first semester in order to inform parents of their children's progress, and the third time is at the end of the school year in order to inform parents about the final marks for the year and also to deal with any problems. Teachers met with Postlethwaite and Ross1 showed that, in many countries, schools in which the parents twice a year. school head and teachers had more contact with parents were more effective in promoting the reading achievement of pupils. In other words, schools where school heads and teachers had more contact with parents scored better than could be expected after taking into account the socio-economic background of their pupils. From Table 4.6 it can be seen that the average number of times was twice per year. Given that it is often necessary to meet parents of individual pupils more frequently than at the general class meetings of all parents one would have expected means of 3 to 4. Clearly, the notion of having to meet parents individually is not one that is seen as important by teachers. (d) Time spent earning extra money Teachers were poorly In 2001, a starting teacher in primary school in Vietnam had a base salary of paid... 329700 (210000 dong * 1.57, from the year 2000) Dong per month (the 1Postlethwaite T. N., and Ross, K, N. (1992) Effective Schools in Reading Hamburg: IEA 114 Vietnam Reading and Mathematics Assessment Study multiplier number 1.57 is for the beginning teachers. This multiplier number increases according to the years of teacher teaching). The equivalent in US dollars is 22 (US$1=15000dong in June 2002). The average per capita income in 2001 when the testing was carried out was 390 dollars per year. It can easily be seen that the teachers were hardly paid a living wage, especially in the larger cities. When this occurs, people tend to take on second, third and even fourth jobs. Another secondary side-effect that frequently occurs is that corruption sets in and it has even been known in some countries for teachers to find all sorts of ways of extracting more money from parents. It is clear that with other competing jobs, those working more hours on other tasks will be more tired for their ordinary teaching task. The mean number of hours spent earning extra money for the whole of and said that they Vietnam was 1.5 hours per day. At the same time, the standard deviation was worked extra 1.5 hours 1.9 hours per day. The minimum number of extra hours worked was zero and per day for more money. the maximum was 10 hours per day. There is some doubt about the veracity of these replies and MOET may wish to consider undertaking a separate study of this matter. Policy suggestion 4.4: a) the MOET might wish to reconsider the minimum and maximum contact hours per week for teachers, given the existing low number of hours compared with other countries; b) the MOET may wish to undertake a special study of teachers meeting parents and then improve the teacher training programs for improving school, family, and community interactions. Which materials and books did pupils have? No matter how well the classrooms may be equipped, it is clear that the Having the wherewithal pupils must have their pens, pencils, paper, and above all the textbooks they to learn in school is need. At the time of testing the government policy was that parents should important... buy these books for their children. Thus, a child possessing the materials is dependent on the parents being able to pay for them. In needy cases, the provincial authorities may provide the major textbooks free of charge. The major textbooks are Vietnamese 5, Vols. 1 & 2, Math 5, Moral Education 5, and Nature and Society (History, Geography, and Science). The number of books possessed can thus, in one way, be seen as an index of parental wealth. On the other hand, it can also be seen as the extent to which the pupils were able to cover the curriculum content. The data collected on the above have been summarized in Tables 4.21, Table 4.22, Table 4.23, and Table 4.25. The percentages have been presented as the percentages of pupils NOT possessing the items. In Table 4.21 there are ten items of materials that each pupil should have. These are: notebook for 115 Vietnam Reading and Mathematics Assessment Study writing, notebook for drafts, black pencil, coloured pencil, calculator, fountain pen or ballpoint pen, 'small table' (this is a small kind of slate for writing on), white chalk or other writing material, school bag, and ruler. but, some pupils did not Nearly all pupils had a notebook for writing and a notebook for drafting. have notebooks, pencil or Nearly all had a school bag and a ruler. About seven percent did not have a pen, or white chalk etc. black pencil and 18 percent did not have coloured pencils. About 78 percent of pupils did not have a calculator and it is assumed that the primary school, curriculum in mathematics did not require a calculator. This may change as the curriculum changes and an eye should be kept on this. Sixty-six to 75 percent of pupils had a 'small table' and white chalk, usually to write on the small tables. Twelve percent did not have a pen and it was 'pen' that had a reasonable sized correlation with achievement (0.17). It could well be that a pen could only be afforded by the better-off families. This was particularly in In general it was the pupils in the Mekong Delta and also in isolated areas, the case in the Mekong especially in the Central Highlands and Northwest, who did not have some Delta and in isolated of the materials. The correlation between the total materials possessed and areas in the Central achievement was also in the order of 0.17 and hence every effort should be Highlands and made by the provincial and district authorities to have all children supplied Northwest. with the required materials. It might be that the MOET does not fully agree with the particular list of materials used in this study. But it is not a difficult task to conduct an audit of which materials are possessed by the pupils. Table 4.21: Percentages and sampling errors of materials' items NOT possessed by pupils % SE Notebook for writing 1.2 0.1 Notebook for draft 3.0 0.1 Black pencil 6.8 0.2 Coloured pencil 18.0 0.3 Calculator 78.6 0.3 Pen 11.3 0.3 Small table 33.1 0.6 White chalk 24.8 0.5 School bag 2.3 0.1 Ruler 1.5 0.1 What was the situation concerning textbooks? It is the task of the provincial authorities to furnish the major textbooks for the pupils in isolated and difficult areas. In all other areas the parents have to buy the books. Without textbooks it is very difficult for pupils to learn. The percentages of pupils NOT having the different textbooks have been presented in Tables 4.22, Table 4.23, Table 4.25, and Table 4.26. These results need to be read in conjunction with Table 4.28 where the percentages of pupils actually studying the various subjects have been given. Over 99 percent of pupils 116 Vietnam Reading and Mathematics Assessment Study studied Vietnamese, Mathematics, Moral Education and Science, History and Geography and more than 97 percent of pupils studied the rest of the subjects. Again it was the pupils in the isolated areas who tended not to study a subject. Table 4.22: Percentages and sampling errors of Vietnamese books NOT possessed by pupils % of pupils Not possessing SE Nearly all pupils had Vietnamese 5 Vol.1 1.0 0.1 two main Vietnamese textbooks, but many did Vietnamese 5 Vol.2 0.7 0.0 not have the prescribed Vietnamese story book 22.8 0.4 supplementary books. Vietnamese exercise book 5 Vol.1 48.5 0.5 Vietnamese exercise book 5 Vol.2 53.9 0.5 Vietnamese other books 61.1 0.4 Vietnamese dictionary 91.8 0.2 Vietnamese Essay Grade 5 25.6 0.4 The provinces with more than two percent of pupils not having the two major Vietnamese textbooks have been given in the table below. Province Vietnamese 5 Vol.1 (% of Vietnamese 5 Vol.2 (% of pupils NOT possessing) Province pupils NOT possessing) % SE % SE Ha Giang 3.93 1.10 Hoa Binh 2.65 0.59 Son La 2.72 0.85 Ha Giang 2.57 0.87 Hoa Binh 2.67 0.71 Gia Lai 2.35 0.88 Soc Trang 2.19 0.63 Kien Giang 2.00 0.54 Nearly all pupils had the two main Vietnamese textbooks. But, again it was in the isolated schools in the Northeast, Northwest, North Central and Central Highlands where there were up to four percent of pupils without these basic textbooks. The Vietnamese story book is not supplied by the government and hence fewer pupils had them and again pupils in urban areas were more likely to have these story books than the pupils in rural and isolated areas. The correlation between possession of the Vietnamese story book and total possessions in the home is 0.24 and thus it is suspected that non-possession of a story book is associated with poverty. The Vietnamese exercise books are also not supplied by the government. Only half of the pupils had them. Pupils from isolated areas were less likely to have them. Some of the provinces with high non-possession have been listed below. 117 Vietnam Reading and Mathematics Assessment Study In some provinces, Vietnamese exercise book 5 Vietnamese exercise book 5 more than 60% of Vol.1 (% of pupils NOT Vol.2 (% of pupils NOT pupils did not have Province possessing) Province possessing) the Vietnamese % SE % SE Grade 5 exercise Tay Ninh 81.4 2.27 Tay Ninh 83.6 2.24 books. Lao Cai 80.2 3.38 Lao Cai 83.1 3.03 Ha Tinh 77.6 3.75 Thua Thien-Hue 81.7 2.37 Thua Thien-Hue 77.5 2.64 Quang Tri 79.5 2.67 Quang Tri 76.3 2.68 Lang Son 79.0 3.00 Tuyen Quang 75.1 3.23 Ha Tinh 78.8 3.67 Lang Son 74.7 3.12 Soc Trang 78.3 2.50 Yen Bai 73.6 3.25 Tuyen Quang 78.0 2.95 Soc Trang 72.2 2.75 Yen Bai 75.2 3.21 Tien Giang 70.3 2.94 Nghe An 74.4 3.25 Ha Giang 69.6 2.84 Tien Giang 74.1 2.83 Hoa Binh 68.8 3.93 Ha Giang 73.6 2.76 Nghe An 68.4 3.63 Hoa Binh 72.3 3.66 Khanh Hoa 68.4 2.39 Ninh Binh 71.5 3.43 Phu Yen 67.1 3.20 Phu Yen 71.2 3.00 Thanh Hoa 66.0 3.93 Khanh Hoa 70.8 2.30 Quang Nam 65.1 3.99 Thanh Hoa 69.6 3.78 Phu Tho 64.2 3.80 Vinh Phuc 68.5 2.81 Da Nang 62.9 2.88 Bac Lieu 68.4 3.13 Quang Binh 62.8 4.28 Phu Tho 67.2 3.82 Vinh Phuc 62.0 2.97 Quang Nam 66.9 3.87 Cao Bang 61.5 3.18 Ninh Thuan 66.7 3.03 Lai Chau 61.0 4.34 Cao Bang 65.9 3.14 Ninh Binh 60.6 3.74 Tra Vinh 65.6 3.42 There is only a near zero correlation between the possession of the exercise books and reading achievement as measured by the Grade 5 pupil reading test. It is up to the curriculum development unit to rethink the necessity of these exercise books. Either they are not being used effectively or they are of no use. Most pupils did not have a dictionary or 'Vietnamese other books'. Again it is for the curriculum unit to rethink the necessity of these books. It is possible to calculate the number of pupils in the whole of Grade 5 not having these books. An example of how to do this has been given in Appendix 4.1. It can be seen that if all [pupils not having books were to be given them by the State it would cost about 12 million US dollars. But, the question to be answered first is if all of the books are really required. What was the situation for mathematics books? The results have been presented in Table 4.23. As seen from Table 4.28 over 99 percent of pupils studied mathematics. 118 Vietnam Reading and Mathematics Assessment Study Table 4.23: Percentages and sampling errors of mathematics books NOT possessed by pupils Items % pupils NOT possessing SE The picture for Math 5 0.6 0.04 mathematics Math Exercise book 41.6 0.40 textbooks was Math Exercise book 5 Vol.1 46.6 0.47 similar to that Math Exercise book 5 Vol.2 52.2 0.45 for Vietnamese Math other books 68.1 0.30 textbooks. Nearly all pupils had the major textbook, Math 5, but again pupils in isolated areas were less likely to have this book - even though it is the Provincial Office that decides where to distribute the free books. But, about half of the pupils did not have the other math books. These can be purchased from a book shop. It can be seen in the table below that in rural and urban areas (but not in isolated areas) there was a difference in math achievement between those having 'other books' and those not having them. Table 4.24: Mean pupil math score by school location and possession of math other books School location Math other books Pupil Math score Mean SE Isolated do not have 469.2 4.25 have 472.6 6.23 Rural do not have 488.0 1.60 have 509.2 2.06 Urban do not have 517.5 2.67 have 556.8 3.26 It is a little unclear who exactly was using the 'other books' and for what reason. This matter should be explored further. Only about half of the pupils had exercise books and again the possession of such books was not related to pupil achievement. The necessity of having these books should be rethought. Although this study had reading and mathematics as its main focus, the opportunity was there to ask about the possession of all the other books in the curriculum. The results have been presented in Table 4.25 below. For Vietnam as a whole it can be seen that high percentages of pupils had In other subject Moral Education 5 (but more than 20 percent of pupils did NOT have this matters, the book in the provinces of Vinh Phuc, Hoa Binh, Kon Tum, Ha Nam, Phu Tho possession of and Lao Cai), Nature and Society 5 Science, History and Geography, textbook situation Technology 5, Music 5, Art 5, and Health 5. But many pupils did not have was worse. 119 Vietnam Reading and Mathematics Assessment Study the Moral Education Exercise book, the exercise books for Science, combined History and Geography, Music and Health. Fewer than 25 percent of pupils had an English for Primary Education textbook and only three percent had the Informatics for Primary Education texts. Given that more than 97 percent of pupils stated that they were studying all of the above subjects some of the high percentages not possessing books must be worrying to the authorities. Table 4.25: Percentages and sampling errors of other books NOT possessed by pupils Item % pupils NOT possessing SE Moral Education 5 10.1 0.20 Moral Ed.Exercise book 70.8 0.36 Nature and Society 5 Science 5.9 0.14 Nature and Society 5 History and Geography (combined) 11.3 0.18 N&S 5:Exercise book-Science 71.8 0.45 N&S 5:Exercise book-History 73.1 0.45 N&S 5:Exercise book-Geography 72.3 0.45 Technology 5 15.6 0.28 Music 5 22.0 0.36 Music exercise 5:Vol1 82.2 0.36 Music exercise 5:Vol2 80.4 0.34 Art 5 13.1 0.26 Heath 5 12.0 0.24 Heath 5 Exercise book 82.8 0.35 Informatics for Primary Education 97.0 0.15 English for Primary Education 76.5 0.56 The picture of the pupils in the isolated areas having fewer books of any kind held true for the books reported in Table 4.25. In some cases, there were up to 30 percent of pupils not having books in isolated areas. In all, there were 10 materials' items, 8 Vietnamese books, 5 Mathematics books, and 16 books covering other subjects. These totals have been summarised for all regions in Table 4.26. It can be seen that it is in the isolated areas that pupils tended to have fewer materials and books. From further analysis, it The correlations between each of the totals in Table 4.11 and pupil total was seen that it was possessions in the home are high (about 0.30). This means that schools mainly the children of reporting more materials and textbooks were schools that had pupils who poor parents who did reported more possessions at home. This is a sign that it is poverty that plays not have textbooks. a role in whether pupils have the materials and textbooks. There may be a Many were in isolated case for the MOET having to ensure that provincial authorities supply pupils areas. from needy families with materials and books. The possession of materials and Vietnamese books, mathematics books and books in other subjects has been depicted in Figure 4.4. 120 Vietnam Reading and Mathematics Assessment Study Table 4.26: Means and sampling errors of number of items possessed Material Vietnamese Mathematics Other subjects' Region School (Max:10) books (Max:8) books (Max:5) books (Max:16) location Mean SE Mean SE Mean SE Mean SE Isolated 7.8 0.48 4.9 0.50 3.6 0.19 8.2 1.00 Rural 8.2 0.03 5.4 0.04 3.7 0.03 8.8 0.11 Red River Delta Urban 8.6 0.08 6.2 0.08 4.3 0.07 10.9 0.22 Total 8.2 0.03 5.5 0.03 3.8 0.03 9.1 0.10 Isolated 7.5 0.09 4.2 0.11 2.1 0.10 6.6 0.23 Rural 8.1 0.04 4.8 0.04 2.7 0.04 7.9 0.08 Northeast Urban 8.6 0.06 5.4 0.10 3.3 0.09 9.4 0.24 Total 8.1 0.03 4.8 0.04 2.7 0.04 8.0 0.07 Isolated 7.0 0.16 4.2 0.15 2.3 0.13 5.9 0.30 Rural 7.8 0.12 4.2 0.13 2.2 0.09 6.1 0.22 Northwest Urban 8.3 0.18 5.1 0.19 2.8 0.21 8.0 0.16 Total 7.6 0.09 4.3 0.09 2.4 0.07 6.3 0.15 Isolated 7.0 0.14 4.1 0.14 2.2 0.12 6.4 0.35 Rural 7.9 0.07 4.5 0.05 2.4 0.06 7.2 0.10 North Central Urban 8.4 0.12 5.2 0.15 3.0 0.17 8.9 0.28 Total 7.9 0.06 4.6 0.05 2.5 0.05 7.4 0.10 Isolated 7.4 0.24 4.3 0.16 2.4 0.12 6.9 0.26 Rural 8.6 0.04 4.6 0.05 2.5 0.06 7.8 0.07 Central Cost Urban 9.0 0.06 5.4 0.07 3.1 0.06 8.4 0.09 Total 8.6 0.06 4.8 0.05 2.7 0.05 7.8 0.06 Isolated 6.8 0.21 4.4 0.15 2.4 0.11 6.2 0.35 Rural 7.7 0.12 4.8 0.10 2.8 0.09 7.4 0.17 Central Highlands Urban 8.5 0.11 5.7 0.12 3.5 0.11 8.2 0.15 Total 7.7 0.09 4.9 0.07 2.9 0.07 7.4 0.11 Isolated 8.1 0.11 4.8 0.13 2.6 0.15 7.6 0.11 Rural 8.6 0.04 4.9 0.06 2.8 0.06 8.0 0.07 Southeast Urban 8.7 0.04 5.8 0.06 3.6 0.06 8.6 0.10 Total 8.6 0.03 5.3 0.04 3.1 0.04 8.2 0.06 Isolated 8.0 0.08 4.5 0.07 2.3 0.08 7.7 0.13 Rural 8.2 0.04 4.8 0.04 2.7 0.03 8.0 0.07 Mekong Delta Urban 8.7 0.06 5.2 0.08 2.9 0.08 8.5 0.11 Total 8.3 0.03 4.8 0.04 2.7 0.03 8.0 0.05 Isolated 7.6 0.05 4.4 0.05 2.4 0.04 7.0 0.09 Rural 8.2 0.02 4.9 0.02 2.9 0.02 7.9 0.04 Vietnam Urban 8.7 0.03 5.6 0.03 3.4 0.04 9.0 0.07 Total 8.2 0.02 5.0 0.02 2.9 0.01 8.0 0.03 121 Vietnam Reading and Mathematics Assessment Study Figure 4.4: Materials and textbooks by region It can be seen that where pupils in particular regions had materials they also had books. The Central Highlands and Mekong Delta pupils had above average materials but below average books. Pupils in the North Central region were particularly below the average for Vietnam. Most pupils did not have Given that some of the pupils did not possess books, how many of them to share their textbooks. actually shared their books with other pupils? This information was collected for the Vietnamese textbook and the mathematics textbook and the results have been presented in Table 4.27. Over 98 percent of pupils had their own textbooks. Where pupils had to share or where it was only the teacher who had a textbook, they tended to be in isolated areas, especially in the Northwest and Central Highlands regions. Table 4.27: Percentages and sampling errors of pupils with and sharing textbooks Textbooks Has/sharing % SE Only teacher has book or do not have 0.1 0.02 Vietnamese Share with others 1.6 0.06 Use book by myself 98.3 0.07 Only teacher has book or do not have 0.5 0.03 Mathematics Share with others 1.0 0.06 Use book by myself 98.5 0.07 For information the provinces with the most problems have been shown below. 122 Vietnam Reading and Mathematics Assessment Study Province Vietnamese textbooks Province Mathematics textbooks Only teacher has book Share with others Use book by my self Only teacher has book Share with others Use book by my self % SE % SE % SE % SE % SE % SE Hoa Binh 0.4 0.19 4.5 0.93 95.1 0.96 Kon Tum 0.5 0.21 4.4 3.11 95.0 3.08 Son La 0.6 0.30 4.2 0.96 95.2 1.05 Gia Lai 1.7 0.73 2.8 1.01 95.4 1.45 Kon Tum . . 4.4 1.33 95.7 1.33 Kien Giang 0.4 0.23 3.3 1.45 96.3 1.44 Gia Lai 0.5 0.30 3.7 1.02 95.8 1.23 Ha Giang 1.8 0.46 1.5 0.56 96.6 0.87 Kien Giang 0.3 0.17 3.7 1.07 96.0 1.06 Cao Bang 1.1 0.43 2.1 0.68 96.8 0.85 What 'opportunity to learn' did the pupils have? From the above, some doubt exists about the pupils being able to learn all of the curriculum material because they did not have the required textbooks in order to do so. In September 2001, the government announced that all pupils in Grade 4 and 5 should study all nine subjects. A further attempt was made to discover the likelihood of the pupils being taught all of the curriculum. The pupils were asked which subjects they learned. The results have been presented in Table 4.28. Table 4.28: Percentages and sampling errors of pupils learning different subjects Subject % SE Nearly all pupils said Vietnamese 99.9 0.01 they learned all 9 Mathematics 99.9 0.02 subjects in Grade 5, Moral Education 99.1 0.10 even if ometimes they Science, History, Geography 99.5 0.05 did not have a Technical work 97.2 0.17 textbook. Music 97.3 0.16 Art 98.1 0.11 Physics Education 97.0 0.17 Health Education 98.9 0.09 What is known as the 165 week curriculum is the full curriculum (165 weeks across 5 years) and it has been implemented in many schools. But there used to be the 120 week (now no longer existing) and 100 week curricula (existing only in a few schools). These were used in the areas which were more remote and isolated and where it was hard to recruit teachers. When the figures presented in Table 4.28 were examined in terms of isolated areas and regions it was in the Northeast, Northwest and North Central as well as in the Central Highlands where the problems occurred. The two provinces with the most problems were Lai Cai and Lai Chau. 123 Vietnam Reading and Mathematics Assessment Study Are all the textbooks The following steps would seem to be desirable. First the Curriculum Unit really necessary? should review the necessity of having all of the books and materials and then come up with a list of books and materials that ALL pupils should have. Each provincial director of education should be asked to conduct an audit of lacking materials and textbooks. A plan should then be made to furnish all pupils with the required materials and books starting in the regions where there were problems of scarcity of materials and books in specific isolated areas. At that point in time (when the provision of good teachers to isolated areas can be undertaken) it will be possible to insist on the 165 week curriculum being taught or at least a fuller curriculum than is now the case. There is in fact nothing new in these suggestions since the problem has been known to the authorities for some time. The results of this study have just served to remind all in the system of the problem. Policy suggestion 4.5: The MOET might (a) wish to review and state which materials and textbooks as a core packet are really required by all pupil; b) conduct a provincial review of the lack of materials and books in all schools; c) re-allocate teachers to schools especially in the poor and isolated areas to achieve an equity of the allocation of quality teachers to all schools, and d) work towards one national consolidated curriculum in all schools. What aspects of the teaching function were designed to improve the quality of education in place? Several indicators were examined about the educational environment in the schools. The teaching load and the extent to which teachers met with parents have already been examined. In this section, teachers' perceptions of the role of the inspectors (sometimes called education advisors), and the major sources of teacher satisfaction have been examined. (a) Teachers' perceptions of the role of inspectors The changing role of the inspectorate has become an important issue in many educational systems. Rather than visiting schools and classrooms in order to 'judge' the teachers, it is now the role of inspectors to help teachers in many ways. In Vietnam, the current government policy is that inspectors are selected at the district level from among the teaching force in the district. The provincial education offices organize the training of the inspectors and advisors. It is the Heads of the District Education offices who decide which schools should receive a visit from the inspectors. The first question, however, must be how many times the teachers have been visited by an inspector over the last three years. Teachers were asked how 124 Vietnam Reading and Mathematics Assessment Study Vietnam report: Volum 2, Chapter 4: Teachers Characteristic 125 many times they had been visited by an inspector in the years 1999, 2000, and up to April, 2001. The results have been summarised in Table 4.29 together with the percentage of teachers reporting no visits by inspectors. According to the rule books, teachers should have an inspector visit at least once per three to five years. Schools should have it at least once a year. Table 4.29: Means and sampling errors of the number of inspectors' visits Year Number of visits to schools Teachers no visits Schools received one Mean SE % SE visit per year from 1999 0.99 0.02 29.4 0.93 inspectors. Teachers 2000 0.99 0.02 32.3 0.79 received a visit every 3-5 2001 1.03 0.02 30.2 0.80 years. Total 2.99 0.06 14.7 0.64 It can be seen that 14.7 percent of pupils had teachers who said that over the three year period they had not had a visit from an inspector. In any one year, there were about 30 percent of pupils who had teachers who said that an inspector had not visited. The average pupil was in a school that had been visited about once a year. There were some differences among the regions. For example, pupils in the Red River Delta had teachers who said that they had been visited 3.7 times over the three years whereas in the Mekong Delta it was only 2.5 times. In general, pupils in urban schools had more teachers who said that they had received more visits than those in isolated areas. For example, pupils in isolated schools in the Northwest region only had 2.4 visits over the three years. Although visits to schools were not related to the distance the school was But, schools in isolated away from many amenities (xdist), schools with satellite campuses received areas and with satellite 2.8 visits over the three years while main schools received 3.3 visits. Pupils campuses received fewer in schools receiving more visits had higher achievement in mathematics visits. (correlation 0.125 after controlling for type of area in which the school was located). In terms of provinces, Binh Duong had less than one visit over the three years In some provinces, and 57 percent of pupils had teachers who said that there had not been an schools received fewer inspector's visit over the three years. The province of Lao Cai was well visits and in other served by the inspectorate. The provincial director of education should be provinces more visits. asked how he managed to have such a success rate. Is there anything that he was doing that could be copied by other provincial directors? The teachers were also asked what the inspector(s) did while visiting. The results have been summarised in Table 4.30. The table can be interpreted as the teachers' perceptions of the role of the inspectorate as of April, 2001. 125 Vietnam Reading and Mathematics Assessment Study Table 4.30: Teachers' perceptions of the role of the inspectorate In general, Teachers agreeing teachers Aspect of roles % SE appreciated Pedagogical role Suggests new ideas 50.6 0.8 what the Clarifies education objectives 48.5 0.8 inspectors Explains curriculum content 38.0 0.9 did to help Recommends new teaching materials 36.4 0.8 them. Suggests new teaching methods 73.9 0.7 Critical and/or advisory role Advises me 58.6 0.9 Criticizes me 28.0 0.9 Finds faults and reports them 36.3 0.9 Provides a little to help me 87.0 0.6 Professional development role Encourages professional contacts with other teachers 43.7 0.9 Provides in-service training 60.0 1.0 Provides information for professional self-development 52.8 0.8 But, 30% of pupils had The fact that 74 percent of pupils had teachers who said that the inspectors teachers who felt that suggested new teaching methods was pleasing since the government is inspectors came to find focusing now on the introduction of new teaching methods. However, the fact fault with them with that only 38 percent of pupils had teachers who said that the inspectors them and criticize them. explained curriculum content was rather disappointing. It was also heartening The view of the to see that many teachers (87%) thought that the inspectors were of some inspectorate was very help to them. Nevertheless there were quite a few who also perceived the different from province inspectors as finding faults and reporting them. In isolated areas in the to province. Southeast 'criticise me' was 40 percent. The inspectors are the main source of professional development for the teachers so that with only about 53 percent of pupils having teachers who endorsed the statements concerning professional development, the overall picture was not good. There were wide disparities in the responses to nearly all of these statements within each of the provinces. It would seem that the various roles of the inspectorate are viewed rather differently in different schools and that a clear statement of the overall role of the inspectorate would be welcomed. Policy suggestion 4.6: The MOET should clarify the roles of the inspectorate to all schools emphasizing the supportive rather monitoring role, and then have a good control mechanism to ensure that that the system works well. (b) Being observed by colleagues and the School Head Teachers were observed Within the Vietnamese system, it is expected that the school head and by their peers between colleagues visit each others classrooms, observe teaching, and then give once and twice a month. feedback on the teaching to the teacher concerned. On average, 45.4 per cent of pupils had teachers who said that they had been observed teaching by the 126 Vietnam Reading and Mathematics Assessment Study school head once a term and another 45.6 percent whose teachers said that they had been visited by the school head once a month. About 1.6 percent said that they had never been visited by a head and about 7.4 percent said that they had only been visited about once a year. Pupils in isolated areas had teachers who said that they had been visited by colleagues (other teachers in the same school for the purposes of having their teaching observed) about 1.83 times (SE was .06) in the last month, those in rural areas 1.93 times (SE .03) in rural areas and 1.66 (SE 0.05) times in urban areas. Overall for Vietnam the number of times having been observed by colleagues in the last month was 1.86 (SE .02) times. It is reported that the observation and feedback by colleagues is useful to the teachers and has a beneficial effect on their work. If this is so, then presumably the more that teachers are observed by colleagues, then the more their pupils should learn. (c) Sources of teacher satisfaction Teachers feeling motivated is an important pre-requisite for good and caring teaching in order to improve the quality of learning and education in general. Teachers were given the opportunity to respond to 16 possible reasons for satisfaction with their jobs. These reasons have been grouped under five headings in Table 4.31: living conditions, general conditions, school facilities/equipment, relationships with others, career advancement, and pupil achievement. The results have been presented in terms of the percentage of pupils having teachers who said that the particular sources of satisfaction were NOT important for them. Table 4.31: Percentages and sampling errors for sources of teacher satisfaction Not important % SE Travel distance to school 29.0 0.85 Living conditions Availability of teacher housing 24.8 0.72 Quality of teacher housing 25.2 0.62 Location of school 30.4 0.82 Level of salary 1.4 0.20 General conditions Timely payment of salary 4.9 0.39 Preferential allowances of teachers 1.5 0.19 Quality of school buildings 0.3 0.09 School facilities/equipment Quality of classroom furniture 0.2 0.08 Availability of classroom supplies 1.3 0.17 Quality of school management and administration 0.8 0.18 Relationships with others Amicable relationships with other teachers 0.3 0.07 Good relations with community 0.9 0.15 Opportunities for promotion 52.5 0.82 Career advancement Opportunities for professional development 0.6 0.11 Perception of teacher status Pupil's perception of status of teachers 6.3 0.39 Education outcomes of pupils Seeing pupils learn 0.1 0.03 127 Vietnam Reading and Mathematics Assessment Study Teachers were asked About 30 percent of pupils had teachers who thought that the travel distance about what they liked to school was not important. Interestingly, teachers in isolated areas did not about their jobs. They find that travel distance was a problem and it is assumed that most of these were concerned about teachers are in government housing (see Table 4.33). About a quarter of all their salary and preferential allowances, pupils had teachers who said that neither the availability of teacher housing and about the lack of nor the quality of teacher housing was important. But about the same good resources in number said that it was very important. Thus although there is a problem it schools and classrooms. is perhaps not as bad as originally thought. More teachers in isolated areas They were somewhat believed that the availability of housing was important although not in the concerned about the public's perception of Red River Delta. The same observation was true for the quality of teacher the status of teachers, housing. but they gained a lot of satisfaction from seeing Under the rubric of 'general conditions' it can seen that there were few who their pupils learn. thought that that the questions of the level of salary, the timely payment of salary and the preferential allowances for teachers were not important. Indeed, those saying that it was a very important source of satisfaction for them were 61 percent, 42 percent and 61 percent respectively. There is obviously a problem here that should be reviewed. Under school facilities/equipment the first factor was the quality of school buildings and 99 percent of pupils had teachers who thought that was important. More or less the same was true for the quality of classroom furniture and the availability of classroom supplies. Very high percentages of pupils had teachers who thought that the other factors mentioned in the table above were important with the exception of opportunities for promotion. The question is why this percentage is so much lower than the other percentages. Perhaps it is because the system of promotion is related to not only 'doing a good job' but also to other factors. Hence, teachers tended to think that promotion was not important factor for satisfaction. The teacher experts in the MOET must have thoughts on this matter. The teachers were asked to give their three most important reasons for satisfaction. These results have been presented in Table 4.32 below. In the first choice the two most important reasons were 'seeing my pupils learn' followed by 'the quality of the school building'. In the second choice the first two reasons were 'seeing my pupils learn' followed by 'the quality of school building'. In the third choice the two first reasons were 'policy of preferential salary for teachers' and 'seeing pupils lean'. 128 Vietnam Reading and Mathematics Assessment Study Table 4.32: Percentages and sampling errors for three most important reasons for satisfaction Percentage rating as most important MOST IMPORTANT % SE Seeing pupils learn 43.6 0.86 Quality of school buildings equipment 17.2 0.64 Level of salary's teacher 9.8 0.49 Quality of school management and administration 6.5 0.45 Opportunities for professional development 5.6 0.42 Percentage rating as most important SECOND MOST IMPORTANT % SE Seeing pupils learn 20.4 0.67 Quality of school buildings equipment 15.1 0.50 Quality of school management and administration 13.0 0.62 Opportunities for professional development 10.9 0.56 Quality of classroom furniture 9.2 0.47 Percentage rating as most important THIRD MOST IMPORTANT % SE Policy of preferential salary for teachers 21.2 0.63 Seeing pupils learn 13.0 0.55 Opportunities for professional development 11.4 0.60 Level of salary's teacher 10.2 0.42 Quality of school buildings equipment 9.5 0.42 There are clearly major problems with the teachers' perceptions of the quality of school buildings and equipment and the system of preferential payment to teachers. Indeed there would also appear to be a problem with many of the general conditions, the school classroom furniture and supplies, relationships with others, opportunities for professional development and the teachers' perception of the public's perception of teachers. Given the various problems associated with the teachers' perceptions of the views of teachers, whether it be about the inspectorate or educational advisors or the teachers' perceptions of their sources of satisfaction, it would seem to be wise to establish a special working group on the views of teachers and to undertake further work on this matter in order to discover the ways in which a teacher's lot may be improved. What were the living conditions of teachers? It is clear that for teachers to be content and to be able to concentrate on their work as teachers that they must have good living conditions. For example, where did they live and what items did they possess at home, and what were their sources of satisfaction with their work? The relevant data have been presented in Table 4.33. 129 Vietnam Reading and Mathematics Assessment Study Table 4.33: Percentages and sampling errors of teacher accommodation, possessions and satisfaction with accommodation Where Teacher live Satisfaction with accommodation Teacher Region School Leased location With parent accommodation Gov. housing Own house possessions Not satisfied Satisfied Very satisfied % SE % SE % SE % SE Mean SE % SE % SE % SE Isolated 11 26.7 34 35.0 . . 55 14.7 10.6 0.86 81 43.7 19 43.7 . . Red River Rural 38 1.8 2 0.4 4 0.7 57 1.8 11.7 0.10 26 2.2 70 2.2 5 0.8 Delta Urban 28 3.3 4 2.1 15 3.5 54 4.5 15.6 0.31 15 3.7 81 3.7 4 1.7 Total 36 1.6 3 0.7 5 0.8 56 1.6 12.3 0.12 25 2.1 71 2.0 5 0.7 Isolated 19 3.2 0 0.5 34 3.9 47 3.9 8.8 0.25 34 3.7 61 3.7 5 1.5 Northeast Rural 21 1.3 1 0.4 11 1.3 67 1.7 10.7 0.14 20 1.5 77 1.7 4 0.8 Urban 10 1.8 1 0.8 11 2.4 78 2.7 12.9 0.21 19 4.2 76 4.1 5 1.4 Total 19 1.1 1 0.3 16 1.2 65 1.4 10.6 0.10 22 1.3 73 1.2 4 0.6 Isolated 17 5.2 2 1.4 46 6.1 35 5.5 8.1 0.30 49 6.3 47 6.7 5 2.6 Northwest Rural 17 3.4 5 3.5 10 2.4 68 4.5 10.2 0.25 27 4.6 69 4.6 5 1.9 Urban 3 2.0 12 7.0 11 4.8 74 9.5 12.3 0.73 25 8.5 72 8.6 3 2.1 Total 15 2.5 5 2.1 24 2.9 56 3.3 9.7 0.23 35 3.3 61 3.4 4 1.4 Isolated 34 5.3 0 0.3 32 5.4 34 7.0 9.3 0.50 30 6.6 66 6.1 4 1.8 North Rural 29 2.6 2 0.7 16 2.7 53 3.1 10.8 0.23 33 3.4 62 3.5 5 1.4 Central Urban 13 3.3 1 1.0 4 2.4 82 3.6 14.2 0.42 15 4.1 79 4.2 6 2.0 Total 28 2.1 2 0.5 16 2.1 54 2.8 11.1 0.20 30 2.6 65 2.7 5 1.1 Isolated 26 5.7 . . 29 6.2 45 6.1 8.8 0.49 34 8.4 63 8.6 3 1.6 Central Rural 42 3.0 1 0.5 3 0.7 54 3.0 10.6 0.16 28 2.8 68 2.8 4 1.1 Coast Urban 25 3.2 4 1.6 9 2.4 62 4.5 13.6 0.29 23 3.8 72 3.6 5 1.7 Total 36 2.1 2 0.5 8 1.4 54 2.1 11.0 0.18 28 2.4 68 2.4 4 0.8 Isolated 16 5.2 2 0.9 28 6.9 55 8.9 8.8 0.49 46 8.9 50 8.2 4 3.1 Central Rural 12 3.0 8 3.9 22 3.9 58 4.8 10.0 0.34 32 5.7 62 6.0 7 2.4 Highlands Urban 6 3.0 . . 16 4.5 78 4.8 12.8 0.40 16 5.0 76 6.1 8 4.1 Total 12 2.1 4 2.1 22 2.9 62 3.3 10.3 0.25 32 3.9 62 4.0 6 1.6 Isolated 32 7.5 6 2.6 14 3.0 48 5.8 10.2 0.58 29 5.1 66 6.6 5 3.0 Southeast Rural 42 3.1 3 0.8 13 1.5 43 2.6 11.0 0.21 27 2.4 66 2.7 7 1.5 Urban 38 3.2 4 1.2 5 1.4 53 3.0 14.5 0.30 20 2.8 70 3.3 9 2.1 Total 39 2.1 4 0.7 10 1.0 47 2.0 12.1 0.17 25 1.8 67 2.0 8 1.2 Isolated 38 4.2 5 1.9 19 3.7 38 4.7 9.1 0.23 29 3.2 63 3.4 8 2.8 Mekong Rural 45 1.8 2 0.6 6 0.8 47 1.9 10.2 0.14 24 1.2 70 1.4 6 0.9 Delta Urban 32 4.3 4 1.6 11 2.9 52 5.0 12.7 0.36 26 4.4 66 4.4 7 2.2 Total 42 1.4 3 0.6 9 0.9 46 1.7 10.4 0.11 25 1.2 68 1.2 7 0.8 Isolated 27 2.0 3 1.1 27 1.7 43 2.1 9.1 0.16 35 2.0 60 2.0 5 0.9 Vietnam Rural 34 0.9 2 0.3 9 0.7 54 0.9 10.8 0.07 26 1.0 69 1.1 5 0.4 Urban 25 1.6 3 0.6 9 0.9 62 1.8 13.9 0.13 20 1.5 74 1.6 7 0.9 Total 31 0.7 2 0.3 12 0.5 54 0.7 11.1 0.06 26 0.8 68 0.7 6 0.3 (a) Teacher accommodation Of half of all pupils had Just over half of all pupils had teachers who said that they lived in their own teachers who lived in homes. Thirty-one percent had teachers who lived with their parents, 12 their own houses and percent in government housing and two percent lived in rented 31% who lived with accommodation. In isolated areas 27 percent of pupils had teachers who lived their parents. In isolated in government housing. Since many teachers (see Table 4.1) in isolated areas areas, many lived in were non-local they could not live with their parents and were very unlikely government housing. to buy their own accommodation. Hence many lived in government housing. This was particularly true in Kon Tum, Lao Cai, and Gia Lai. (b) Teacher possessions in the home Teachers were asked how many of the following 21 items they had in the 130 Vietnam Reading and Mathematics Assessment Study homes where they lived: daily newspaper, weekly or monthly magazine, Teachers had about radio, TV set, video cassette recorder (VCR), cassette player, telephone, half of the 21 items refrigerator/freezer, motorcycle, bicycle, piped water, electricity (mains, listed as home generator, or solar), table to write on, chair, bookcase or bookshelf, watch, possessions. lamp to read books, wardrobe, cooker (electric or gas), washing machine, and calculator. These items were summed for each teacher and then averaged for each area in each region. The means have been presented in Table 4.33. It can be seen that the average pupil was taught by a teacher who possessed 11 items. As might be expected the pupils in isolated areas had teachers with fewer items and this was particularly true in the isolated areas in the Northeast, Northwest, North Central, Central Coast and Central Highlands. It is to be noted that pupils who had teachers who were had more possessions tended to have higher achievement (.11 and .12 for reading and mathematics after controlling for pupil home background). (c) Teacher satisfaction with living accommodation A question 'Are you satisfied with your living accommodation?' was asked of Three quarters were the teachers. The responses were 'Not very satisfied', 'Satisfied', and 'very satisfied with their satisfied'. The percentages of pupils with teachers in these categories have living conditions. been presented in Table 4.33. In general pupils had teachers who were satisfied/very satisfied with their living accommodation. But, 26 percent of pupils had teachers who were not satisfied with their accommodation. These tended to be more in isolated areas. Those teachers living with their parents or in their own homes tended to be more satisfied (and to have more possessions) than those living in rented accommodation or in government housing (see Figure 4.5). Figure 4.5: Teacher possessions and satisfaction with housing There is clearly a need to devise incentives for teachers, especially in isolated areas, to own their homes, but every effort should be made by the authorities to do so. 131 Vietnam Reading and Mathematics Assessment Study What were some of the teaching methods used? Teachers were asked It is a sine qua non that teachers vary their teaching methods according to about their teaching which pupils they are teaching and the particular educational objectives that methods in mathematics. they are teaching. In other words, the best method to use for Pupil Thanh is not necessarily the best method to use for Pupil Binh, let alone Pupil Linh. It may well require different teaching methods even for the same educational objective for all three pupils. At the same time, it is also clear that the teacher will use different methods for the different educational objectives that she teaches. Some objectives are better learned through frontal teaching, others in small group work, and yet other objectives through individual work. Despite the above, an attempt was made to try and capture something of what teachers of mathematics perceived to be desirable in terms of pupil activities, their goals, and their approaches in the teaching of mathematics. (a) Pupil activities in mathematics arranged by teachers The teachers were asked to rate the importance they attached to several activities when they taught mathematics. The percentages of teachers in Vietnam rating the activities as very important have been recorded in Table 4.34. Table 4.34: Percentages and sampling errors of teacher activities in mathematics Rating very important Most believed in pupils' Activities % SE reciting tables but not in Working in pair or group to solve math problems 33.7 0.76 working alone or in pairs or in studying and Working alone on problems 31.3 0.65 interpreting graphs. Preparing projects or posters to be shown in class 40.6 0.79 Using practical equipment e.g scales, calculators, etc. 49.7 0.93 Homework assignments 66.0 0.80 Studying and interpreting graphs, etc 4.9 0.34 Reciting tables, formulae 92.6 0.42 Quizzer, tests. Exams, etc. 27.9 0.75 The major activities for pupils that teachers deemed to be important were 'reciting tables, formulae, etc., homework assignments, 'using practical equipment'. The last important activity was seen to be studying and interpreting graphs. It was thought that this activity may have not been possible because of the difficulty of acquiring appropriate graphs to study. 'Working in pairs or groups to solve math problems' was much less popular than expected from good math teaching/learning activities. It is suggested that the curriculum centre be asked to review the figures in 132 Vietnam Reading and Mathematics Assessment Study Table 4.34 and compare them with the learning activities suggested in the Teacher Guide for Mathematics. The curriculum group should then be asked to comment on what steps might be undertaken to improve matters. A principal component analysis was run on the activities' variables. Only one component was found. It was: Variable Loading Preparing projects or posters to be shown in class 0.746 Using practical equipment e.g scales, calculators, etc. 0.745 Studying and interpreting graphs, etc 0.603 It has been interpreted as 'practical work in math class'. It can be seen that the preparation of the posters and using practical equipment was slightly more highly correlated with the component than studying graphs. (b) Teacher goals in mathematics Seven typical goals of teachers teaching mathematics were listed for the teachers to rate their importance. The results have been presented in Table 4.35. Table 4.35: Percentages and sampling errors for teacher goals in mathematics Very important' 90% of pupils had Goals % SE teachers who had the Basic numeracy skills 90.18 0.476 pupil acquisition of Problem solving (transfer of skills to everyday life) 43.27 0.66 basic numeracy skills as Thinking skills (different ways of solving problems) 41.87 1.057 the major goal of their Confidence in solving math problem 53.72 0.887 work.. Only 40% had the Satisfaction from doing math 53.22 0.988 goal of having the Opening up career possibilities 12.54 0.649 children think of Development of life skills 22.44 0.75 different ways to solve a problem. From Table 4.35, it can be seen that the only goal rated as 'very important' by many teachers was 'basic numeracy skills'. About half of the Grade 5 pupils had teachers who endorsed 'confidence in solving math problems', 'satisfaction from doing mathematics', 'problem solving (transfer of skills to everyday life)', and 'thinking skills (different ways of solving problems)' as being 'very important'. Relatively few pupils had teachers who endorsed 'opening up career possibilities' and the 'development of life skills' as 'very important'. The pattern of responses was much the same in all regions of Vietnam. However, in isolated schools the percentage of pupils with teachers endorsing 'basic numeracy' and 'thinking skills' as very important was relatively low. On the other hand, the percentage of pupils with teachers in isolated schools who endorsed 'problem solving' and 'opening up career 133 Vietnam Reading and Mathematics Assessment Study possibilities' was higher than in rural and urban areas. The different approaches used often by teachers have been presented in Table 4371. The approach of 'using local materials' was used more in isolated areas than elsewhere and the correlation with 'opening up career possibilities' was 0.17. Finally, those teachers who believed that basic numeracy was very important had pupils who reached higher skill levels than the pupils of the teachers who considered 'basic numeracy' as only 'important' or 'not important'. Table 4.36: Mean math skill level by the teacher goal of basic numeracy Basic numeracy Mean Math skill level SE Not important 4.19 0.775 important 4.37 0.072 very important 4.65 0.014 When the teachers were asked which was the most important of the seven goals listed they stated that it was 'basic numeracy'. A principal component was undertaken on the variables in teacher goals. Two components were identified. They were: Variable Loading Confidence in solving math problem 0.816 Satisfaction from doing math 0.816 This first component has been named 'emphasizing confidence and satisfaction in math classes'. The second component, shown below, has been named 'math for life use'. (c) Teacher approaches when teaching mathematics Ten different approaches were listed. These have been presented in Table 4.37 together with the percentages of pupils with teachers said that they used them 'often'. Nearly 100% Table 4.37: Percentages and sampling errors of teacher approaches of pupils had teachers who Approaches Pupils with teachers who used the approach 'often' often taught % SE the whole Using everyday problems 91.7 0.41 class as a Teaching the whole class as a group 97.2 0.24 group. Few Teaching in small group 18.1 0.70 did small Teaching individually 28.9 0.72 group teaching Teaching through question and answer techniques 88.9 0.54 and nearly a Relating to everyday life situations 50.4 0.86 third taught Basic skills training 95.5 0.32 pupils Explaining math processes 82.6 0.69 individually. Using available local materials 16.8 0.73 134 Vietnam Reading and Mathematics Assessment Study From the above table it can be seen that more than 90 percent of pupils were in classes where the teachers used 'everyday problems' for the pupils to gain 'basic skills training' through 'teaching the whole class as a group'. More than 80 percent of pupils had teachers who 'explained the math processes' through 'teaching through question and answer techniques'. Fifty percent of pupils had teachers who related what they taught to 'everyday life situations'. Fewer than 20 percent of pupils had teachers who taught in small groups and only 29 percent ever taught 'individually' often. Examining the relationship between goals and approaches (what teachers want to do and what they actually do) there should be a relationship between 'basic numeracy skills' and 'basic skill training'. The correlation was 0.17 which is not high and means that there is still quite a lot of difference between goals and approaches for basic literacy. Pursuing this line of thinking it should be noted that the correlation between 'problem solving' and 'relating to everyday life situations' was 0.20. The correlation between 'development of life skills' and 'relating to everyday life situations' was 0.22. A principal component analysis was run on the variables in 'teacher approaches'. Two components were identified. They were: Variable Loading Teaching the whole class as a group 0.776 Teaching through question and answer techniques 0.776 This first component has been named 'traditional teaching'. Variable Loading Teaching in small group 0.756 Teaching individually 0.710 Using available local materials 0.510 The second component has been named 'modern teaching'. The relationship between what was reported for activities, goals and approaches and what the curriculum centre considers to be desirable (as in their teacher guides) should be examined. It should also be borne in mind how long it takes for 'new' methods to permeate from teacher pre- and in- service training down into normal classroom practice. Then the guides and in-service program content could, if required, be revised. Policy suggestion 4.7: It is suggested that a small R&D team conduct a detailed study of new teaching methods that will result in appropriate methods being recommended. The team could use the results just reported above as a starting point, but then review the international literatures on the matter, devise the new methods and experiment with them until the best new and appropriate methods are known. 135 Vietnam Reading and Mathematics Assessment Study (d) Giving written tests Finally teachers were asked about how often they gave written tests in both reading and mathematics to their pupils. The answers have been recorded in Figure 4.6. Figure 4.6: Frequency of giving tests in Vietnamese and mathematics Most teachers gave tests It can be seen that pupils were tested more frequently in math than in reading at regular intervals. in Vietnamese. In reading there were 4.5 percent of pupils who were either never tested or only once per year or once per term. Although nothing is written in the guidebook about applying the new curriculum the teachers seemed to test well. There were very few differences among regions as can be seen from Figure 4.7. Figure 4.7: Frequency of testing in reading Vietnamese by region 136 Vietnam Reading and Mathematics Assessment Study The same was true of testing in mathematics in the various regions. There were very few differences in the frequency of testing between isolated areas, rural areas and urban areas as illustrated by Figure 4.8 below. Figure 4.8: Frequency of giving math tests in schools in isolated, rural and urban areas. There is nothing much to be said about the frequency of giving tests for feedback to teachers. No information was available about the quality of the tests given and about how the results were used. Relationships of teacher variables with pupil achievement The results so far presented in the chapter have described teacher characteristics and what teachers do. However, what is the relationship of these characteristics and activities and perceptions with pupil achievement in reading and mathematics. Table 4.38 has presented the simple correlations (r)and partial correlations (rp) with most of the variables presented in this chapter and pupil achievement. The first criterion reported is pupil reading and the second is pupil mathematics. The correlations have been calculated at the pupil level. That is the teacher variables were disaggregated to the pupil level and attached to the pupils in the sample whom they taught. 137 Vietnam Reading and Mathematics Assessment Study Table 4.38: Correlations, partial correlations of teacher variables and pupil achievement. Reading Math Teacher variable r rp r rp Age of teacher 0.10 0.05 0.09 0.05 Sex of teacher 0.16 0.13 0.17 0.14 Ethnic group of teacher -0.12 0.01 -0.11 0.01 Teacher from local area 0.00 -0.03 -0.02 -0.05 Academic education 0.08 0.03 0.08 0.04 Professional education 0.11 0.07 0.13 0.09 Teacher subject knowledge: reading 0.23 0.19 0.24 0.21 Teacher subject knowledge: math 0.24 0.19 0.29 0.25 Years teaching experience 0.09 0.04 0.09 0.04 Married 0.07 0.05 0.07 0.05 No.of children 0.04 0.02 0.04 0.02 Classified as 'excellent' teacher 0.14 0.08 0.18 0.13 Perception of INSET 0.04 0.02 0.04 0.03 Bookcase books 0.08 0.06 0.09 0.07 Places variable -0.03 -0.02 -0.05 -0.04 Desks moveable -0.01 0.00 0.00 0.01 Classroom furniture 0.23 0.14 0.23 0.14 Classroom supplies 0.16 0.07 0.16 0.08 Classroom resources 0.24 0.14 0.24 0.15 Hours teaching per week 0.09 0.06 0.10 0.08 Hours preparing and marking -0.01 0.00 0.00 0.01 Frequency meeting parents 0.05 0.04 0.05 0.04 Percentage meeting parents 0.15 0.13 0.19 0.18 Hours earning extra money 0.01 0.00 -0.01 -0.02 Pupils' materials items 0.18 0.09 0.16 0.08 Vietnams books possessed by pupils 0.15 0.06 0.16 0.08 Math books possessed by pupils 0.13 0.05 0.14 0.07 Other books possessed by pupils 0.12 0.03 0.13 0.05 All books possessed by pupils 0.16 0.05 0.17 0.08 Sharing textbook Vietnamese 0.07 0.05 0.06 0.04 Sharing textbook mathematic 0.05 0.03 0.05 0.03 Opportunity to learn 0.11 0.06 0.10 0.05 No.inspection visits in last 3 years (imputed variable) 0.10 0.07 0.13 0.11 Teacher possession in home 0.20 0.12 0.21 0.13 Practical work in math class -0.01 0.02 -0.01 0.02 Emphasizing confidence/satisfaction in math 0.01 0.01 -0.01 -0.01 Math for life use -0.01 0.01 -0.02 -0.01 Traditional teaching 0.02 0.00 0.02 0.00 Modern teaching 0.03 0.02 0.02 0.01 Total resources (school+class) 0.29 0.18 0.28 0.18 138 Vietnam Reading and Mathematics Assessment Study It is the partial correlations that are important. As stated earlier it is relatively easy for teachers to teach pupils from 'good' home backgrounds but much more difficult to teach children from 'poor' home backgrounds. The partial correlations indicate the relationship of teacher variables with achievement after the home background of the pupils has been taken into account. It will be recalled that only 51 percent of pupils were linked with teachers but as seen in Chapter 1 there was no difference between the pupils in the pupil- teacher linked sub-sample and the non-pupil-teacher linked sub-sample. It can be seen that the variables of teacher subject-matter knowledge in both reading and mathematics had strong correlations with pupil achievement. It will also be seen that after home background was partialled out that the correlations were still strong but less large than the simple correlations. Take, for example, the correlation of teacher subject-matter knowledge in mathematics and pupil achievement in mathematics. The simple correlation was 0.29 and the partial was 0.25. The simple correlation included the home background, i.e. the fact that pupils from better homes were in schools with teachers who had more subject matter knowledge. But when the effect of home background was removed, the correlation decreased in size to 0.25. But 0.25 is still large and therefore important. No sampling errors were reported in Table 4.38 because any correlation greater then about 0.02 would be significant with the large N. Therefore, only partial correlations greater than 0.07 have been selected as worthy of further study. These are: Sex of teacher (Pupils having female teachers had higher achievement) After taking into account Teacher training (more teacher training associated with better pupil socio-economic status of achievement) the pupils in the system, Teacher subject matter knowledge in both subjects (more teacher certain teacher and knowledge = more pupil achievement) classroom variables were identified as being Being classified as an excellent teacher (excellent teachers associated strongly related to with higher pupil achievement) pupils' achievement. Classroom resources (more furniture and supplies associated with higher pupil achievement) Percentage meeting parents (more percentage meeting parents associated with higher pupil achievement) Pupils' materials in the classroom (the more the pupils have the materials they need, the higher their achievement) Number of inspection visits in the last three years (the more pupils were in schools visited by inspectors the higher the pupils achievement - but was it that inspectors decided to visit better schools) Teacher possessions in the home (teachers with more possession associated with higher pupil performance) The relative importance of these variables will be examined in a later chapter. 139 Vietnam Reading and Mathematics Assessment Study Conclusion Several aspects of the characteristics and daily lives of the teachers of Grade 5 pupils in Vietnam have been reviewed in this chapter. There were large differences between the kinds of teachers in isolated areas and the other areas. For measures of teacher quality such as the years of teacher training they had received or the teacher subject matter achievement, those in the isolated areas were of less quality than the teachers in rural and urban areas. Female teachers tended to have pupils with better achievement than did male teachers. There were more male teachers in the isolated schools (probably for good reasons) but these teachers had a lower mastery of the subject-matter than did the female teachers. Teachers with more years of pre-service teacher training had better results with their pupils than did teachers with fewer years of pre-service teacher training. The teachers in the isolated area schools tended to have fewer years of pre-service teacher training. The allocation of teachers to schools in terms of teacher quality was not equitable. In some ways it can be argued that the better teachers should be in the remote schools in order to compensate for the poorer teachers they have had for years. It was not clear how teacher quality is taken into account when allocating teachers to schools. Thus the first major suggestion was that those responsible for the allocation of teachers to schools should review the current system and affect some improvement in it. Some easy indicators of teacher quality should be used. A second suggestion was concerned the poor levels of reading and mathematics achievement on the part of some teachers. it was suggested that the content of the current teacher training courses to be updated to take account of the new curriculum, that a national level of subject matter knowledge should be established for teachers to become certified, and that special efforts to be made through the in-service programs to up-date the existing teachers. The excellent teachers at the school and national levels had pupils who did not achieve more than the pupils who had teachers not classified as excellent teachers at these levels. On the other hand the excellent teachers at the district and provincial levels had pupils with much higher achievement levels than those pupils without excellent teachers at these levels. Teachers in general did not think that the in-service teacher training courses were very effective. A third suggestion was that the MOET should make every effort to strengthen the supply to schools of many books for their classroom and school libraries. There were adequate places for the pupils to sit and write in the classrooms. Nearly half the desks were moveable so that the teachers could organize different groupings of pupils. But, over 50 percent were not moveable. It is assumed that the MOET is moving toward a replacement of desks so as to have all moveable desks. The number of books available in classrooms was very limited. This was serious if pupils were to get the opportunity to do lots of reading. It was also the case that where pupils were in classrooms with some books in the classroom library, they were also in schools where there 140 Vietnam Reading and Mathematics Assessment Study was a school library. What was bad was the number of pupils in schools where there was neither a school or classroom library. In terms of classroom furniture there were a few pupils in classrooms where there was a useable blackboard. This should be remedied immediately. There were high percentages of pupils in classrooms without a cupboard, without a classroom bookcase and without bookshelves. Again it tended to be the isolated schools that had the least furniture. About five percent of pupils were in classrooms where the teacher did not have access to teacher guides in Vietnamese and mathematics. This seems to be odd. A fourth suggestion was that the provincial authorities (or even the regional authorities since there were large differences between regions) should consider undertaking as part of the annual census a review of the furniture and supplies in all schools in their jurisdiction. Teachers in isolated areas only taught 16.1 hours per week, those in rural and urban areas 16.8 hours per week. This is a low teaching load by international standards. This is a low teaching load by international standards. It is not known why Vietnamese primary teachers teach such low hours but it is suggested that the matter be reviewed. The teachers professed to be preparing lessons ands marking homework for a further 20 hours per week but perhaps the MOET might like to check the validity of these data. The extent to which teachers meet parents seems to be a rather rote affair. Given the importance of meeting parents and bearing in mind what was written in Chapter 3 about school intervention programs a further suggestion was that the MOET should conduct a study on how, when and why teachers meet parents and how this might be improved. The issue of the materials and textbooks that pupils possessed for the purpose of learning was examined. Whereas nearly all pupils had a copy of the major textbooks this was not true of many of the other textbooks, even though the pupils said that they were studying all of the subjects. Nearly all of the pupils had the materials but there were small percentages not having some of the required materials. Having the wherewithal to learn is so important - and of course it was the pupils in the isolated schools who had least - that it was felt that the curriculum centre should review which materials and books a Grade 5 pupil must have and then make sure that all pupils have what they need to be able to study. The teachers' perceptions of the role of inspectors and educational advisors were examined as well as the teachers' sources of satisfaction with their work. Finally, the teaching methods used by the math teachers were examined. It was felt that a special study group should review the findings on teaching methods (activities, goals and approaches) and on the basis of their deliberations and a review of the international literature revise the teacher guidebooks. A priority needs to be set on the different suggestions and a plan made as to where the MOET must act alone and where the donor agencies might help. This matter will be raised again in the last chapter. 141 Vietnam Reading and Mathematics Assessment Study Appendix 1.4: Estimating actual number of pupils needing textbooks and costs1 It is possible to estimate the actual number of persons having a certain characteristic or, in this case, needing a textbook by using Raising Factor 3 on the data file. The total population of Grade 5 pupils for 2001 is known and the RF3 is a weight by which to multiply an existing figure in the sample to know what it would constitute for the whole target population. By using the IIEPJACK program it is also possible to calculate the standard error of sampling at the same time. In the table below, the first column is the estimated number of pupils minus two standard errors. Thus the estimated figure was 18938 pupils needing a textbook in Vietnamese 5 Vol. 1. One standard error of sampling was 1137. Thus two standard errors of sampling constituted 2274 pupils. Using two standard errors allows us to be 95 percent confident that the actual population figure lies between two standard errors. The estimate of 18938 minus 2274 was 16664 and the estimate plus 2274 was 21212. We can be 95 percent confident that the 'true' population figure lay between 16664 and 21212. The MOET supplied the cost of each book. This was the actual cost of the book plus the average cost of sending it to schools. Multiplying this cost by each of the population estimates produces the minimum, average, and maximum costs for supplying any one book to those pupils saying that they did not have one. Appendix 4.1: Target population estimates of pupils needing textbooks, and costs. Items Number of pupils who do not possess Cost (dong) minus 2SE mean plus 2SE unit cost transport fee minimum cost average cost maximum cost Vietnamese 5 Vol.1 16664 18938 21212 5100 690 96484560 109651020 122817480 Vietnamese 5 Vol.2 12404 13942 15480 4700 690 66857560 75147380 83437200 Vietnamese story book 425618 439732 453846 4200 690 2081272020 2150289480 2219306940 Vietnamese exercise book 5 Vol.1 915577 937143 958709 4500 690 4751844630 4863772170 4975699710 Vietnamese exercise book 5 Vol.2 1021087 1042363 1063639 4900 690 5707876330 5826809170 5945742010 Vietnamese other books 1160890 1180362 1199834 16000 690 19375254100 19700241780 20025229460 Vietnamese dictionary 1756207 1774875 1793543 21000 690 38092129830 38497038750 38901947670 Vietnamese Essay Grade 5 481491 495465 509439 6000 690 3221174790 3314660850 3408146910 Math 5 9813 11395 12977 5400 690 59761170 69395550 79029930 Math Exercise book 789011 804645 820279 5500 690 4883978090 4980752550 5077527010 Math Exercise book 5 Vol.1 881499 900709 919919 3800 690 3957930510 4044183410 4130436310 Math Exercise book 5 Vol.2 990901 1009759 1028617 3800 690 4449145490 4533817910 4618490330 Math other books 1300261 1315785 1331309 24500 690 32753574590 33144624150 33535673710 Moral Education 5 188275 195951 203627 1500 690 412322250 429132690 445943130 Moral Ed.Exercise book 1350821 1368259 1385697 4500 690 7010760990 7101264210 7191767430 Nature and Society 5 Science 107552 113056 118560 5600 690 676502080 711122240 745742400 Nature and Society 5 History and Geography 211322 217824 224326 5900 690 1392611980 1435460160 1478308340 N&S 5:Exercise book-Science 1366727 1388081 1409435 3300 690 5453240730 5538443190 5623645650 N&S 5:Exercise book-History 1391040 1412494 1433948 2100 690 3881001600 3940858260 4000714920 N&S 5:Exercise book-Geography 1376882 1397788 1418694 2300 690 4116877180 4179386120 4241895060 Technology 5 290609 301519 312429 1800 690 723616410 750782310 777948210 Music 5 412005 425835 439665 4700 690 2220706950 2295250650 2369794350 Music exercise 5:Vol1 1569684 1588412 1607140 2700 690 5321228760 5384716680 5448204600 Music exercise 5:Vol2 1537005 1554005 1571005 2500 690 4903045950 4957275950 5011505950 Art 5 242482 253068 263654 2900 690 870510380 908514120 946517860 Health 5 222830 231512 240194 4000 690 1045072700 1085791280 1126509860 Heath 5 Exercise book 1581491 1600241 1618991 3200 690 6151999990 6224937490 6297874990 Informatics for Primary Education 1855913 1874987 1894061 4200 690 9075414570 9168686430 9261958290 English for Primary Education 1450288 1477648 1505008 5000 690 8252138720 8407817120 8563495520 Total million US$ 12.1 12.3 12.4 1This appendix was written by Christopher P. Shaw of the World Bank in Vietnam 142 Vietnam Reading and Mathematics Assessment Study Chapter 5 WHAT WERE THE CHARACTERISTICS OF SCHOOL HEADS AND SCHOOLS? Introduction S chools are different within every country in the world. Within Vietnam it is the same. Griffin1 showed that there was tremendous variation among schools in the five provinces in which the NIES collected data in 1999. In the present study the sample of schools and Grade 5 pupils covered the whole of Vietnam. The major thrust of this chapter is to examine the characteristics of school heads and of the schools themselves in order to establish the levels of inputs to schools but also the amount of variation among schools in isolated, rural, and urban areas in different regions. These data, as in the previous two chapters, help establish a baseline against which schools can be compared in future studies in order to see if the situation has improved, remained the same or deteriorated. At the same time these are the data that will be analysed more extensively in later chapters. And again the percentages presented are the percentages of pupils in schools with different characteristics rather then the simple count of schools. A mean indicates that the average pupil in Vietnam is in a school with the characteristic in question. ------------------------------- 1 Griffin, Patrick (1999) Vietnam Primary School Monitoring Report. Mimeographed report. World Bank in Vietnam 143 Vietnam Reading and Mathematics Assessment Study What were the general characteristics, training and experience of school heads? (a) Gender of School Head It will be recalled that the percentage of pupils whose teachers were female Whereas most teachers was 73 percent. The percentage of pupils having school heads who were were female 65% of female have been presented in Table 5.1. It can be seen that on average 35.2 pupils were in schools percent of pupils had heads who were female. In isolated areas the percentage where the school heads was 22.9, in rural schools it was 32.3, and in urban areas it was 51.0. There were male. But in Hanoi was also some variation among regions. In the Central Coast region only 12 province 71% were percent of pupils had female heads. At the province level, only three percent female heads. of pupils had heads who were female in Binh Dinh and Phu Yen whilst in Hanoi the percentage was 71. (b) Age of school heads The mean age of the average Grade 5 pupil's school head in Vietnam was 45 The average age of head years. Pupils in isolated schools tended to have younger school heads than was 45 years, but in pupils in other areas. The reason for having younger heads in the isolated isolated areas they were schools was probably that those being Heads in such areas required fewer younger. years of teacher training. It was interesting to note that younger heads were heads of more recently built schools. (c) Time taken to travel to school The average Grade 5 pupil had a head who travelled 13.5 minutes to reach The school head's school each day. In isolated schools the average travel time was 18 minutes. average travel time to In isolated schools in the Northeast and Northwest regions the average travel school was about 15 time was 22 and 28 minutes respectively. Only half a percent of pupils had minutes. teachers who said that they travelled an hour to go to work. There would not appear to be a problem with travel time to school. Table 5.1: Means, percentages and sampling errors of selected characteristics of School Heads Minutes travel to Marital status % % non-majority School Sex (% female) Age in years No. of children Region school yes ethnic group location % SE Mean SE Mean SE % SE Mean SE % SE Isolated 29.5 30.03 52.4 3.92 10.6 1.97 100.0 0.00 2.83 0.57 . . Rural Red River 44.3 2.38 49.1 0.33 12.1 0.35 98.9 0.47 2.95 0.06 0.1 0.13 Delta Urban 72.7 4.70 49.1 0.55 12.8 0.62 96.9 1.78 2.26 0.10 1.3 1.26 Total 49.0 2.26 49.2 0.28 12.2 0.28 98.6 0.51 2.83 0.05 0.3 0.24 144 Vietnam Reading and Mathematics Assessment Study Isolated 38.3 4.06 40.4 0.75 21.5 1.87 95.1 2.08 2.47 0.12 45.5 4.86 Rural 43.8 2.56 45.6 0.30 13.8 0.42 99.0 0.64 2.75 0.06 18.1 1.30 Northeast Urban 69.1 4.84 47.1 0.51 11.6 1.23 96.6 1.70 2.44 0.12 9.3 1.96 Total 47.0 2.17 45.0 0.28 14.7 0.47 97.9 0.61 2.65 0.05 21.1 1.18 Isolated 26.6 6.14 41.1 1.10 27.8 3.41 95.1 3.40 2.98 0.24 50.6 7.86 Rural 31.9 4.99 44.6 0.83 15.7 1.17 100.0 0.00 2.91 0.16 47.5 5.72 Northwest Urban 73.3 9.00 44.9 0.83 9.7 1.52 98.4 1.66 2.44 0.14 . . Total 36.2 3.30 43.4 0.64 19.0 1.50 98.0 1.23 2.86 0.12 41.5 4.60 Isolated 7.9 6.16 40.7 1.55 17.6 2.96 98.8 0.86 2.48 0.27 18.9 8.52 Rural North 33.0 4.06 47.2 0.57 13.8 0.78 98.0 0.99 2.90 0.10 1.8 1.28 Central Urban 45.6 8.25 46.3 1.22 10.2 0.85 94.9 4.22 2.19 0.15 . . Total 32.4 3.40 46.4 0.53 13.6 0.66 97.6 0.94 2.76 0.09 3.2 1.27 Isolated 9.3 4.57 41.7 0.80 15.4 2.15 100.0 0.00 2.45 0.15 6.9 5.12 Rural Central 6.8 1.72 44.7 0.36 13.0 0.55 96.5 1.21 2.73 0.09 . . Coast Urban 27.3 4.06 46.4 0.45 10.6 0.88 97.3 1.35 2.38 0.14 . . Total 12.2 1.47 44.8 0.28 12.7 0.50 97.1 0.79 2.61 0.07 0.8 0.61 Isolated 19.6 8.30 39.1 1.50 17.8 3.07 97.1 1.54 2.51 0.20 17.4 7.04 Rural Central 29.7 5.86 41.5 0.58 12.4 1.12 96.9 2.16 3.00 0.17 11.3 4.02 Highlands Urban 29.0 8.40 44.6 0.90 8.5 0.65 100.0 0.00 2.90 0.20 4.2 3.59 Total 27.4 3.92 41.8 0.49 12.5 0.76 97.8 1.14 2.88 0.10 10.7 2.62 Isolated 37.5 6.98 38.6 0.74 13.1 1.60 92.1 6.04 2.07 0.17 4.2 1.97 Rural 38.1 3.10 42.8 0.37 11.6 0.52 87.6 2.43 2.24 0.09 3.8 1.19 Southeast Urban 52.3 4.53 45.9 0.63 12.9 0.65 88.8 2.95 2.03 0.08 0.6 0.45 Total 43.9 2.64 43.6 0.36 12.3 0.42 88.7 1.84 2.13 0.06 2.5 0.61 Isolated 12.9 3.86 40.2 0.70 14.8 1.26 96.4 1.60 2.23 0.13 2.9 1.71 Rural Mekong 17.6 1.90 42.3 0.29 15.3 0.62 91.5 1.35 2.20 0.07 3.3 0.87 Delta Urban 40.0 4.73 44.9 0.69 11.1 0.65 81.9 3.87 1.91 0.16 3.0 1.65 Total 20.7 1.56 42.4 0.30 14.5 0.44 90.6 1.07 2.16 0.06 3.2 0.71 Isolated 22.9 1.98 40.4 0.36 17.6 0.82 96.1 1.07 2.40 0.05 19.3 1.87 Rural 32.3 1.08 45.5 0.15 13.4 0.23 95.9 0.45 2.67 0.04 5.7 0.41 Vietnam Urban 51.0 2.31 46.4 0.29 11.6 0.32 92.2 1.28 2.21 0.04 2.2 0.44 Total 35.2 1.00 45.0 0.12 13.5 0.18 95.1 0.40 2.54 0.03 6.6 0.37 . = no data (d) Marital status and number of children In Table 5.1 it is the percentage of Grade 5 pupils who had school heads who Nearly all were married were married that has been reported. It can be seen that 96 percent of pupils and... were in schools where the school heads were married. This average masks the fact that 99 percent of pupils had male heads and only 89 percent of pupils had female heads who were married. In urban areas only 85 percent of pupils had female heads who were married. For reasons that are unknown in the provinces of Dong Nai and Vinh Long 15 percent of pupils had heads who were unmarried. These statistics have been given because when the policy issues were discussed in the Ministry of Education at the very onset of the study many were asking for information about the number of unmarried female teachers and heads in remote and isolated areas. Apparently the Ministry was disturbed that too many unmarried females were in isolated areas and hence not being 145 Vietnam Reading and Mathematics Assessment Study able to meet appropriate males for marrying. It is suggested that the fears of the MOET were not realised ­ at least for Grade 5 Heads. ...had 2.5 children. On average the Heads had 2.5 children. In general there was not much variation among school location areas and among regions. But the highest number of children (3) was in the rural Central Highlands. In the provinces of Vinh Phuc, Ha Tay, Ha Nam, Hoa Binh and Quang Binh the Heads had more than three children. It should be noted that the government rule about not having more than 2 children was not strictly enforced until after 1980. It is the older heads who tended to have more than two children. The correlation between school heads' age and the number of children he or she had was 0.50. Table 5.2: Means, percentages and sampling errors of Heads' education and teaching Attended Hours per Years academic Years teacher Years Teaching Years head Years head School management week teaching Region education training experience this school altogether location course now Mean SE Mean SE % SE Mean SE Mean SE Mean SE Mean SE Isolated 10.5 1.58 2.2 0.62 42 40.3 32.2 3.44 2.38 0.88 11.4 3.73 14.7 6.07 Red River Rural 12.8 0.11 2.5 0.04 90 1.5 28.5 0.35 1.89 0.06 10.6 0.27 13.7 0.33 Delta Urban 13.6 0.24 3.0 0.08 96 2.3 28.9 0.60 0.83 0.12 9.1 0.57 14.2 0.86 Total 13.0 0.10 2.6 0.04 91 1.3 28.6 0.30 1.72 0.05 10.3 0.25 13.8 0.33 Isolated 11.7 0.21 2.2 0.08 65 4.2 19.5 0.79 2.58 0.16 7.2 0.49 9.4 0.52 Rural 11.8 0.09 2.4 0.04 83 1.5 24.9 0.32 2.60 0.06 8.2 0.24 12.3 0.35 Northeast Urban 12.4 0.22 2.7 0.08 88 3.3 26.8 0.59 1.85 0.14 8.1 0.52 13.6 0.84 Total 11.9 0.07 2.4 0.03 81 1.4 24.3 0.31 2.47 0.05 8.0 0.20 12.0 0.28 Isolated 11.1 0.29 2.0 0.11 67 6.3 19.6 1.24 2.43 0.26 8.2 0.82 10.9 0.99 Rural 11.5 0.24 2.1 0.08 77 4.7 24.5 0.90 2.91 0.20 8.3 0.55 12.8 0.75 Northwest Urban 12.8 0.51 2.5 0.18 82 7.2 24.9 1.00 2.04 0.32 6.6 0.80 10.7 1.41 Total 11.6 0.18 2.1 0.06 75 3 22.8 0.72 2.61 0.14 8.0 0.41 11.8 0.62 Isolated 12.0 0.32 2.2 0.07 79 8.5 18.7 1.72 2.29 0.37 6.3 1.03 11.0 1.56 Rural 12.4 0.14 2.7 0.06 89 2.4 25.9 0.64 1.79 0.12 7.9 0.44 12.7 0.49 North Central Urban 13.6 0.42 3.1 0.14 92 2.7 25.9 1.23 1.91 0.58 6.7 0.74 13.6 1.38 Total 12.5 0.13 2.7 0.05 89 2.1 25.2 0.58 1.86 0.14 7.6 0.35 12.6 0.44 Isolated 12.5 0.27 2.5 0.13 81 6.5 21.6 0.84 1.35 0.24 9.1 0.86 11.8 1.00 Rural 13.1 0.13 2.7 0.06 83 2.6 24.0 0.28 2.01 0.10 8.9 0.42 14.2 0.50 Central Coast Urban 14.0 0.21 3.1 0.10 90 3.6 25.8 0.40 1.35 0.14 7.7 0.66 14.2 0.81 Total 13.3 0.10 2.8 0.05 85 2 24.1 0.25 1.77 0.09 8.6 0.33 13.9 0.41 Isolated 12.5 0.25 2.2 0.18 69 9.3 18.1 1.61 1.49 0.31 6.9 1.02 10.6 1.20 Central Rural 12.8 0.26 2.3 0.15 83 4.8 20.2 0.58 2.01 0.19 7.7 0.61 10.4 0.64 Highlands Urban 13.9 0.37 2.9 0.17 84 6.5 23.4 0.66 1.26 0.21 7.1 0.71 12.9 1.16 Total 13.0 0.16 2.4 0.09 80 3.8 20.6 0.49 1.69 0.14 7.4 0.40 11.1 0.53 Isolated 12.1 0.27 2.1 0.12 65 5.6 17.2 0.70 1.55 0.28 7.3 0.67 9.5 0.79 Rural 12.9 0.14 2.4 0.08 74 2.7 21.2 0.42 0.85 0.07 8.9 0.33 11.9 0.37 Southeast Urban 13.4 0.20 2.6 0.10 85 2.9 24.9 0.55 0.73 0.13 7.8 0.53 13.2 0.65 Total 13.0 0.10 2.4 0.06 77 1.8 22.2 0.35 0.89 0.07 8.2 0.32 12.1 0.37 Isolated 12.4 0.17 2.2 0.10 65 4.6 19.9 0.61 0.89 0.17 8.8 0.39 11.0 0.57 Rural 12.9 0.08 2.3 0.05 77 1.8 21.4 0.26 0.83 0.08 8.8 0.28 11.8 0.34 Mekong Delta Urban 13.1 0.20 2.6 0.08 81 3.5 24.1 0.68 0.71 0.11 7.4 0.46 12.4 0.70 Total 12.8 0.06 2.4 0.04 76 1.6 21.6 0.27 0.82 0.06 8.5 0.20 11.8 0.26 Isolated 12.0 0.10 2.2 0.04 68 2.3 19.4 0.35 1.75 0.08 7.8 0.26 10.5 0.35 Rural 12.6 0.04 2.5 0.02 83 0.7 24.5 0.16 1.70 0.04 8.9 0.14 12.6 0.17 Vietnam Urban 13.4 0.09 2.8 0.04 87 1.2 25.7 0.27 1.12 0.08 7.8 0.24 13.3 0.34 Total 12.7 0.04 2.5 0.02 82 0.6 24.2 0.13 1.58 0.03 8.5 0.10 12.5 0.15 146 Vietnam Reading and Mathematics Assessment Study (e) Ethnic group affiliation of School Heads Some 6.6 percent of pupils were in schools with ethnic minority heads. Nearly 7 % of pupils When the ethnic minority affiliation in the different school location areas were in schools with was examined, it was found that it was 2.2 percent of pupils in urban areas, ethnic minority heads. In 5.7 percent of pupils in rural areas, and 19.3 percent of pupils in remote and some provinces this was isolated areas. In isolated schools in the Northeast and Northwest regions over 50% of pupils. 46 and 51 percent of pupils respectively had ethnic minority heads. In Cao Bang 94 percent of pupils had heads who were non-Kinh. In Bac Kan, Lang Son, and Ha Giang more than 50 percent of pupils had Heads who were non-Kinh. (f) Academic qualifications As with teachers, heads were asked about their academic education. This was The average years of the number of years of primary and secondary schooling they had received. academic education of It can be seen from Table 5.2 that the mean for Vietnam was 12.7 years school heads was 12.7 meaning that the average pupil in Grade 5 had a school head who had years... received this amount of education. This varied among school location areas. It was 12.0 in isolated schools, 12.6 in rural areas and 13.4 years in urban areas. There were some provinces where the number of years of the School Heads' education could be considered to be somewhat low. These have been listed below. Level of academic education Province Mean SE Lao Cai 10.67 0.25 Lai Chau 10.69 0.21 Ha Giang 10.93 0.24 Bac Kan 10.97 0.24 Tuyen Quang 10.99 0.21 Yen Bai 11.05 0.22 Lang Son 11.3 0.22 Son La 11.53 0.28 Binh Phuoc 11.58 0.20 Bac Ninh 11.87 0.26 Hung Yen 11.9 0.27 Vinh Phuc 11.94 0.21 (g) School head's teacher professional training School Heads were asked about their pre-service teacher training. The question they were asked was: Which kind of teacher training did you receive? (Please tick only one box.) 147 Vietnam Reading and Mathematics Assessment Study (1) I did not receive any teacher training. (0) (2) Less than secondary teacher training (1) (3) Secondary teacher training (9+3) (1.5) (4) Secondary teacher training (12+2) (2) (5) College (12+3) (3) (6) University (12+4) (4) ...and had had 2.5 years The types of training were converted into the number of years they of teacher training. represented in teacher training. The numbers of years have been given after each type of training. The national average was 2.5 years. The `rule' is that heads in isolated areas should have 1.5 years of teacher training and in rural and urban areas it should be two years. In general those heads in urban areas had more years of teacher training than those in rural areas who in turn had more than those in isolated areas. There was some variation between provinces and in Tuyen Quang, Lai Cai, Lai Chau, and Binh Phuc the figure was less than two years. (h) Specialised training in school management 80% of pupils were in The school heads were asked if they had received specialised training in schools where heads had school management at the Ministry or Provincial level. It can be seen from received training in Table 5.2 that, for Vietnam as a whole, 80 percent of pupils were in schools school management. where the school head had received a school management course. However, there was some variation among school location areas. It can be see, for example, that only 68 percent of pupils in isolated areas had school heads who had attended a management course. In isolated areas in the Red River Delta and the Southeast regions fewer than 50 percent of pupils had heads who had attended such a course. In the provinces of Ninh Thuan, Lai Chau. Binh Phoc, and Binh Thuan more than half of the pupils had heads who had not attended such courses. (i) Years teaching experience The average Grade 5 The average Grade 5 pupil had a head teacher who had had 24 years of pupils had school head teaching experience altogether including the years that he or she had been a with 24 years of school head. But from Table 5.2, it can also be seen that there was some teaching experience. variation. Pupils in urban areas tended to have heads who had had more years of teaching experience. In the isolated schools in the Southeast the heads had only 17 years of teaching experience. It is sometimes suggested that the more teaching experience a school has had in teaching the better a school head he/she will be and that the pupils will learn more. The association between the school heads' years of teaching experience and pupil achievement was in the order of about 0.15 (and 0.08 after controlling for home background). 148 Vietnam Reading and Mathematics Assessment Study (j) Head's hours per week teaching now School heads were asked how many lessons or periods a week they taught in Most heads taught for a typical school week and how long each lesson was on average. From these 1.6 hours per week in two questions it was possible to derive how many hours they taught in a their schools, but in typical week. It is often suggested that a good school head teaches each grade urban areas this was for a short time each week and that in this way he or she learns of the only 1.1 hours. problems in the school rather than waiting for them to be reported by teachers. The Vietnamese government has laid down that a school head should teach 1.5 hours per week. From Table 5.2, it can be seen that the average pupil had a head who taught for 1.6 hours but this was only 1.1 hours in urban areas and 1.7 hours elsewhere. Heads in the Mekong Delta were below par on this variable. There were 26 provinces where the number of hours did not reach the government norm. Hours per week teaching now Province Mean SE Tra Vinh 0.06 0.05 Dong Thap 0.27 0.09 Bac Lieu 0.29 0.11 Ho Chi Minh 0.38 0.16 Hai Phong 0.40 0.13 Ha Noi 0.43 0.13 Binh Duong 0.44 0.14 Ba Ria - Vung Tau 0.49 0.14 Tien Giang 0.52 0.12 Ca Mau 0.52 0.14 Dong Nai 0.58 0.18 Long An 0.66 0.15 An Giang 0.77 0.41 Thua Thien - Huu 0.95 0.20 Kien Giang 1.02 0.19 Da Nang 1.08 0.16 Lam Dong 1.12 0.19 Kon Tum 1.13 0.20 Vinh Long 1.13 0.23 Thai Nguyen 1.22 0.17 Can Tho 1.30 0.19 Binh Phuoc 1.31 0.20 Ha Nam 1.35 0.06 Soc Trang 1.42 0.22 Binh Thuan 1.47 0.31 Binh Dinh 1.48 0.21 It is to be expected that heads in urban areas are typically in charge of larger Not all school heads schools and with more administrative work and perhaps they do not have taught classes. time to teach. No information was available as to whether in these large schools there was a deputy head who might have been the contact with the pupils. The heads in isolated areas have often by necessity to teach more. 149 Vietnam Reading and Mathematics Assessment Study Nevertheless, there would appear to be a problem with the heads teaching and the MOET should examine this problem more. (k) Years experience as a head The average years as a The school heads provided information on how many years they had been school head was 12, but head of the current school and how many years they had been a head this was less in isolated altogether. Again from Table 5.2 it can be seen that the average was 8.5 years areas. for being head of this school and that there was no difference by school location area. The national average for being a head altogether was 12 years. As might be expected the heads in isolated areas had less experience. Of all of the characteristics reviewed thus far, it would seem that many things are in order as far as School Heads are concerned. However, there were some inequities as far as the number of years of general education and teacher training that the Heads of different schools had received. This was also true for the numbers having attended management courses (see Figure 5.1 below). In particular there were several provinces where the heads were not teaching even the minimum required of 1.5 hours per week. In general, school heads in urban areas had higher levels of Figure 5.1: School head education, teacher training, and management academic education and course, by school location teacher training and had also attended management courses more than heads in rural and isolated areas.. 150 Vietnam Reading and Mathematics Assessment Study Policy suggestion 5.1: The MOET may wish to review the data on the distribution of school heads' characteristics across schools and, together with the relevant provincial directors of education, devise methods whereby there will be a more equitable allocation of heads to schools in terms of their education, teacher training and management training. Every effort should also be made to have the heads teaching at least two or three periods a week. What were the general living conditions of school heads? The Ministry of Education and Training was interested to discover more about the living conditions of school heads. It was considered that if the heads had good living conditions they would not be worrying about them and could devote more energy and time to being a school head. Questions were asked about where they lived, the possessions they had at home, and the extent to which they were satisfied with their living accommodation. (a) Where school heads lived School heads were asked if they lived with their parents, in leased Heads tended to live in accommodation, in government accommodation or in their own homes. From their own homes.... Table 5.3, it can be seen that four-fifths of pupils had school heads who were living in their own homes. About 12 percent lived with their parents and five percent were in government accommodation. In the Central Highlands the percentage of pupils with heads living with their parents was quite low. More heads in isolated areas were in government housing. Those living in their own homes tended to be the most satisfied and those living leased accommodation were the least satisfied. Less than one percent of pupils had heads who lived in leased accommodation. (b) `Wealth' of school heads homes The `wealth' of the homes was measured by the number of possessions they ...where they had 12 out had in their homes. In all, 20 possible possessions were listed and the heads of 20 possessions items.. had to say whether or not they had then in their homes. The possessions were: daily newspaper, weekly or monthly magazine, radio, TV set, VCR, cassette player, telephone, refrigerator/freezer, motorcycle, bicycle, piped water, electricity, table to write on, bookcase, watch, lamp to read books, wardrobe, cooker, washing machine, and computer. The average Grade 5 pupil had a Head who had nearly 12 items number of these 20 possessions. In urban areas this was 14.6 items, in rural areas it was 11.6 items and in isolated areas it was 10.1 items. The correlation between total possessions owned and satisfaction with living conditions was 0.24, but as can be seen from Figure 151 Vietnam Reading and Mathematics Assessment Study 5.2 those Heads in leased accommodation were least satisfied. It would seem that the type of accommodation that heads have is more important than the items possessed in determining satisfaction with the living conditions. In general, there was slightly more dispersion of wealth among school heads in urban areas than in rural and isolated areas. Figure 5.2: Teacher satisfaction with living conditions by possessions and type of accommodation (c) Satisfaction with living accommodation ...and were satisfied with If school heads are to be devoting their energy to running the school and their living conditions. teaching they should not have to be worried about their living accommodation. They were asked to what extent they were satisfied with their living accommodation. The scale was 1 = not very satisfied, 2 = satisfied, and 3 = very satisfied. It can be seen from Table 5.3 that the last two categories were combined to make a `satisfied'category. Eighty-eight percent of pupils had heads who were satisfied with their living accommodation. It should be noted that at the provincial level, less than 70 percent of pupils had Heads who were satisfied in Cao Bang, Ca Mau, and Hue. Cao Bang and Ca Mau also had the least possessions. On the other hand Hanoi and Hue had high possessions but were least satisfied. Again, this may well be the problem of leased accommodation. 152 Vietnam Reading and Mathematics Assessment Study Table 5.3: Means, percentages and sampling errors of living conditions of school heads Where living Satisfaction with Possession (20 items) Region School location With parent Leased Govt. Own home accommodation % SE % SE % SE % SE Mean SD SE % SE Isolated . . . . 27.5 34.83 72.5 34.83 10.3 1,77 1.43 89.7 1.43 Rural 11.5 1.58 . . 2.9 0.83 85.7 1.84 11.8 2,65 0.13 88.2 0.13 Red River Delta Urban 8.7 3.02 8.4 3.03 7.8 3.20 75.1 5.36 15.6 2,54 0.27 84.4 0.27 Total 10.9 1.50 1.5 0.53 4.0 0.85 83.7 1.70 12.5 2,99 0.12 87.6 0.12 Isolated 7.5 2.76 0.6 0.47 12.9 2.96 79.0 3.94 9.4 2,97 0.29 90.6 0.29 Rural 9.0 1.27 . . 3.0 0.73 88.0 1.51 11.3 2,81 0.14 88.7 0.14 Northeast Urban 3.1 1.66 . . 0.6 0.46 96.3 1.70 13.8 3,23 0.31 86.2 0.31 Total 7.8 1.06 0.1 0.08 4.2 0.75 87.9 1.29 11.4 3,17 0.12 88.6 0.12 Isolated 12.2 4.64 . . 12.9 4.15 74.9 5.80 8.8 3,67 0.58 91.3 0.58 Rural 12.6 3.39 1.3 1.28 1.8 1.27 84.4 3.66 10.5 2,79 0.33 89.6 0.33 Northwest Urban 1.6 1.66 . . . . 98.4 1.66 13.0 3,03 0.60 87.0 0.60 Total 10.8 2.60 0.6 0.64 5.5 1.58 83.1 2.90 10.2 3,45 0.30 89.8 0.30 Isolated 11.5 4.85 1.9 1.95 19.7 7.75 66.9 9.88 10.5 2,52 0.50 89.5 0.50 Rural 8.3 2.65 0.4 0.28 5.0 1.66 86.3 3.07 12.2 2,43 0.20 87.8 0.20 North Central Urban 9.7 4.74 . . 1.2 1.22 89.2 4.87 14.6 2,68 0.45 85.4 0.45 Total 8.8 2.31 0.5 0.28 5.9 1.43 84.8 2.68 12.4 2,68 0.19 87.6 0.19 Isolated 11.5 5.85 . . 11.6 5.27 76.9 6.95 9.9 3,23 0.49 90.1 0.49 Rural 20.8 3.06 . . 0.9 0.71 78.3 3.16 11.3 2,90 0.19 88.7 0.19 Central Coast Urban 11.1 2.97 1.8 1.77 2.0 1.26 85.2 3.62 14.3 3,49 0.43 85.8 0.43 Total 17.3 2.18 0.4 0.44 2.5 0.83 79.8 2.31 11.9 3,42 0.20 88.1 0.20 Isolated 1.9 1.40 . . 15.1 5.87 83.0 5.96 9.3 2,84 0.48 90.7 0.48 Central Rural 3.2 2.24 0.4 0.35 0.9 0.89 95.6 2.38 11.4 2,88 0.37 88.6 0.37 Highlands Urban . . . . 3.7 3.70 96.4 3.70 13.4 2,25 0.37 86.6 0.37 Total 2.1 1.18 0.2 0.18 4.7 1.63 93.1 1.98 11.5 3,06 0.27 88.5 0.27 Isolated 17.0 6.99 2.3 2.27 6.6 2.73 74.2 7.51 10.9 3,41 0.54 89.1 0.54 Rural 17.3 2.59 0.8 0.46 4.8 1.43 77.2 2.78 12.4 3,13 0.21 87.6 0.21 Southeast Urban 18.0 3.39 4.4 2.20 5.7 1.91 71.9 4.23 15.5 3,19 0.26 84.5 0.26 Total 17.5 2.02 2.5 0.96 5.4 1.15 74.6 2.43 13.5 3,64 0.17 86.5 0.17 Isolated 10.3 2.85 . . 10.1 3.09 79.6 4.20 10.6 3,44 0.24 89.4 0.24 Rural 17.5 1.71 0.7 0.38 4.6 0.88 77.2 1.99 11.0 3,28 0.16 89.0 0.16 Mekong Delta Urban 20.9 3.80 1.5 1.11 8.3 2.62 69.3 4.66 13.7 3,23 0.29 86.3 0.29 Total 17.0 1.42 0.7 0.31 6.1 0.94 76.2 1.74 11.4 3,46 0.14 88.6 0.14 Isolated 10.2 1.58 0.7 0.42 12.4 1.62 76.7 2.13 10.1 3,25 0.15 89.9 0.15 Rural 12.8 0.85 0.3 0.12 3.5 0.43 83.3 0.96 11.6 2,90 0.07 88.4 0.07 Vietnam Urban 12.3 1.46 3.0 0.80 4.8 0.91 80.0 1.88 14.6 3,16 0.12 85.4 0.12 Total 12.4 0.77 0.9 0.18 4.9 0.40 81.8 0.87 12.1 3,33 0.06 87.9 0.06 . = no data What were the school types and the contexts in which they were located? (a) Type of school 153 Vietnam Reading and Mathematics Assessment Study Three types of school exist in Vietnam. There is the ordinary public school and then there are semi-public and people-founded schools. Nearly 100 percent of pupils were in ordinary public schools and only very few were in the other two kinds of schools. No pupils in isolated areas were in semi-public or people- founded and indeed the pupils in such schools tended to be in urban areas. Table 5.4: Means, percentages and sampling errors of school types and contexts in which located Type of school Average no.of % pupils in primary Age of school kilometres to school joined with Region School location Public Semi-public People founded (in years) public services Sec.Sch % SE % SE % SE Mean SE Mean SE % SE Isolated 100.0 0.00 . . . . 28.0 19.90 5.1 1.45 . . Red River Rural 100.0 0.00 . . . . 31.9 0.88 2.4 0.06 0.8 0.45 Delta Urban 99.4 0.60 0.6 0.60 . . 23.0 1.92 1.5 0.15 1.2 1.24 Total 99.9 0.10 0.1 0.10 . . 30.3 0.81 2.2 0.06 0.8 0.54 Isolated 100.0 0.00 . . . . 29.7 1.60 10.6 0.89 26.8 4.58 Rural 99.8 0.22 . . 0.2 0.22 29.8 0.84 4.2 0.14 12.6 1.38 Northeast Urban 99.3 0.69 0.7 0.69 . . 22.7 1.80 1.7 0.15 5.3 2.18 Total 99.7 0.18 0.1 0.11 0.2 0.15 28.7 0.76 4.9 0.17 13.8 1.02 Isolated 100.0 0.00 . . . . 35.5 1.62 17.4 1.43 37.6 6.88 Rural 100.0 0.00 . . . . 33.9 1.80 5.2 0.50 20.2 4.10 Northwest Urban 100.0 0.00 . . . . 18.0 3.07 1.7 0.28 13.8 6.00 Total 100.0 0.00 . . . . 32.1 1.27 9.0 0.66 25.4 3.30 Isolated 100.0 0.00 . . . . 26.6 2.46 12.5 1.22 16.4 7.02 Rural 100.0 0.00 . . . . 28.6 1.37 4.1 0.20 1.2 0.95 North Central Urban 100.0 0.00 . . . . 23.9 3.36 1.4 0.08 . . Total 100.0 0.00 . . . . 27.7 1.13 4.5 0.24 2.5 0.97 Isolated 100.0 0.00 . . . . 17.8 1.12 7.1 0.79 14.6 6.01 Rural 100.0 0.00 . . . . 14.6 0.63 3.8 0.16 3.9 1.31 Central Coast Urban 99.6 0.43 0.4 0.43 . . 18.0 1.49 1.4 0.06 . . Total 99.9 0.11 0.1 0.11 . . 15.8 0.50 3.6 0.17 4.2 1.03 Isolated 100.0 0.00 . . . . 13.2 1.58 9.6 1.10 42.6 9.77 Central Rural 100.0 0.00 . . . . 13.3 1.26 4.8 0.45 15.2 4.89 Highlands Urban 96.4 3.60 . . 3.6 3.60 15.4 1.49 1.5 0.11 4.3 4.37 Total 99.0 0.97 . . 1.0 0.97 13.9 0.81 5.0 0.37 18.1 3.32 Isolated 100.0 0.00 . . . . 15.6 1.59 7.0 0.80 5.4 2.72 Rural 99.8 0.22 0.2 0.22 . . 20.1 0.93 4.0 0.22 1.2 0.68 Southeast Urban 98.9 1.15 . . 1.1 1.15 25.9 1.84 1.8 0.11 1.4 1.08 Total 99.4 0.48 0.1 0.10 0.5 0.48 22.0 0.85 3.5 0.17 1.8 0.63 Isolated 100.0 0.00 . . . . 16.8 0.93 5.6 0.50 5.5 1.87 Rural 99.9 0.12 0.1 0.12 . . 19.6 0.53 4.6 0.20 1.3 0.51 Mekong Delta Urban 100.0 0.00 . . . . 28.8 1.81 1.5 0.11 1.4 0.96 Total 99.9 0.08 0.1 0.08 . . 20.7 0.54 4.2 0.15 2.0 0.53 Isolated 100.0 0.00 . . . . 22.0 0.73 9.1 0.35 17.6 1.69 Rural 99.9 0.05 0.1 0.03 0.0 0.03 25.2 0.36 3.8 0.07 4.1 0.44 Vietnam Urban 99.2 0.56 0.2 0.13 0.6 0.55 23.8 0.69 1.6 0.04 2.0 0.61 Total 99.8 0.12 0.1 0.03 0.1 0.12 24.5 0.28 4.0 0.06 5.3 0.37 154 Vietnam Reading and Mathematics Assessment Study (b) Age of school During the war many school buildings were damaged and since the war there has been an extensive program of building schools. School heads were asked to give the year in which the school was opened or, if they did not know, at least to estimate the year in which it was opened. As can be seen from Table 5.4, the average Grade 5 pupil was in a school that was 24.5 years old. In other words, it was started in 1977. The range of the number of years schools had existed was from 0 to 119 years. The distribution of pupils in different ages of schools was as follows: Less than 5 years old 9.1 The average pupils were 6-10 years old 20.6 in the school there was 11-20 years old 19.3 24.5 years old, but 26% 21-30 years old 14.5 of pupils were in schools 31-40 years old 10.0 more than 40 years old. 41-50 years old 20.9 More than 50 years old 5.5 There was some variation among regions, namely that in the Central Highlands the average pupil was in a school that was only 14 years old. Isolated and rural schools in the Northwest were about 35 years old on average. At the provincial level, pupils in Soc Trang, Bac Lieu, Quang Nga, Dak Lak and Ben Tre were in schools that were less than 14 years old. In Cao Bang, the figure was 41 years old. (c) Average number of kilometres to public services Two attempts were made to judge the degree of urbanisation and rurality in which the schools were located. They have been reported here and under (d) below. In all cases it must be remembered that the head, when answering, was referring to the main campus of the school where it was a school with several campuses. School Heads were asked to state how many kilometres it was from the Isolated schools were, school to each of the following: nearest health clinic, nearest road for cars, on average, were nine nearest public library, nearest book shop, nearest lower secondary school, kms. from public and nearest market. These kilometres were then summed and divided by 6 ­ services. In rural areas, the number of public services. These averages were then again averaged for this was 3.6 kms. And in provinces and for all of Vietnam and these figures have been reported in urban areas, 1.6 kms. Table 5.4. In all regions the pupils in isolated schools were further from public amenities than pupils in rural or urban areas. The national average distance for isolated schools was nine kilometres which is considered to be quite far. In the Northwest in isolated areas the distance was 16 kms. And in the Northeast, North Central and Central Highlands it was more than ten kilometres. In Son La, the average distance was nine kms. 155 Vietnam Reading and Mathematics Assessment Study (d) Urban, rural, and remote context of schools A second question was asked to try and establish the urbanness/ruralness of school locations. School Heads were asked which of the following best described the location of the school: Isolated/remote Rural In or near a small town In or near a large town or city The final two categories were collapsed into one category, namely urban. The percentages of pupils in the different regions/areas have been presented in Table 5.5 below. School location Region Isolated Rural Urban % SE % SE % SE Red River Delta 1.2 0.66 81.5 1.49 17.3 1.43 Northeast 16.3 1.32 67.7 1.57 16.1 1.36 Northwest 35.3 3.48 49.7 3.40 15.0 2.67 North Central 9.7 1.76 75.8 2.68 14.5 2.12 Central Coast 12.0 1.95 63.0 3.14 25.0 2.85 Central Highlands 21.7 3.73 51.0 4.48 27.3 4.22 Southeast 12.5 1.79 45.9 2.31 41.6 2.22 Mekong Delta 15.6 1.38 67.3 1.86 17.2 1.69 Vietnam 12.3 0.63 66.5 0.77 21.3 0.68 Table 5.5: The percentage of pupils in different school locations by region 67% of pupils were in It can be seen that 12 percent of pupils were in isolated areas, 67 percent in rural areas, 21% in rural areas and 21percent in urban areas. There was considerable variation urban areas, and 12% in among regions. Whereas the Red River Delta had just above one percent in isolated areas. isolated schools the Northwest had 35 percent. The Southeast, Central Highlands and Central Coast had the most pupils in urban areas. e) Pupils in primary schools attached to a secondary school 95% of pupils were in From Table 5.4 it can be seen that five percent of pupils were in primary primary schools only schools attached to secondary schools. This also means that 95 percent of and 5% in primary pupils were in schools that were strictly only primary schools. In joint schools joined with schools it is usually the head of the secondary school who has prime secondary schools. In responsibility for the site. In isolated areas 17 percent of pupils were in isolated areas, 17% of primary schools attached to secondary schools, while in rural areas the figure pupils were in joint was four percent and in urban areas two percent. In the Central Highlands the schools. figure was 42 percent and in the Northwest it was 37 percent. In the following 156 Vietnam Reading and Mathematics Assessment Study provinces there were more than 20 percent of pupils in schools attached to secondary schools. It should be noted that only primary education has been compulsory so far but that several provinces made junior secondary education compulsory from December 2001. It is sometimes argued that it is better for pupils to be in schools attached to secondary schools because they then know what is in store for them. Often it is the case that governments have combined the schools because there are not enough pupils for just a primary school. On the other hand, some argue that when schools are attached to secondary schools then the schools are too big and that this can cause learning problems. They also argue that the two establishments are educationally different and therefore should be separate. The correlation between this variable and pupil achievement was near to zero (see Table 5.23). Primary school joined with Sec. School Province % SE Tuyen Quang 41.1 6.59 Lang Son 38.2 6.50 Ha Giang 36.7 7.08 Son La 34.8 6.37 Gia Lai 33.4 6.45 Lai Chau 32.3 6.35 Yen Bai 29.3 6.20 Cao Bang 27.7 6.08 Kon Tum 25.3 6.51 Lao Cai 22.5 5.73 Policy suggestion 5.2: a) Given the age of some school buildings, the MOET may wish to conduct a review of the quality of all primary school buildings. This point is raised later when the conditions of school buildings are discussed; b) The distance from public amenities for some very isolated communities is large and the possibility of increasing community cultural activities associated with school learning may be desirable. What was the size and quality of the teaching force in the primary schools? The last section dealt with the type of school and context of schools. In this 157 Vietnam Reading and Mathematics Assessment Study section the size and quality of the teaching force in the primary schools have been described. The school heads were asked about all of the teachers in their schools and not only the Grade 5 teachers. However the schools were those in which Grade 5 pupils were to be found. That is, the primary schools where there may have been pupils in Grades 1, 2 , 3, and 4 but not in Grade 5 were not included in the statistics reported here. % female Years of Pupil-teaching Total Boys enrolled N.of classes No.of classes Region School location teachers teacher training staff ratio enrolment in Grade 5 in school in Grade 5 % SE Mean SE Mean SE Mean SE Ratio SE Mean SE Mean SE Isolated 82 14.2 2.15 0.122 20.5 3.44 676 278.2 0.52 0.03 21.3 10.30 4.8 2.55 Rural 91 0.3 2.18 0.01 25.2 0.19 778 13.3 0.51 0.00 23.7 0.38 5.1 0.09 Red River Delta Urbal 96 0.4 2.50 0.026 28.4 0.65 1151 53.4 0.52 0.01 28.1 1.04 5.9 0.22 Total 91 0.3 2.24 0.011 25.7 0.19 841 15.6 0.51 0.00 24.4 0.40 5.3 0.09 Isolated 75 2.2 1.85 0.025 21.1 0.44 570 20.7 0.53 0.01 23.5 0.69 4.2 0.25 Rural 87 0.4 2.07 0.01 23.3 0.17 704 12.0 0.52 0.00 24.5 0.43 5.0 0.09 Northeast Urbal 96 0.4 2.31 0.028 25.5 0.67 802 34.1 0.52 0.01 23.2 0.82 4.9 0.17 Total 86 0.5 2.07 0.009 23.3 0.16 698 11.3 0.52 0.00 24.1 0.36 4.8 0.07 Isolated 59 2.8 1.57 0.042 19.5 0.59 639 44.6 0.58 0.02 27.6 1.70 4.2 0.31 Rural 78 2 1.90 0.025 18.0 0.40 671 32.4 0.55 0.01 27.9 1.16 4.7 0.19 Nothwest Urbal 93 0.9 2.20 0.084 16.6 0.95 588 61.9 0.54 0.02 20.6 1.72 4.3 0.35 Total 73 1.8 1.83 0.025 18.3 0.35 648 26.4 0.56 0.01 26.7 0.96 4.5 0.15 Isolated 62 3.3 1.93 0.042 20.8 1.13 538 41.9 0.53 0.01 23.2 1.88 4.0 0.33 Rural 87 0.9 2.06 0.014 28.1 0.38 739 26.9 0.52 0.00 23.1 0.71 4.6 0.14 North Central Urbal 91 0.9 2.50 0.052 26.0 0.62 850 33.8 0.54 0.01 24.5 1.11 5.1 0.31 Total 85 0.9 2.11 0.016 27.1 0.36 736 21.2 0.52 0.00 23.3 0.57 4.6 0.11 Isolated 70 3.1 1.90 0.059 25.7 1.02 696 55.2 0.53 0.01 24.1 1.31 4.1 0.25 Rural 73 0.9 2.04 0.014 28.6 0.26 883 18.5 0.52 0.00 26.6 0.49 5.1 0.11 Central Coast Urbal 87 0.9 2.27 0.04 30.2 0.94 1168 69.6 0.52 0.01 31.4 1.35 6.1 0.27 Total 76 0.8 2.08 0.015 28.6 0.32 932 25.4 0.52 0.00 27.5 0.54 5.2 0.11 Isolated 65 3.4 1.71 0.053 26.0 0.71 910 77.3 0.55 0.02 31.1 2.15 3.8 0.40 Central Rural 81 1.7 1.63 0.036 27.0 0.48 830 38.3 0.52 0.01 27.1 1.06 4.5 0.21 Highlands Urbal 93 1.2 1.83 0.064 28.9 0.55 990 58.2 0.52 0.01 28.7 1.52 5.8 0.34 Total 81 1.3 1.70 0.027 27.3 0.29 891 29.1 0.53 0.01 28.4 0.80 4.7 0.17 Isolated 72 1.8 1.74 0.027 28.3 0.73 734 44.5 0.52 0.01 23.8 1.35 4.4 0.21 Rural 80 0.8 1.83 0.011 27.9 0.36 835 26.2 0.51 0.00 25.9 0.74 5.1 0.14 Southeast Urbal 86 0.8 1.96 0.019 32.4 0.59 1271 61.7 0.52 0.00 31.6 1.22 6.3 0.25 Total 82 0.5 1.87 0.01 29.9 0.35 1003 31.0 0.52 0.00 28.0 0.65 5.5 0.13 Isolated 47 1.9 1.76 0.03 28.4 0.60 783 31.8 0.53 0.01 26.1 1.00 4.5 0.20 Rural 55 0.8 1.74 0.009 26.1 0.21 740 14.1 0.52 0.00 24.7 0.44 4.7 0.09 Mekong Delta Urbal 78 1.5 1.90 0.022 29.9 0.52 1094 50.1 0.52 0.01 30.4 1.04 6.0 0.23 Total 58 0.8 1.77 0.008 27.1 0.22 808 15.3 0.52 0.00 25.9 0.38 4.9 0.08 Isolated 63 1.1 1.79 0.015 24.7 0.31 691 15.0 0.53 0.00 25.2 0.49 4.3 0.10 Rural 79 0.3 1.98 0.004 26.1 0.11 766 7.6 0.52 0.00 24.6 0.21 4.9 0.04 Vietnam Urbal 89 0.4 2.16 0.014 29.2 0.26 1085 25.0 0.52 0.00 28.7 0.54 5.8 0.11 Total 79 0.2 1.99 0.004 26.5 0.10 825 8.1 0.52 0.00 25.5 0.20 5.0 0.04 Table 5.6: Means, percentages and sampling errors of teaching force characteristics The average Grade 5 (a) Percentage of female teachers pupils were in schools where79% of teachers As can be seen from Table 5.6, the average Grade 5 pupil was in a school were female, but in where 79 percent of the teaching staff was female. In other words, 21 percent isolated areas this was were male. This figure is different from the figure reported for Grade 5 less. teachers only where only 73 percent were female (see Table 4.1). The figure 158 Vietnam Reading and Mathematics Assessment Study reported here is for all of the teaching staff. There were proportionately more female teachers in schools in urban areas than in rural or isolated areas. The figure for the Red River Delta and urban Northeast was 90 percent and for the Mekong Delta it was only 58 percent. The percentages of female teachers were high in Ha Noi, Hai Phong and Thai Nguyen but low in Soc Trang and Tra Vinh. The correlation between this variable and achievement was high (even after home background had been partialled out). This has implications for the allocation of teachers to schools. (b) Teaching or professional qualifications Again the coding of this indicator was the same as for the Grade 5 teacher The average Grade 5 professional qualification. The school heads reported that their teachers had pupil was in a school the following teacher training. where 56% of teaching Percentage staff had 12+2 training, No teacher training = 0 yrs 0.6 21% 9+3, 12% 12+3, Below pedagogical HS = 1 yrs 6.8 and 4% 12+4 training. Pedagogical HS (9+3) = 1.5 yrs 20.9 Pedagogical HS (12+2) = 2 yrs 55.8 College (12+3) = 3 yrs 12.0 University (12+4) = 4 yrs 3.9 In Table 5.6 the average value for years of teacher-training in Vietnam was 1.99 years. This meant that the average Grade 5 pupil was in a school where all of the teachers on average had the Pedagogical High School Training of two years. The urban schools had teachers with 2.2 years and the isolated school had teachers with 1.8 years. Again this raises the question of the criteria used when allocating teachers to schools. It should be noted that the teachers in the Northwest isolated schools had only 1.6 years and in Soc Trang province the average was 1.5 years which is actually below the government minimum. It would seem that the provinces with low figures should be the first to receive in-service training. (c) Pupil-teaching staff ratio The pupil-teaching staff ratio must not be confused with the pupil-teacher On average, primary ratio for a class. The pupil-staff ratio is the total number of pupils in a school schools had 26.5 pupils divided by the total number of teachers. In other words it is the number of per teaching staff pupils per average staff member. Not all teachers teach all of the time. Some member. have lesson periods when they do not teach and can be used, when necessary, to help other teachers or to take over classes in the case of a teacher being absent. The indicator can be regarded as the teacher `wealth' of a school for the pupils. 159 Vietnam Reading and Mathematics Assessment Study From Table 5.6 it can be seen that the average pupil-teaching staff ratio for Vietnam was 26.5 pupils per staff member. Urban schools had 29.2 pupils per staff member and isolated had 24.7. But, isolated schools in the Mekong Delta had 28.4 pupils per teacher. The highest figure was for urban schools in the Southeast region. There did not appear to be a problem with the pupil-staff ratio and the relationship to pupil achievement was near zero (see Table 5.18). (d) Total enrolment or size of school There has been much discussion about the `size of school' or `total enrolment' of a primary school. On the one hand, the larger the enrolment a school has the less is the cost of providing some amenities such as laboratories for science or sports facilities. On the other hand, a criticism has been that the larger the school the more overwhelming it is for small children and that it is not easy for the social integration of the children into the school. What then were the different sizes of school? The average school The average Grade 5 pupil was in a primary school with an enrolment of 825. enrolment was 825, but As expected, the size of school was smaller in isolated areas (691) and in in isolated areas, it was urban areas it was 1085. The urban areas in the Central Coast and the Red 691, in urban areas River Delta had the highest enrolments. Some of the schools had several 1085. campuses and hence the enrolment was spread over the various campuses. In the IEA studies it was, in some cases, found that more than 800 pupils in a school was associated with lower achievement. In Vietnam, 44.6 percent of children were in schools with more than 800 pupils. In Vietnam, a school is allowed to up to a maximum of 30 classes with a maximum of 35 pupils per class. Hence, the maximum total enrolment is 1050. In Chapter 6 it was seen that 21 percent of pupils were in schools where this maximum figure was exceeded. However, the correlation between total enrolment and pupil achievement was near to zero. (e) Percentage boys in Grade 5 But, only 48% girls in All schools in Vietnam are co-educational. In theory, therefore, there should Grade 5. be a 50:50 split between boys and girls unless there has been sex-biased drop- out before that point. At the same time, it is of interest to determine if a grade was more boy-biased or more girl-biased. From Table 5.6 it can be seen that there were 52 percent boys and 48 percent girls. In the Northwest there were only 44 percent girls and 56 percent boys in Grade 5 and in the isolated schools in this province there were only 42 percent girls. 160 Vietnam Reading and Mathematics Assessment Study At the provincial level, the provinces of Kon Tun, Lao Cai, Son La, and Lai Chau all had fewer than 45 percent girls in Grade 5. This is a problem of girls either not entering school or dropping out early from school. Again, this is a matter of persuading the parents to enrol the girls and keeping them in school. This can be helped by other measures such as incentives to mitigate opportunity costs, cost barriers and the like. The school intervention programs mentioned in Chapter 3 should be implemented and monitored. (f) Number of classes in school and in Grade 5 Another indicator of the size of the school and how it is structured is the On average there were 25 number of classes in the school. Two questions were asked of School Heads. class groups per school, The first concerned the number of classes in the school and the second but in some provinces concerned the number of classes in Grade 5. From Table 5.6 it can be seen this was over 30. that the average number of classes was 25 but in the isolated schools in the Central Highlands this was 31. The number of Grade 5 classes was 5 and the range across areas in regions was from 4.1 to 6.1. It is said that the maximum number of class groups in a school should be 30. The average number of class groups was higher than 30 in the following provinces. Number of classes in school Provice Mean SE Kien Giang 36.15 1.82 Ho Chi Minh 31.84 1.83 Dong Nai 31.72 1.44 Binh Phuoc 30.94 1.23 Da Nang 30.56 1.34 Son La 30.01 1.45 (g) Average class size In the school questionnaire, there was a question about the number of classes in the school and a separate question about enrolment. It was possible, therefore, to divide the number of pupils by the number of classes and thus arrive at a rough estimation of the average class size in primary schools. The results have been presented in Table 5.7 below. 161 Vietnam Reading and Mathematics Assessment Study Table 5.7: Average class size by region and school location The average number of Class size (from Shquest) Region School location pupils per class was 32.8, Mean SE Isolated 39.1 12.13 but 5% of pupils were in Rural 35.2 0.18 Red River Delta classes with over 40 Urban 40.0 0.65 Total 36.1 0.19 pupils and 5% in classes Isolated 26.3 0.65 with fewer than 22. Rural 31.0 0.24 Northeast Urban 35.1 0.69 Total 30.7 0.24 Isolated 25.9 0.90 Rural 26.9 0.61 Northwest Urban 28.3 1.38 Total 26.7 0.49 Isolated 25.1 1.16 Rural 33.7 0.42 North Central Urban 36.2 0.77 Total 33.1 0.39 Isolated 28.5 1.58 Rural 34.3 0.34 Central Coast Urban 36.6 0.60 Total 33.9 0.39 Isolated 29.4 0.90 Rural 31.8 0.56 Central Highlands Urban 35.3 1.13 Total 31.9 0.43 Isolated 30.6 0.72 Rural 32.3 0.30 Southeast Urban 36.9 0.55 Total 33.7 0.29 Isolated 30.6 0.41 Rural 31.3 0.25 Mekong Delta Urban 34.6 0.62 Total 31.7 0.20 Isolated 28.4 0.45 Rural 32.8 0.11 Vietnam Urban 36.3 0.24 Total 32.8 0.11 It can be observed that the schools in urban areas had larger class sizes than in schools elsewhere. As expected, the isolated areas had the smallest class sized. The class size range was quite large. Five percent of pupils were in schools with classes with over 40 pupils in them, and five percent in classes with fewer than 22 pupils. What has emerged from the review of the size and quality of the teaching force? A higher proportion of female teachers on the staff was associated with higher pupil achievement. The more teacher training that teachers on the whole staff had received then the higher was the pupil achievement. These findings need to be reviewed and a more careful allocation of teachers to schools made. Although no problem appeared in the review of total enrolment it must remembered only its relationship with cognitive achievement has been examined. The MOET should also examine the social development of the child and identify if larger schools inhibit social integration or not. Policy suggestion 5.3: The MOET may wish to review its policy of the allocation of teachers to schools in order to take account of the proportion of female teachers on the staff and the teacher training qualifications of the staff. 162 Vietnam Reading and Mathematics Assessment Study Policy suggestion 5.4: This study has been concerned with cognitive learning. Social development is also an important feature of primary schooling. It might be desirable for the MOET to initiate a study on the social development of pupils in primary school and in that context examine the relationship between the social integration of pupils and total enrolment of schools. What were some characteristics of the operation of the school? This section is concerned with the operation of the school. How many shifts did schools have? How many satellite campuses did a school have? How often had it been inspected in the last few years? What was the purpose of the inspectors' visits? What were the main activities of the school heads? What were the major problems in the schools? How many days were lost last year in the school's schedule? Table 5.8: Percentages and sampling errors of indicators of the operation of the schools % pupil in % pupil in % pupil in No. of satellite Region School location first shift second shift full day campuses % SE % SE % SE Mean SE Isolated 17 16.6 22 22.1 60 38.7 1.73 1.94 Rural 22 1.0 22 1.0 56 1.8 1.63 0.08 Red River Delta Urban 20 2.7 18 2.2 62 4.3 0.54 0.09 Total 22 0.8 21 0.8 57 1.4 1.44 0.07 Isolated 67 2.2 31 2.0 2 1.1 4.49 0.34 Rural 48 1.0 38 0.9 14 1.0 2.64 0.12 Northeast Urban 35 1.7 30 1.9 35 3.0 0.90 0.14 Total 49 0.9 36 0.7 16 0.9 2.66 0.11 Isolated 71 3.3 28 3.1 1 0.8 5.97 0.57 Rural 59 2.1 39 2.1 2 0.9 4.64 0.38 Northwest Urban 36 5.5 30 5.2 34 6.8 0.89 0.28 Total 60 2.0 34 1.7 6 1.4 4.55 0.28 Isolated 59 4.0 38 4.1 3 1.8 3.58 0.50 Rural 44 1.6 42 1.3 14 2.3 1.17 0.12 North Central Urban 28 3.8 24 3.4 48 6.8 0.66 0.25 Total 43 1.6 39 1.0 18 2.0 1.33 0.12 Isolated 53 1.6 45 1.4 2 1.2 3.66 0.29 Rural 46 1.1 45 1.0 9 1.5 2.48 0.10 Central Coast Urban 42 1.4 39 1.5 19 2.6 1.32 0.15 Total 46 0.8 44 0.8 11 1.2 2.33 0.09 Isolated 57 2.2 42 2.2 1 0.6 4.97 0.53 Rural 48 1.6 44 1.5 7 2.0 2.54 0.33 Central Highlands Urban 39 3.9 34 3.1 27 6.1 1.16 0.18 Total 48 1.5 41 1.3 11 2.2 2.69 0.22 Isolated 53 1.5 46 1.5 1 0.9 2.27 0.23 Rural 51 0.9 43 0.9 6 1.3 1.64 0.09 Southeast Urban 40 2.0 35 1.7 24 2.9 0.64 0.08 Total 47 1.0 40 0.8 13 1.4 1.30 0.06 Isolated 51 1.0 47 0.9 1 0.5 3.74 0.23 Rural 52 0.6 45 0.5 3 0.5 2.90 0.09 Mekong Delta Urban 45 1.6 40 1.5 15 2.3 1.14 0.13 Total 50 0.5 45 0.4 5 0.5 2.73 0.08 Isolated 57 1.0 40 0.8 3 0.9 3.90 0.14 Rural 43 0.5 38 0.4 19 0.7 2.13 0.04 Vietnam Urban 36 1.0 32 0.9 32 1.7 0.84 0.05 Total 43 0.4 37 0.3 20 0.6 2.08 0.03 163 Vietnam Reading and Mathematics Assessment Study (a) Shift system Some schools ran a system of two shifts ­ a morning shift and an afternoon shift. Indeed, 80 percent of pupils were enrolled in these two shifts. A few schools only in the Mekong Delta had a third shift although efforts have been made to eliminate three shifts per day. Some schools have what is called a `full-day' school on top of one or two shifts. The government has been encouraging all schools to move to a full-day system. What is a full-day system? A full-day system requires the pupils to attend school morning and afternoon, although there is certain flexibility in the actual number of hours this will depend on school location and weather. Parent must pay for their children to attend a full-day school whereas a half-day shift is free. 20% pupils attended a From Table 5.8 it can be seen that 43 percent of all primary school pupils full-day school, 43% a attended a first shift, 37 percent a second shift, and 20 percent a full-day morning half-day shift, school. There was considerable variation among school location areas and and 37% a second half- regions. It can be seen that for first shift schools while only 36 percent of day shift.. pupils in urban areas attended them the figure was 57 percent for isolated schools. The differences were not very large for second shifts but for full-day schools there were differences: 32 percent for urban schools, 19 percent for rural schools and only three percent for isolated schools. The Red River Delta schools had most pupils in full-day schooling (57%) while the Northwest had only six percent. In the provinces of Nam Dinh, Thai Binh, Hai Duong, Ha Noi, and Bac Ninh the average pupil was in a school with more than 50 percent full-day shift pupils. In Nam Dinh the figure was 96 percent in full- day schooling. Nam Dinh is a province well known for its progressiveness in education. But the provinces of Ninh Thuan and Soc Trang, the average pupil was in schools with no pupils in full day schooling. (b) Satellite campuses In certain areas it was deemed desirable to set up a satellite campus system because either the children lived too far from the main school or it was not possible to extend the current main school building. The average pupil was in a school with 2.1 satellite campuses. In isolated areas the figure was 3.9 satellite campuses, in rural areas it was 2.13 and in urban areas it was less than one satellite campus. Thirty-eight percent of pupils were in schools with one or two satellite campuses, twenty percent in schools with three or four campuses, and twelve percent in schools with five to ten campuses. Thirty percent were in schools without satellite campuses. In the isolated Northwest schools and Central Highlands the figures were six and five respectively. In the province of Son La the average pupil was in a school with six satellite campuses. The correlation between the number of satellites and the number of school resource items was ­0.39 indicating that there were fewer resources in schools with more satellite campuses. The correlation between the percentage of pupils in full-day schooling and schools with satellites was ­0.25 indicating that there were fewer pupils in full day-schooling in schools with more satellite campuses. 164 Vietnam Reading and Mathematics Assessment Study From Figure 5.3, it is clear that the pupils in satellite campus schools were disadvantaged in terms of many school features. But, there were different kinds of satellite schools. The satellite schools in Ha Noi were created because there was not enough room in the ordinary schools whereas the satellite schools in several remote provinces (for example Cao Bang) were built because it was too far for the pupils to go to the main school. It is also the case that resources may be allocated to each school irrespective of how many campuses it had. In isolated areas there is multi-grade teaching whereas in Ha Noi this is not the case. It is suggested that the MOET should undertake a special study of satellite campus schools in order to identify the kind of help they need to be brought up to par in terms of conditions of schooling (buildings, resources, teacher quality). Schools with satellite campuses had fewer classroom and school resources, a teaching staff with less professional training, and a lower ratio of female teaching staff.. Figure 5.3: Differences in material and human resources in main schools and main schools with satellites Policy suggestion 5.5: The MOET might wish to conduct a special study of satellite campus schools in order to identify the kind of help they need to be brought up to par in terms of conditions of schooling (buildings, resources, and teacher quality.) How often and for what purposes had the schools been inspected? a) How often had the schools been inspected? The data for this survey were collected in April, 2001. If the authorities are to know what is happening in their schools, it is clear that full inspections of schools should be made. In Chapter 4, it was already seen how often 165 Vietnam Reading and Mathematics Assessment Study inspectors had visited the classrooms and how the teachers perceived the various roles of the inspectorate. In this chapter a description has been made of full inspections of schools. A full inspection is where a team of national inspectors visits the school to examine the whole school as compared with provincial and BOET inspectors' visits to individual teachers. As can be seen from Table 5.9 the percentage of Grade 5 pupils in schools that were inspected within the last five years (i.e. from 1996 to 2001) was 95. The MOET regulations state that a school should be visited once every three to five years. There was some variation among school location areas. In the urban areas in the Red River Delta and the Central Highlands the figure was 100 percent but in the isolated schools in the Central Highlands the figure was only 85 percent. In the Cao Bang province the figure was only 75 percent. Table 5.9: Percentage of pupils and sampling errors in schools inspected School Inspected in last 5 years Times visited since 1999 Region location % SE Mean SE Isolated 100.0 0.00 2.1 0.83 Red River Rural 97.7 0.74 3.6 0.17 Delta Urban 100.0 0.00 5.2 0.61 Total 98.1 0.61 3.8 0.17 Isolated 89.2 3.20 3.4 0.32 Rural 96.9 0.72 4.1 0.24 Northeast Urban 93.4 3.36 3.7 0.33 Total 95.1 0.86 3.9 0.17 Isolated 89.6 4.44 1.9 0.16 Rural 95.1 2.44 2.0 0.16 In general, 95% of Northwest Urban 85.1 7.36 3.3 0.94 pupils were in schools Total 91.6 2.17 2.2 0.17 inspected in last 5 years. Isolated 90.6 6.91 2.4 0.35 On average, schools North Rural 98.5 0.79 3.5 0.29 Central Urban 98.0 1.48 3.6 0.30 were visited once or Total 97.7 0.86 3.4 0.24 twice per year. Isolated 98.1 1.90 4.3 0.90 Central Rural 91.9 1.81 4.2 0.36 Coast Urban 86.8 4.15 5.9 0.46 Total 91.4 1.63 4.6 0.27 Isolated 84.7 8.29 1.8 0.23 Rural 96.6 2.04 2.6 0.26 Central Highlands Urban 100.0 0.00 3.2 0.88 Total 95.0 2.37 2.6 0.28 Isolated 93.5 3.39 4.7 0.88 Rural 93.1 1.74 6.4 0.40 Southeast Urban 97.9 0.71 8.4 0.79 Total 95.1 0.95 7.0 0.36 Isolated 88.8 3.03 5.7 0.69 Mekong Rural 93.4 0.97 5.3 0.27 Delta Urban 96.9 1.48 6.6 0.68 Total 93.3 0.74 5.6 0.20 Isolated 90.5 1.82 3.9 0.26 Rural 95.8 0.41 4.3 0.11 Vietnam Urban 96.3 0.68 5.9 0.29 Total 95.3 0.35 4.6 0.08 166 Vietnam Reading and Mathematics Assessment Study A further question was asked about the number of times a school had been visited by any kind of inspector for whatever reason since 1998. The figures have been reported in Table 5.9. It can be seen that an average pupil was in a school visited four times in the period. The isolated schools had had the least number of visits (3.9), the rural schools four and the urban schools six visits. The Northwest region and Central Highlands had the fewest visits. It is clear that in most, but not all regions, isolated schools were visited least. (b) Reasons for inspection since 1999 The average number of times that schools were inspected together with the reasons for such visits have been recorded in Table 5.10. Table 5.10: Means and sampling errors for the number of times schools inspected since 1999 Inspection of To assist teachers To address crisis or Routine To advise the School Full inspection individual to improve their problem Other purposes Region inspection school head location teachers teaching skills in the school Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Isolated 0.5 0.56 0.3 0.46 3.2 2.30 0.0 0.00 0.3 0.46 1.0 1.38 0.3 0.46 Red River Rural 1.0 0.06 1.2 0.07 1.8 0.12 1.1 0.10 0.8 0.08 0.1 0.02 0.2 0.04 Delta Urban 1.2 0.11 2.1 0.38 2.6 0.51 1.4 0.29 1.4 0.46 0.1 0.06 0.5 0.15 Total 1.0 0.05 1.3 0.09 2.0 0.13 1.1 0.10 0.9 0.09 0.1 0.03 0.3 0.05 Isolated 1.1 0.09 1.2 0.27 1.4 0.15 1.1 0.13 1.1 0.14 0.1 0.03 0.4 0.10 Rural 1.1 0.06 1.3 0.09 1.9 0.17 1.1 0.09 0.9 0.08 0.1 0.04 0.3 0.04 Northeast Urban 1.1 0.07 1.0 0.16 1.6 0.14 0.9 0.13 0.9 0.10 0.1 0.04 0.2 0.05 Total 1.1 0.04 1.2 0.07 1.8 0.11 1.0 0.06 0.9 0.06 0.1 0.03 0.3 0.03 Isolated 0.9 0.12 0.5 0.17 0.8 0.16 0.7 0.16 0.5 0.11 0.1 0.05 0.3 0.13 Rural 1.1 0.09 0.4 0.07 0.9 0.12 0.8 0.11 0.6 0.10 0.0 0.01 0.1 0.05 Northwest Urban 0.7 0.12 0.6 0.26 1.2 0.27 0.5 0.23 0.4 0.18 0.2 0.11 0.3 0.15 Total 1.0 0.07 0.4 0.08 0.9 0.09 0.7 0.09 0.5 0.07 0.1 0.02 0.2 0.05 Isolated 0.9 0.14 0.4 0.13 1.2 0.25 0.7 0.21 0.7 0.27 0.1 0.08 0.1 0.08 Rural 1.1 0.05 1.1 0.13 1.5 0.16 0.8 0.08 0.8 0.10 0.0 0.01 0.1 0.02 North Central Urban 1.2 0.13 0.8 0.19 1.4 0.19 0.6 0.14 0.5 0.12 0.1 0.05 0.1 0.05 Total 1.1 0.05 1.0 0.10 1.5 0.13 0.8 0.07 0.7 0.08 0.0 0.01 0.1 0.02 Isolated 0.6 0.10 1.2 0.27 2.3 0.73 1.1 0.71 0.4 0.12 0.1 0.03 0.2 0.10 Central Rural 0.5 0.04 1.2 0.12 2.0 0.19 0.7 0.09 0.5 0.07 0.0 0.01 0.3 0.06 Coast Urban 0.6 0.08 1.6 0.19 2.6 0.25 1.1 0.28 0.6 0.14 0.2 0.08 0.6 0.13 Total 0.6 0.04 1.3 0.11 2.2 0.15 0.8 0.12 0.5 0.06 0.1 0.02 0.4 0.05 Isolated 1.0 0.12 0.4 0.13 0.9 0.20 0.5 0.13 0.7 0.18 0.1 0.06 0.3 0.10 Central Rural 1.0 0.11 0.5 0.09 1.2 0.20 0.8 0.21 0.7 0.14 0.1 0.03 0.2 0.07 Highlands Urban 0.8 0.12 1.0 0.24 3.1 1.70 0.6 0.15 0.7 0.19 0.1 0.06 0.1 0.06 Total 1.0 0.07 0.6 0.09 1.7 0.49 0.7 0.11 0.7 0.09 0.1 0.03 0.2 0.04 Isolated 0.7 0.12 1.2 0.21 2.4 0.43 0.9 0.33 0.6 0.16 0.1 0.05 0.2 0.10 Rural 0.7 0.05 1.2 0.12 3.7 0.33 1.1 0.19 0.7 0.10 0.1 0.02 0.5 0.09 Southeast Urban 0.8 0.05 2.3 0.28 5.4 0.71 1.9 0.39 1.2 0.26 0.2 0.05 0.6 0.20 Total 0.7 0.03 1.7 0.13 4.2 0.29 1.4 0.18 0.9 0.13 0.1 0.03 0.5 0.09 Isolated 0.8 0.09 1.4 0.19 3.4 0.60 1.4 0.33 0.6 0.11 0.1 0.02 0.2 0.06 Mekong Rural 0.7 0.03 1.4 0.12 3.1 0.19 1.0 0.13 0.8 0.09 0.1 0.02 0.4 0.08 Delta Urban 0.7 0.08 1.9 0.24 3.5 0.68 1.0 0.21 0.8 0.25 0.2 0.06 0.6 0.12 Total 0.7 0.03 1.5 0.09 3.2 0.15 1.0 0.11 0.8 0.08 0.1 0.02 0.4 0.06 Isolated 0.9 0.05 1.0 0.09 2.0 0.19 1.0 0.12 0.7 0.06 0.1 0.02 0.3 0.04 Rural 0.9 0.02 1.2 0.04 2.2 0.07 1.0 0.05 0.8 0.04 0.1 0.01 0.3 0.02 Vietnam Urban 0.9 0.04 1.7 0.11 3.3 0.24 1.2 0.13 1.0 0.11 0.2 0.02 0.4 0.07 Total 0.6 0.04 1.3 0.04 2.4 0.06 1.0 0.04 0.8 0.03 0.1 0.01 0.3 0.02 167 Vietnam Reading and Mathematics Assessment Study Most of the visits were It can be seen that the most frequent type of inspection was that of the for routine inspections `individual teacher' (an average of more than twice since 1999). `Subject or inspecting individual matter inspection' was the second most frequent, then in descending order teachers. `to assist teachers to improve their teaching skills', `to advise the School Head', and full inspections. Visits to a school `to address a crisis or problem in the school' were only 0.1 times in the two years preceding when the data were collected ­ except in the Red River Delta where it was 1.0! Individual teacher inspection was more frequent in urban areas (except for the two Delta regions) and full inspections in isolated areas. In the regulations for inspectors great emphasis is laid on the visiting schools in order `to assist teachers improve their teaching skills' and `to advise the School Head'. It was therefore somewhat surprising to find such a low number of visits for these reasons. The provinces with the most inspections for these reasons also tended to be rich provinces. To assist teachers to improve their teaching skills To advise the school head PROVINCE Mean SE PROVINCE Mean SE Bac Ninh 2.39 0.48 Ha Noi 1.59 0.58 Ho Chi Minh 2.31 0.58 Lao Cai 1.53 0.24 An Giang 2.23 0.68 Ho Chi Minh 1.47 0.36 Da Nang 1.93 0.32 An Giang 1.43 0.44 Ninh Thuan 1.70 0.47 Da Nang 1.41 0.34 Ba Ria-Vung Tau 1.70 0.39 Bac Ninh 1.40 0.24 Lao Cai 1.69 0.22 Ninh Thuan 1.39 0.35 Can Tho 1.66 0.44 Hai Duong 1.28 0.32 Ha Noi 1.57 0.31 Tien Giang 1.24 0.48 Lam Dong 1.55 0.65 Bac Giang 1.22 0.21 The relationship between inspectors' visits, as reported by school heads, and pupil achievement was nearly zero. This may be in order because the inspectors only visited the schools that were performing poorly in order to improve them or it may be because there was little effect from the inspectors visits as they are now conducted. It is clear from what has been reported above and what was reported in Chapter 4 that there is some confusion about the role of inspectors. Furthermore the fact that the inspectors had not visited some schools according to what was required from the regulations makes it important that the role of the inspectorate is reviewed and, if required, amended and made clear to all school heads and teachers. It is assumed that the isolated schools probably need more visits and this is not happening. There was no correlation between the number of inspector visits and the number of behavior problems. Might it be the case that the inspectors only went to the schools close to the office or easy to access? 168 Vietnam Reading and Mathematics Assessment Study Policy suggestion 5.6: The several roles of the inspectorate should be made clearer. There should be a system of monitoring that the inspectors are doing their jobs well. What were the Head's activities and those he/she planned for the school? The amount of teaching that heads do has already been discussed. It is one thing to gain insights into the school by teaching at various grade levels in it but there are many more responsibilities that a head has. (a) School Heads' activities The major activities as perceived by the MOET were listed and each head School heads rated the was asked to rate how important that he or she perceived the activities to professional be. A scale was constructed where `not important' was coded 1, `of some development of teachers importance' 2, and `very important' 3. Thus a mean value of 2 meant `of and improving school conditions as the most some importance' From Table 5.11 it can be seen that nearly all Heads important activities. ticked either `important' or `very important'. The `professional Organising extra development of teachers' was rated to be the most important activity of curricular activities School Heads. `Extra curricular activities' had the lowest ratings. This received the lowest was surprising because these activities often give an aura to a school and rating. can make the difference between an outstanding school and an average school. `Contacts with the local community' that were deemed to be important were rated more important in isolated areas than in rural or urban areas. So were `discussing educational objectives'. This may be because more community help is needed in the isolated areas. Furthermore, because the teachers in the isolated areas had less experience as teachers, then the heads deemed that it was important to discuss educational objectives more. But, in general there were few differences between school locations and between regions. When asked which were the most important activities, the Heads rated administrative tasks to be the most important, then the professional development of teachers and thirdly the taking care of the school buildings and resources. 169 Vietnam Reading and Mathematics Assessment Study Table 5.11: Means and sampling errors of School Heads' perceived importance of activities Contacts with Monitoring Discuassing Professional Professional Extra- Improving Organising Administrativ School local pupils' Edu. Dev. of Dev. of curricula school excellence Region e tasks location community progress objectives teachers Head activittie conditions competition Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Isolated 2.56 0.47 2.15 0.19 2.43 0.47 2.58 0.40 2.57 0.47 3.00 0.00 2.15 0.19 2.85 0.19 2.30 0.30 Red River Rural 2.39 0.03 2.47 0.02 2.70 0.02 2.56 0.03 2.75 0.02 2.65 0.03 2.04 0.01 2.74 0.02 2.58 0.03 Delta Urban 2.25 0.05 2.56 0.06 2.72 0.04 2.53 0.06 2.83 0.04 2.55 0.05 2.07 0.04 2.82 0.04 2.45 0.05 Total 2.37 0.02 2.48 0.02 2.70 0.02 2.56 0.02 2.76 0.02 2.64 0.02 2.05 0.01 2.75 0.02 2.55 0.03 Isolated 2.58 0.05 2.64 0.05 2.79 0.04 2.73 0.04 2.77 0.04 2.71 0.04 2.14 0.04 2.82 0.04 2.56 0.05 Rural 2.53 0.03 2.54 0.03 2.72 0.02 2.60 0.02 2.76 0.02 2.61 0.03 2.13 0.02 2.77 0.02 2.61 0.02 Northeast Urban 2.47 0.05 2.51 0.05 2.73 0.05 2.55 0.05 2.78 0.04 2.55 0.05 2.08 0.02 2.77 0.04 2.56 0.05 Total 2.53 0.02 2.55 0.02 2.73 0.02 2.62 0.02 2.77 0.02 2.62 0.02 2.12 0.01 2.78 0.02 2.59 0.02 Isolated 2.73 0.07 2.52 0.08 2.60 0.07 2.71 0.08 2.74 0.07 2.68 0.06 2.10 0.04 2.79 0.06 2.58 0.07 Rural 2.69 0.06 2.41 0.05 2.62 0.06 2.58 0.06 2.72 0.05 2.66 0.06 2.07 0.03 2.81 0.05 2.62 0.06 Northwest Urban 2.42 0.11 2.51 0.10 2.64 0.11 2.58 0.10 2.87 0.06 2.69 0.09 2.12 0.08 2.79 0.08 2.62 0.10 Total 2.66 0.05 2.46 0.04 2.61 0.04 2.62 0.04 2.75 0.03 2.67 0.04 2.09 0.02 2.80 0.04 2.61 0.04 Isolated 2.78 0.07 2.41 0.10 2.62 0.09 2.78 0.07 2.74 0.09 2.63 0.08 2.00 0.02 2.87 0.06 2.44 0.10 Rural 2.46 0.03 2.54 0.04 2.68 0.04 2.58 0.04 2.75 0.03 2.63 0.04 2.01 0.02 2.80 0.03 2.54 0.04 North Central Urban 2.33 0.07 2.52 0.08 2.80 0.06 2.53 0.09 2.77 0.07 2.64 0.08 2.11 0.04 2.77 0.08 2.40 0.09 Total 2.47 0.03 2.52 0.03 2.69 0.03 2.59 0.03 2.75 0.03 2.63 0.04 2.02 0.02 2.80 0.03 2.51 0.03 Isolated 2.46 0.08 2.59 0.09 2.81 0.07 2.46 0.08 2.78 0.07 2.40 0.10 2.11 0.06 2.78 0.08 2.51 0.09 Rural 2.57 0.04 2.69 0.04 2.77 0.03 2.46 0.03 2.74 0.03 2.68 0.03 2.07 0.02 2.71 0.03 2.46 0.04 Central Coast Urban 2.56 0.05 2.71 0.06 2.76 0.05 2.56 0.05 2.84 0.04 2.60 0.05 2.17 0.04 2.83 0.04 2.61 0.05 Total 2.56 0.03 2.68 0.03 2.77 0.02 2.49 0.03 2.77 0.02 2.63 0.03 2.10 0.02 2.75 0.02 2.50 0.03 Isolated 2.55 0.10 2.52 0.11 2.53 0.11 2.58 0.10 2.71 0.09 2.51 0.10 2.00 0.12 2.72 0.09 2.62 0.10 Central Rural 2.55 0.07 2.54 0.07 2.70 0.06 2.56 0.07 2.71 0.06 2.60 0.06 2.04 0.03 2.82 0.05 2.56 0.07 Highlands Urban 2.58 0.08 2.70 0.09 2.79 0.07 2.54 0.10 2.85 0.06 2.49 0.08 2.09 0.07 2.81 0.07 2.51 0.11 Total 2.56 0.04 2.58 0.05 2.69 0.04 2.56 0.05 2.75 0.04 2.55 0.05 2.05 0.03 2.79 0.04 2.56 0.05 Isolated 2.53 0.07 2.74 0.07 2.65 0.06 2.66 0.07 2.87 0.04 2.56 0.07 2.11 0.04 2.69 0.06 2.54 0.06 Rural 2.48 0.03 2.69 0.03 2.66 0.03 2.46 0.03 2.79 0.03 2.47 0.04 2.07 0.03 2.64 0.03 2.42 0.03 Southeast Urban 2.56 0.04 2.75 0.04 2.70 0.04 2.59 0.04 2.77 0.04 2.53 0.05 2.05 0.03 2.77 0.04 2.29 0.04 Total 2.52 0.03 2.72 0.02 2.68 0.03 2.54 0.03 2.79 0.02 2.51 0.03 2.07 0.02 2.70 0.02 2.38 0.03 Isolated 2.62 0.05 2.70 0.05 2.71 0.05 2.59 0.05 2.77 0.05 2.53 0.06 2.07 0.04 2.81 0.04 2.46 0.05 Mekong Rural 2.58 0.03 2.64 0.02 2.70 0.02 2.58 0.02 2.76 0.02 2.54 0.03 2.08 0.02 2.73 0.02 2.50 0.03 Delta Urban 2.51 0.04 2.67 0.04 2.66 0.04 2.54 0.04 2.80 0.04 2.53 0.06 2.10 0.04 2.78 0.04 2.45 0.06 Total 2.57 0.02 2.65 0.02 2.69 0.02 2.58 0.02 2.77 0.02 2.53 0.02 2.08 0.01 2.75 0.02 2.48 0.02 Isolated 2.60 0.02 2.61 0.02 2.68 0.03 2.65 0.03 2.77 0.02 2.59 0.02 2.08 0.02 2.79 0.02 2.51 0.02 Rural 2.50 0.01 2.57 0.01 2.70 0.01 2.56 0.01 2.76 0.01 2.60 0.01 2.06 0.01 2.75 0.01 2.54 0.01 Vietnam Urban 2.47 0.02 2.64 0.02 2.72 0.02 2.56 0.02 2.80 0.02 2.56 0.02 2.09 0.01 2.79 0.02 2.44 0.02 Total 2.51 0.01 2.59 0.01 2.70 0.01 2.57 0.01 2.77 0.01 2.59 0.01 2.07 0.01 2.76 0.01 2.51 0.01 (b) Activities planned for school There are several activities that are planned for the whole school that can have a beneficial effect on learning. For example, in a study of 32 different school systems it was found that where schools had pupils produce a school magazine this was related to the general level of reading comprehension of pupils (Elley, 1993). It is to be expected that when there is a general ambience 170 Vietnam Reading and Mathematics Assessment Study in the school in certain activities then this will be beneficial in certain school subjects. For example, where there are competitions in spelling, then spelling should improve. Where there are mathematics competitions between schools then this will improve mathematics achievement. And so on. The NIES group responsible for the questionnaires selected six activities that they believed were beneficial to learning. The extent to which they occurred in schools was recorded by the School Heads. From Table 5.12 it can seen that seen that none of the activities was undertaken a great deal. Table 5.12: Percentages and sampling errors of pupils in schools where certain activities were practised Never Once per year More than once per year Activities % SE % SE % SE Production of school magazine 75 0.9 16 0.8 9 0.6 Story telling competition 6 0.4 73 0.9 21 0.8 general knowledge quizzes 12 0.6 52 1.0 37 1.0 Camping expeditions 29 1.1 60 1.1 11 0.7 School day celebrations 60 1.0 27 1.0 13 0.7 Open meeting of school education council 5 0.4 70 0.8 25 0.8 Academic knowledge quizzes was the most undertaken extra curricular activities. Figure 5.4: Frequency of school activities 171 Vietnam Reading and Mathematics Assessment Study From Table 5.12 and Figure 5.4 above, it can be seen that more than half pupils were in schools which never had `production of school magazine' and `school day celebrations'. The most frequent was `general knowledge quizzes' that do require pupils to read widely. The second most frequent activity was `open meeting of the school education council', and the third was `story telling competitions'. The least frequent activity was the `production of a school magazine', the one variable from other research that is known to have a positive effect on reading achievement. Only two of the variables had correlations with achievement larger than 0.08. They were the `production of a school magazine' and `story telling competitions'. It was these two variables and `general knowledge quizzes that were also correlated with home background (.11, .18, and .18 respectively). The other variables had virtually zero correlations with achievement as was expected. This is an important issue if the reading levels are to be improved. The `culture' of reading in a school does lead to higher levels of reading comprehension. If levels are to be raised it is important that the authorities provide the wherewithal for pupils to read and then encourage them to do so. At the same time it is good to encourage school activities to do with reading such as the first two activities mentioned above. It is to be hoped that teacher training courses, management courses and inspectors' visits will all include many ideas about how to initiate such activities in schools and that the inspectors will actively encourage them. Policy suggestion 5.7: In order to strengthen the culture of reading in schools, the Curriculum Centre may wish to decide on which reading associated activities in school are desirable and then experiment on the best way to have them run. What were the major problems that schools experienced in Vietnam? There has been considerable anecdote about the different kinds of problems that schools had experienced but no systematic evidence was available. An attempt was made in this survey to list several of the problems that schools had been said to experience both from pupils and from teachers and to ask heads to rate them in terms of how often the heads had to deal with them. (a) Pupil problem areas School heads perceived Twelve problems were listed. The percentages of pupils in schools where the that there were many Head said they were experiencing problems never, sometimes, or often have pupils health problems. been presented in Table 5.13 below. 172 Vietnam Reading and Mathematics Assessment Study Table 5.13: School Heads' perception of pupil behavioural problems at school Never Sometimes Often Pupil behavioural problems % SE % SE % SE Pupils arriving late at school 20.4 0.79 74.8 0.84 4.9 0.40 Pupil absenteeism 27.3 0.83 65.4 0.92 7.3 0.50 Pupils skipping classes 52.5 0.91 42.4 0.88 5.1 0.41 Pupils dropping out of school 29.1 0.82 57.7 0.99 13.2 0.66 Classroom disturbance by pupils* 48.8 1.00 48.2 0.97 3.0 0.32 Cheating by pupils* 50.7 0.95 47.5 0.97 1.8 0.24 Use of abusive language by pupils* 36.7 0.96 59.4 0.95 3.9 0.31 Vandalism by pupils* 47.9 0.87 46.3 0.79 5.7 0.42 Theft by pupils* 71.0 0.92 26.0 0.82 3.1 0.34 Intimidation or bullying of pupils by pupils* 52.4 0.98 44.6 1.00 3.0 0.26 Intimidation/verbal abuse of teachers by pupils* 89.3 0.64 7.7 0.52 2.9 0.30 Drug abuse by pupils 96.5 0.41 0.9 0.19 2.6 0.33 Fights among pupils 26.7 0.94 70.2 0.97 3.2 0.29 Pupil health problems 9.7 0.52 72.9 0.88 17.5 0.79 Whereas, some problems can be expected to be experienced some times, those said to be experienced often must be serious. The most frequent problem was a health problem. Unfortunately, nothing was asked about the kinds of health problems being experienced. It sounds serious and, if not already known, it would be a desirable for the MOET to conduct a small study just to discover what these health problems are. Many of the problems were more prevalent in the Southeast and Mekong Delta regions although absenteeism was high in isolated schools in the Northeast and Northwest regions. However, the drug problem, although low was mostly concentrated in the border provinces and the provinces with big cities. The problems marked with an asterisk are those that are said to form Pupils coming late to part of the moral education classes that the pupils receive in the sense schools and absenteeism that an effort is made to have pupils understand why these particular as well as fights among behaviours and undesirable. It is very difficult to have good teaching if pupils were seen as behavioural problems are found frequently in schools and teachers have serious. to deal with them instead of teaching. It would appear that the authorities may well have to take some action. But it will be important to look at teacher behavioural problems before making comment on 173 Vietnam Reading and Mathematics Assessment Study what might be done. Figure 5.5: Pupil behavioural problems (often + sometimes) by region It can be seen that it is in the Mekong Delta and in the Southeast regions that there were most problems. (b) Teacher problems Nine teacher problems were identified and they have been listed in Table 5.14 and Figure 5.6. Table 5.14: Means and sampling errors of problems with teachers reported Never Sometimes Often Teacher behavioural problems % SE % SE % SE Teacher arriving late at school 35.5 1.05 61.4 1.09 3.1 0.29 Teacher absenteeism 65.6 0.79 30.8 0.75 3.6 0.30 Teacher skipping classes 76.9 0.74 20.0 0.70 3.1 0.29 Intimidation or bullying of pupil by teachers 79.2 0.80 18.7 0.75 2.1 0.28 Teacher lost unity 71.5 0.75 26.0 0.84 2.5 0.28 Use abusive language by teachers 89.7 0.56 8.6 0.51 1.7 0.25 Drug abuse by teachers 98.0 0.30 0.3 0.10 1.7 0.27 Alcohol abuse or possessions by teachers 74.7 0.79 23.9 0.82 1.4 0.23 Teacher health problems 16.3 0.77 72.8 1.06 10.9 0.71 174 Vietnam Reading and Mathematics Assessment Study Teachers'health problems were considered to be a big problem (80% of pupils in schools where this was said to be a problem). `Teachers arriving late at school' was deemed to be a problem in schools where 65 percent of the pupils were. Teacher absenteeism was also deemed to be a problem. As can be seen from Figure 5.6, there were many problems in the Mekong Delta region. Teacher health problems were seen as a issue by heads. To some extent teachers turning up late Figure 5.6: Teacher behavioural problems (sometimes + often) by region was also a problem. It was interesting to note that in isolated areas about 40 percent of non-Kinh pupils were in schools where alcohol abuse was perceived to be a problem. For Kinh pupils the percentage was 25. Soc Trang province had the highest percentages of pupils in schools with the four problems: teacher coming late, teacher absenteeism, skipping class, and alcohol abuse. The teacher problems were inter-related as can be seen from the correlations in Table 5.15. Table 5.15: Correlations among teacher behavioural problems Late Absent Skip class Bullying Non Abusive lang. Drugs Alcohol Health cooperation Late 1.00 0.51 0.50 0.32 0.39 0.32 0.26 0.34 0.22 Absent 1.00 0.73 0.42 0.47 0.45 0.40 0.42 0.20 Skip class 1.00 0.45 0.49 0.51 0.47 0.46 0.24 Bull 1.00 0.48 0.53 0.46 0.33 0.19 No cooperation 1.00 0.54 0.44 0.36 0.24 Abusive language 1.00 0.63 0.44 0.25 Drug 1.00 0.41 0.19 Alcohols 1.00 0.20 Health 1.00 175 Vietnam Reading and Mathematics Assessment Study The highest correlation was between skipping class and being absent (.73) and the second highest was between drug abuse and abusive language (.63). But in general the problems were interlinked. There was a relationship There was also a strong relationship between the pupil behavioural problems between pupil and and the teacher behavioural problems as can be seen from Table 5.16. The teacher behavioural correlations between teacher absent and teacher skip class on the one hand problems. This has and pupil reading achievement on the other were -.12 and -.10 (and ­0.09 and implications for ­0.07 after the effect of home background was removed) and with pupil math changing teacher achievement ­0.15 and ­0.13 (partial correlation was ­0.12 and ­0.11) and behaviour. this indicated that the more teachers were absent or skipped class the lower was the achievement of the pupils. The correlations between pupil behaviours and teacher behaviours have been presented below. The correlations indicate that where a school tended to have teacher behaviour problems it also had pupil behavioural problems. Table 5.16: Correlations among pupil and teacher behavioural problems Teacher Late Absent Skip class Bullying Non- coop. Abuse lang. Drug Alcohol Health Late 0.33 0.20 0.22 0.16 0.17 0.19 0.16 0.19 0.21 Absent 0.40 0.36 0.34 0.22 0.24 0.23 0.18 0.27 0.21 Skip class 0.42 0.42 0.44 0.34 0.36 0.35 0.27 0.31 0.22 Drop out 0.41 0.43 0.42 0.30 0.32 0.29 0.22 0.30 0.20 Disturbance 0.35 0.35 0.35 0.29 0.31 0.28 0.19 0.29 0.21 Cheating 0.29 0.24 0.26 0.27 0.26 0.25 0.18 0.22 0.19 Abusive language 0.29 0.22 0.26 0.25 0.27 0.26 0.18 0.20 0.18 Pupil Vandalism 0.41 0.44 0.42 0.37 0.39 0.36 0.26 0.32 0.20 Theft 0.35 0.42 0.45 0.42 0.45 0.46 0.38 0.30 0.20 Intimidation 0.35 0.34 0.37 0.38 0.40 0.37 0.28 0.25 0.20 Abuse of teacher by pupil 0.32 0.44 0.48 0.48 0.45 0.55 0.54 0.35 0.21 Drug 0.25 0.34 0.40 0.39 0.36 0.51 0.73 0.36 0.18 Fight 0.34 0.32 0.31 0.32 0.33 0.31 0.23 0.26 0.21 Health 0.21 0.20 0.21 0.19 0.20 0.22 0.17 0.15 0.59 The highest correlation was between the heads perceiving drug problems with teachers and perceiving drug problems with pupils (.73). It is suggested that the Ministry officials read the correlations with care. Pupils dropping out, skipping class and being absent are highly correlated with teachers being late, absent, and skipping class. Pupil vandalism, theft and intimidation were highly correlated with teacher non-co-operation. Abusive language by pupils goes together with abusive language by teachers and with drugs. What do these relationships mean and what are their implications for action by the national and provincial authorities? Clearly, it is difficult to run a school well if the teachers are suffering from poor health, often absent, skipping class and coming late. Whether the poor health comes from the quality of the teacher housing or not is unknown. But it likely that the teacher transportation to go to school, the teacher's enthusiasm to go to school, and the weather conditions are also contributory factors. It would seem that the school heads' perceptions of these problems - both pupil and teacher problems ­ are a matter for concern. It is suggested that 176 Vietnam Reading and Mathematics Assessment Study the MOET undertake a special study of the actual existence of these problems and identify the factors that come together to make the problems such that they combine into a major problem for the School Head to run the school well. Policy suggestion 5.8: The MOET might wish to consider including content in the school management program and in-service teacher training courses in order to have school heads and teachers diminish these problems in school. (c) Days lost per year A `lost day' is a day on which no teaching takes place. This can be due to factors like a late start of term, the organisation of examinations, school festivals, storms. Given that at best there are only 165 days in a school year, every school day is very important. Table 5.17: Days lost per year School School lost days Region location Mean SD SE Isolated 4.5 2,90 2.53 Rural 3.9 4,79 0.21 Red River Delta Urban 2.9 3,95 0.41 Total 3.7 4,65 0.19 On average, 4.5 Isolated 3.4 3,95 0.36 schooldays were lost per Rural 2.8 3,29 0.18 Northeast Urban 2.3 3,16 0.32 year. Total 2.8 3,40 0.15 Isolated 4.5 4,47 0.60 Rural 3.6 3,81 0.52 Northwest Urban 4.7 5,01 0.99 Total 4.1 4,28 0.32 Isolated 7.2 5,92 1.02 Rural 5.3 4,84 0.36 North Central Urban 4.6 4,56 0.63 Total 5.4 4,96 0.32 Isolated 6.7 5,24 0.96 Rural 4.7 4,55 0.34 Central Coast Urban 4.5 4,29 0.47 Total 4.9 4,62 0.28 Isolated 4.8 5,35 0.87 Rural 2.8 3,71 0.48 Central Highlands Urban 2.8 3,64 0.62 Total 3.2 4,21 0.34 Isolated 4.2 4,73 0.79 Rural 3.7 4,70 0.31 Southeast Urban 2.0 3,63 0.29 Total 3.1 4,38 0.20 Isolated 8.1 8,81 0.99 Rural 7.9 11,01 0.85 Mekong Delta Urban 4.1 5,39 0.60 Total 7.3 10,04 0.63 Isolated 5.7 6,37 0.34 Rural 4.8 6,60 0.21 Vietnam Urban 3.2 4,28 0.16 Total 4.5 6,20 0.15 177 Vietnam Reading and Mathematics Assessment Study The average number of days lost was 4.5 (see Table 5.17). The range was from no days to 78 days. The standard deviation was 6.20. However, in urban areas there was less dispersion of days lost than in the other areas. The top three schools for the most number of school days lost were in Kieng Giang, An Giang, and Tien Gian provinces in the Mekong Delta. Floods may have caused more than 2 months of school days lost in the Mekong Delta. More days were lost in isolated areas than in rural areas and in urban areas only 3.2 days were lost. Isolated schools in North Central had a value of 7.2. These many days lost are too many and it would seem that the MOET should discover why these many days were lost and take measures to combat this. What were the conditions of the school facilities? In most countries of the world there is a large range in the condition of school buildings. Some are very well kept and maintained. Others appear to be in a state of total disrepair. First of all a question was asked about the number of classrooms in each school that were permanent (with a concrete roof), semi-permanent (with a tile roof), and temporary (with a leaf roof or no roof). (a) Number of classrooms and square metrage per pupil There were somewhat In Table 5.18, the number of permanent, semi-permanent, and temporary more semi-permanent classrooms have been given together with the average square metrage per than permanent pupil for each category. Finally the total square metrage per pupil for the classrooms and only a few temporary whole school has been given. Whereas the square metrage per pupil in classrooms per school. classrooms is a normal statistic, the square metrage per pupil for the school Schools in isolated areas as a whole is not. had the least amount of permanent and the greatest of temporary classrooms. 178 Vietnam Reading and Mathematics Assessment Study Table 5.18: Means and sampling errors for no. of classrooms, square metrage per pupil and total square metrage per pupil Permanent classrooms Semi-permanent classrooms Temporary classrooms Square meters School per pupil per Region No. of Square meters No. of Square meters No. of Square meters location classroom classrooms per pupil classrooms per pupil classrooms per pupil Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Isolated 7.15 0.49 0.61 0.25 6.18 3.84 0.44 0.13 3.68 6.52 0.08 0.11 1.13 0.15 Red River Rural 8.19 0.30 0.51 0.02 8.09 0.30 0.48 0.02 1.94 0.24 0.09 0.01 1.07 0.02 Delta Urban 14.74 0.88 0.80 0.10 5.80 0.64 0.26 0.04 1.23 0.43 0.03 0.01 1.08 0.12 Total 9.43 0.28 0.56 0.03 7.73 0.27 0.44 0.02 1.85 0.21 0.08 0.01 1.07 0.02 Isolated 2.54 0.42 0.16 0.03 8.83 0.53 0.67 0.05 8.97 0.87 0.80 0.14 1.63 0.13 Rural 4.56 0.25 0.26 0.02 9.77 0.31 0.70 0.03 5.19 0.33 0.31 0.03 1.26 0.03 Northeast Urban 7.48 0.66 0.40 0.04 9.50 0.65 0.66 0.05 1.67 0.39 0.09 0.02 1.12 0.05 Total 4.75 0.23 0.27 0.02 9.58 0.26 0.69 0.02 5.37 0.28 0.35 0.03 1.30 0.03 Isolated 3.46 0.63 0.26 0.08 10.95 1.39 0.65 0.10 12.08 1.26 0.79 0.10 1.71 0.13 Rural 2.14 0.37 0.14 0.04 10.80 0.88 0.70 0.06 9.49 0.87 0.44 0.06 1.27 0.07 Northwest Urban 11.08 1.98 0.96 0.18 7.31 1.40 0.91 0.44 2.82 1.00 0.43 0.32 2.37 0.84 Total 4.11 0.49 0.31 0.04 10.35 0.73 0.71 0.08 9.88 0.68 0.56 0.06 1.58 0.13 Isolated 2.17 0.78 0.17 0.05 8.59 1.55 0.82 0.18 8.18 1.52 0.68 0.24 1.68 0.29 North Rural 6.10 0.53 0.34 0.03 8.73 0.44 0.80 0.23 1.71 0.30 0.07 0.01 1.21 0.23 Central Urban 10.32 1.01 0.63 0.09 8.03 1.00 0.45 0.06 1.18 0.68 0.04 0.02 1.13 0.07 Total 6.50 0.49 0.37 0.03 8.62 0.37 0.75 0.18 2.40 0.36 0.12 0.03 1.24 0.17 Isolated 5.76 1.36 0.28 0.07 8.10 1.02 0.71 0.14 4.13 0.99 0.24 0.06 1.23 0.12 Central Rural 7.93 0.68 0.31 0.03 12.45 0.46 0.75 0.04 2.31 0.47 0.05 0.01 1.10 0.03 Coast Urban 14.47 1.32 0.49 0.05 9.58 0.81 0.47 0.05 1.82 0.72 0.04 0.01 0.99 0.05 Total 9.53 0.70 0.35 0.03 11.21 0.40 0.67 0.03 2.53 0.37 0.07 0.01 1.09 0.03 Isolated 5.39 1.29 0.25 0.07 10.16 1.40 0.63 0.10 7.40 1.11 0.46 0.19 1.34 0.25 Central Rural 4.68 0.89 0.22 0.04 11.13 1.07 0.53 0.05 4.00 0.71 0.19 0.04 0.93 0.04 Highlands Urban 8.28 1.67 0.31 0.07 12.19 1.51 0.61 0.07 2.93 0.82 0.09 0.03 1.01 0.04 Total 5.79 0.72 0.25 0.03 11.21 0.76 0.57 0.04 4.53 0.51 0.22 0.04 1.04 0.06 Isolated 2.92 1.02 0.14 0.06 11.78 1.03 0.90 0.07 2.97 0.56 0.11 0.02 1.15 0.05 Rural 4.33 0.51 0.21 0.03 11.56 0.54 0.87 0.08 1.75 0.21 0.08 0.01 1.15 0.08 Southeast Urban 15.75 1.25 0.44 0.04 9.49 0.96 0.43 0.04 0.97 0.26 0.03 0.01 0.90 0.04 Total 9.34 0.64 0.30 0.02 10.78 0.49 0.69 0.05 1.60 0.15 0.06 0.01 1.05 0.04 Isolated 2.99 0.73 0.14 0.03 11.67 0.47 0.84 0.05 3.95 0.39 0.22 0.04 1.20 0.06 Mekong Rural 4.92 0.39 0.28 0.02 10.61 0.27 0.87 0.04 3.10 0.22 0.19 0.03 1.34 0.06 Delta Urban 12.48 1.45 0.43 0.05 12.10 0.95 0.63 0.05 1.20 0.36 0.04 0.01 1.10 0.04 Total 6.02 0.45 0.29 0.02 11.03 0.26 0.83 0.03 2.97 0.17 0.17 0.02 1.27 0.04 Isolated 3.38 0.31 0.19 0.02 10.16 0.31 0.76 0.03 6.55 0.39 0.45 0.05 1.39 0.05 Rural 6.00 0.17 0.33 0.01 9.86 0.14 0.72 0.04 2.99 0.13 0.15 0.01 1.19 0.04 Vietnam Urban 12.88 0.45 0.51 0.03 9.44 0.41 0.49 0.02 1.39 0.17 0.05 0.01 1.05 0.03 Total 7.30 0.17 0.35 0.01 9.81 0.14 0.67 0.03 3.19 0.11 0.16 0.01 1.18 0.03 179 Vietnam Reading and Mathematics Assessment Study It can be seen that there were more semi-permanent classrooms than permanent ones. The urban areas were reasonably well provided for but the isolated areas had many temporary and semi-permanent buildings. This is best seen from Figure 5.7 below. Figure 5.7: Type of classroom by school location It is clear that isolated areas have many more temporary and semi-permanent classrooms than do other areas and it is here that the updating program will need to begin. There is nothing new in these figures. They are known to the MOET and work is progressing. (b) Heads' perceptions of the condition of school buildings School Heads were asked to classify their schools in terms of the general conditions of the school buildings. The percentages of pupils in each of the categories have been presented in Table 5.19. 180 Vietnam Reading and Mathematics Assessment Study Table 5.19: Percentages and sampling errors of pupils in schools where School Heads have differing perceptions of general condition of school buildings Needs complete Some classrooms Most classrooms Some classrooms School In good condition Region rebuilding need major repairs need major repairs need minor repairs location % SE % SE % SE % SE % SE Isolated 57.8 40.23 42.2 40.23 0. 0. 0. 0. 0. .0 Rural 24.9 2.27 33.8 3.26 6.9 1.38 28.1 2.34 6.3 1.34 Red River Delta Urban 9.3 3.83 32.4 6.20 11.0 3.75 33.9 6.13 13.4 4.81 Total 22.6 1.97 33.7 2.92 7.6 1.26 28.8 2.23 7.4 1.39 Isolated 29.4 4.33 35.7 4.40 17.4 3.90 15.7 3.50 1.8 0.92 Rural 32.2 2.01 31.8 2.40 9.1 1.63 22.4 1.66 4.4 1.04 Northeast Urban 30.9 3.79 31.2 4.19 5.7 1.83 24.5 4.93 7.7 3.58 Total 31.6 1.50 32.4 1.98 9.9 1.24 21.7 1.39 4.5 0.87 Isolated 46.8 7.09 34.8 6.62 13.5 4.59 4.9 2.82 .0 0. Rural 51.1 6.08 33.6 5.85 8.5 3.58 5.5 2.42 1.3 1.25 Northwest Urban 22.9 7.47 22.3 7.82 11.7 7.27 33.9 11.03 9.3 6.63 Total 45.5 4.25 32.4 3.86 10.8 2.54 9.4 2.53 2.0 1.12 Isolated 36.1 10.71 46.7 9.95 8.0 5.03 8.0 6.03 1.2 1.24 Rural 16.5 2.71 38.8 3.98 7.2 2.16 26.6 3.50 10.9 2.20 North Central Urban 9.2 2.93 35.7 8.23 8.2 3.96 28.8 7.44 18.1 8.06 Total 17.2 2.30 39.1 3.25 7.5 1.77 25.2 2.67 11.1 2.00 Isolated 20.7 7.36 50.9 8.57 10.7 6.54 15.2 6.93 2.4 2.47 Rural 25.0 3.25 50.1 3.59 6.5 1.76 15.3 2.75 3.1 1.18 Central Coast Urban 16.6 4.28 39.6 5.46 7.4 2.53 25.7 4.36 10.6 2.90 Total 22.4 2.53 47.7 2.75 7.2 1.49 17.8 2.41 4.9 1.12 Isolated 22.0 7.73 49.3 10.22 3.3 2.00 24.3 9.31 1.1 1.14 Rural 15.8 4.40 50.9 6.68 14.9 4.99 18.1 5.16 0.2 0.22 Central Highlands Urban 18.6 6.32 44.2 9.04 7.7 4.82 22.6 7.95 6.9 4.75 Total 17.9 3.13 48.7 4.96 10.4 3.02 20.7 3.66 2.2 1.35 Isolated 18.3 5.77 52.1 8.25 9.8 3.59 15.8 7.07 4.0 4.01 Rural 25.0 3.41 40.2 3.78 13.5 2.55 16.7 2.77 4.6 1.48 Southeast Urban 27.9 4.42 24.1 3.75 4.6 1.74 29.7 4.64 13.6 2.93 Total 25.4 2.44 35.0 2.62 9.3 1.47 22.1 2.37 8.3 1.39 Isolated 17.8 3.24 46.3 4.30 11.6 3.33 21.3 3.72 3.1 2.28 Rural 25.0 1.74 44.1 1.97 12.4 1.52 14.6 1.67 3.9 0.80 Mekong Delta Urban 24.6 4.10 38.3 4.65 10.4 3.32 18.9 4.15 7.8 2.48 Total 23.8 1.51 43.5 1.82 11.9 1.28 16.4 1.63 4.5 0.83 Isolated 26.2 2.33 44.7 2.41 11.2 1.86 15.7 2.04 2.2 0.89 Rural 24.8 0.99 39.2 1.20 9.5 0.80 21.0 0.90 5.5 0.58 Vietnam Urban 21.4 2.01 32.4 2.13 7.5 1.05 27.1 2.12 11.6 1.40 Total 24.2 0.92 38.4 1.05 9.3 0.66 21.6 0.81 6.4 0.45 181 Vietnam Reading and Mathematics Assessment Study About 60% of pupils It can be seen that the school heads perceived the condition of school were in schools where buildings not to be in very good shape. Twenty-four percent of pupils were heads deemed their in schools where the heads thought that the buildings needed to be school to be needing completely rebuilt and 38 percent of pupils were in schools where the heads either a complete thought that some classrooms needed major repairs. It was the isolated rebuilding or major schools (the blue bar in Figure 5.8 below) which were perceived to be in repairs. the worst condition. It can happen that it is the better heads with well- achieving pupils who consider that their schools are in a worse condition. These heads simply have higher standards and expect more. In this case a zero correlation would be expected between the condition of school building and pupil achievement. And, this is the case. It was also the case that the newer schools (less than ten years old) tended to be perceived to be in a better condition. This was particularly the In Figure 5.9, it can be seen that it is in the Northeast and Northwest case in the Northeast regions that rebuilding is deemed to be most required by the heads. It is and Northwest important that the school buildings are in order. It would be wise to request the most recent assessment of the conditions of the buildings by the authority responsible for school buildings at the national level and compare the findings of this study with the other assessment. Figure 5.8: Building condition by school location Figure 5.9 below depicts the Heads' perception of the condition of the school buildings by region. 182 Vietnam Reading and Mathematics Assessment Study Figure 5.9: Condition of school building by region Policy suggestion 5.9: The MOET might wish to compare the findings from this study with their own assessments of the condition of school buildings. If there are grounds for concern, then appropriate plans should be made to have a detailed four year plan of rebuilding. Priority should be given to the isolated schools in the Northwest and Northeast regions. What were the resources in the schools? School resources are very important. In the 5 province study (ibid Griffin, 1999) it was seen that the differences in school resources among schools was very high. The correlation with achievement was also very high. This meant that schools that had more resources had pupils who scored higher on the achievement test than schools with low levels of school resources. This is not a surprising result because many studies have shown the same result. The national results have been presented in Table 5.20. The expanded table by region and by school location has been presented as Appendix 5.1. 183 Vietnam Reading and Mathematics Assessment Study Table 5.20: Percentages and sampling errors of pupils in schools with different resources Pupils in schools with Resources % SE Tape recorder 70 0.8 School or community hall 68 0.6 Teacher /staff room 30 1.0 Separate office for school head 53 0.8 Store room 14 0.7 First aid kit 25 0.9 Separate medical room 69 0.9 Sports area/Play ground 62 1.0 Piped water/Water tank/Borehole/Spring 73 0.8 Drinking water 67 0.9 Electricity (mains or generator) 85 0.7 Telephone 77 0.8 Fax machine 1 0.2 School garden 36 0.9 Typewriter 32 0.6 Duplicator 2 0.3 Microphone/amplifier equipment 66 0.7 Tape recorder 14 0.7 Overhead Projector 8 0.5 TV set 38 0.8 Video cassette recorder 25 0.6 Photocopier 6 0.4 Computer 15 0.5 Fence or hedge around school borders 74 0.8 Cafeteria 13 0.6 Music equipment 35 0.9 Sports equipment 46 0.9 As can be seen from Figure 5.10 below, there are more pupils in schools in urban areas with more resources than in rural schools and, as usual, it is the pupils in the isolated areas that have the fewest resources. 184 Vietnam Reading and Mathematics Assessment Study The MOET decided on 27 different resource items that should be Figure 5.10: Selected resources by school location assessed. The same list of resources was then examined by region. The Red River Delta schools were relatively well- off and the schools in the Northwest were the least well resourced. Figure 5.11: Selected resources by region Isolated schools were badly- off. 185 Vietnam Reading and Mathematics Assessment Study Urban schools were better resourced than rural which in turn were better resourced than isolated schools. However, TV set, typewriter and a staff room existed for more pupils in isolated areas. Isolated schools tended to be lacking a First Aid Kit and special medical room. In Cao Bang, Son La, and Tra Vinh less than three percent of pupils were in schools with these resources. Regionally, schools in the Red River Delta were best resourced and the Northwest (the yellow line in Figure 5.11) was least well resourced. Drinking water and libraries were low in the Northwest and Central Highlands. Only 13 percent of pupils in Cao Bang were in schools with a library. The following resources were associated with home background (correlations of 0.18 or more). This indicated that it was the parents rather than the authorities that were associated with their children's schools having more of these particular resources. Or, that the better off parents lived in areas that had been connected to electricity and water. The list was: School library School Head's office Medical room First aid kit Piped water Drinking water Electricity Telephone Microphone VCR Computer Fence around the school borders Music equipment Sports equipment It should be noted that the correlation between total resources and home background was 0.42. If the items of school resources are summed, it is possible to see for the regions what the general level of provision was. These means have been reported in Table 5.21. It can be seen again that the urban schools in the Red River Valley and Northwest are well resourced and the poorest schools from the resource point of view are in the isolated regions in the Northwest and Central Highlands. The two graphs (Figures 5.12 and 5.13) below depict the situation well. 186 Vietnam Reading and Mathematics Assessment Study Figure 5.12: Total school resources by school location Table 5.21: Means and sampling errors of resources Region School School resources location Mean SE Isolated 14.3 3.86 Rural 12.9 0.13 Red River Delta Urban 16.3 0.43 Total 13.5 0.14 Isolated 7.0 0.43 Rural 9.8 0.19 Northeast Urban 12.8 0.36 Total 9.8 0.15 Isolated 4.9 0.38 Rural 7.0 0.42 Northwest Urban 13.5 0.99 Total 7.2 0.36 Isolated 8.0 0.58 Rural 11.4 0.32 North Central Urban 15.8 0.45 Total 11.7 0.29 Isolated 7.9 0.68 Rural 10.6 0.28 Central Coast Urban 15.6 0.42 Total 11.5 0.27 Isolated 4.0 0.50 Central Rural 7.7 0.63 Highlands Urban 11.8 0.51 Total 8.0 0.44 Isolated 9.0 0.66 Rural 11.3 0.32 Southeast Urban 16.7 0.30 Total 13.2 0.19 Isolated 6.6 0.22 Rural 7.7 0.15 Mekong Delta Urban 13.3 0.41 Total 8.5 0.15 Isolated 7.1 0.24 Rural 10.4 0.10 Vietnam Urban 15.1 0.16 Total 11.0 0.09 187 Vietnam Reading and Mathematics Assessment Study Figure 5.13: Total school resources by region The provinces that had fewer than eight items of resources have been listed below. Mean of resources Province Mean SE Cao Bang 4.91 0.46 Son La 5.59 0.44 Ha Giang 5.82 0.51 Ca Mau 5.97 0.57 Bac Kan 5.98 0.53 Tra Vinh 6.25 0.37 Lang Son 6.57 0.59 Bac Lieu 6.64 0.46 Kon Tum 6.97 0.54 Lao Cai 7.42 0.73 Soc Trang 7.96 0.46 There would appear to be a severe lack of resources in many schools. It might be that the MOET believes that not all of the items mentioned are required by a school in order to make it function well. If this is the case the MOET should decide on the subset of items of resources that it wishes to have and the tables can be recomputed and then interpreted. It should be noted that the list suggested is not the same as the items listed as Ministry benchmarks in Chapter 6. 188 Vietnam Reading and Mathematics Assessment Study Once this has been arrived at, the provinces where there is a general lack of resources is known and the authorities can begin to allocate resources to these provinces first. Policy suggestion 5.10: The MOET may wish to decide on the minimum resources needed for a school to function well and then re-run the analyses above. What help was given by the community to the schools? A school exists for a community and within a community. The general ambience of the school is much dependent on the community. The discipline in the school must be supported by the community. The relationship between the school and community has been shown to be important in many studies. Good relationships are associated with more learning and fewer behavioural problems and even with less crime2. A question was asked of school heads about the contributions that parents and/or the community made to the school. The percentages of pupils in 78% of pupils were in schools where the community made contributions has been presented in schools where parents Table 5.22 and the means for the sum of 14 possible contributions have been helped in the presented for all regions in Figure 5.14 and school locations areas in Figure maintenance of school 5.15. Unfortunately for several of the activities it is not known whether the facilities, 70% in contributions were in cash or kind or whether the parents actually came and building school helped maintain parts of the school as in the first item. facilities, 53% in construction and repair Seventy-eight percent of pupils were in schools where the head stated that furniture. In all cases, parents helped in the maintenance of school facilities, 70 percent building of these contributions were school facilities, and 53 for the construction or maintenance and repair of more frequent in rural furniture, equipment and the like. In all cases it was in the rural areas where than in urban or isolated this was more frequent than in the urban or isolated areas. Twenty-four areas. percent of pupils were in school where heads said that the parents donated books to the school (this time more in urban schools), 30 percent were in schools where the parents helped with extra-curricular activities including school trips (mostly in urban areas), 25 percent where parents assisted in teaching and/or teach or supervise the pupils themselves without pay (again mostly in urban areas), and finally24 percent provided land for the school and teacher housing mostly in isolated areas). --------------------- 2 Coleman J. S. (1994) Family, School and Social Capital In: Husén T and Postlethwaite T N International Encyclopedia of Education, Oxford, Pergamon 189 Vietnam Reading and Mathematics Assessment Study Contributions in the Nineteen percent of pupils were in school where the parents helped towards form of payment and the payment of the non-salaried staff (mostly urban areas), 13 percent in the money were more purchasing of stationery (mostly urban), 12 percent paid towards the topping common in urban than up of the salaries of the non-teaching staff (mostly in urban areas), and 12 in rural and isolated percent towards the topping-up of salaries for normal teachers (again urban). areas. The kinds of parent activity where very few helped the school were: payment of salary of additional teachers (but when they did this it was mostly in urban schools), providing land for teachers to farm (obviously mostly isolated), and the provision of school meals (but where this occurred it was predominantly in urban areas and associated with schools that practice full-day schooling). In Figure 5.14 the main differences between the isolated, rural, and urban areas have been shown. Figure 5.14: Types of parental contributions to school by school location In general it was in urban areas that parents contributed money or time. It was in the isolated areas that parents contributed land and their time for construction purposes. In general, the Mekong Delta region was below the other regions. The contribution towards the payment of non-teaching staff was positively associated with pupil achievement but this was mostly in urban areas where the pupil achievement scores were higher. 190 Vietnam Reading and Mathematics Assessment Study Table 5.22: Percentages of pupils in schools with different contributions from communities 191 Vietnam Reading and Mathematics Assessment Study Figure 5.15: Parental contributions by regions and by school location For this particular combination of parental contributions it can be seen that the Red River Delta, North Central and Central Coast regions were the highest and the Mekong Delta was the lowest. In general more of the contributions (because many of them were money contributions) were from urban areas. This was to be expected. Parents helped the school to a certain extent. They helped in kind in isolated areas and in cash in urban areas. Education is a collaborative effort between home and school. When the school and home collaborate, this has been shown to result in better achievement, fewer pupil behavioural problems, and in the long run less crime in the society. Collaboration involves the setting of standards, whether it be for behaviour, helping the children learn, parents helping the teachers with school tasks, and to a much lesser extent donations in cash or kind (Coleman, ibid). Given that the home-school relationship is so important, it would seem to be desirable for further study to be conducted to see how this can best be enhanced in the Vietnamese context. Policy suggestion 5.11: The MOET might wish to consider establishing a study group to examine which action might be taken for school-home relationship. Relationships of school variables with pupil achievement The results so far presented in the chapter have described the characteristics of schools and school heads and what school heads do. 192 Vietnam Reading and Mathematics Assessment Study However, what is the relationship of these characteristics and activities with pupil achievement in reading and mathematics. Table 5.23 has presented the simple correlations (r)and partial correlations (rp) with most of the variables presented in this chapter and pupil achievement. The correlations have been calculated at the pupil level. That is the school variables were disaggregated to the pupil level and attached to the pupils in the sample in the school. It is to be noted that it is the sums of various variables that have been correlated with achievement since it is expected that such correlations will more stable than the individual items making up a sum. There were very few variables that had a sizeable simple correlation and also a large partial correlation. Only the variables marked in red will be taken through to the next stage of analysis in the penultimate chapter of this report. There were relatively few variables that were correlated to any significant extent with pupil achievement. They were: school head having received management training, school location the amount of teacher training teachers had received, pupils being in full day school, the amount of school resources, parental contributions, and finally, in a negative way, behavioural problems at school. Although some of the other variables will not be taken forward for more analysis, the production of a school magazine should be noted. Conclusion This chapter was about the school heads, the overall staff in the school, full inspections, the condition of the school buildings, problems experienced by the head with pupils and teachers, school resources and parental contributions to the school. For the most part, it could be seen that the primary schools were in reasonably good shape. Only the findings that prompted possible action on the part of the MOET and through that body the provincial and district levels of educational administration were highlighted. The following were the main points where it was felt that action might be undertaken. Some of the findings highlighted inequities among schools and provinces and regions. Others highlighted low levels of provision. Yet other findings highlighted problems in schools as perceived by the school heads. These have been resumed below. In some cases it is the MOET that is responsible. In other cases it is the lower levels of educational administration but it is for the MOET to ensure that action is forthcoming from the lower levels. Whereas in Chapter 4 reference was made to the need for a more equitable allocation of teachers to Grade 5 in schools, the analyses in this chapter reinforced this need to have a more equitable allocation of teachers to the whole school, not just Grade 5. The analyses also revealed a need to have a 193 Vietnam Reading and Mathematics Assessment Study Table 5.23: Correlations, partial correlations and sampling errors of school variables and pupil achievement. Reading Mathematics School variables R pr r pr Age of School Head 0.13 0.03 0.12 0 Sex of School Head 0.12 0.08 0.1 0.03 School Head's travel time to School -0.06 0.07 -0.05 0.05 School Head's marital status 0 -0.04 0.03 0.02 School Head's No. of children -0.02 -0.04 -0.02 -0.02 School Head's ethnic affiliation 0.1 -0.03 0.09 -0.02 School Head's academic qualification 0.06 -0.03 0.06 -0.03 School Head's professional training 0.1 -0.02 0.11 0.02 School Head received management training 0.07 0.08 0.09 0.1 School Head's years teaching experience 0.14 0.04 0.13 0.02 School Head's hours per week teaching now 0.02 0.03 0.04 0.08 Years experience as head 0.07 0.04 0.05 -0.01 School Head's possessions at home 0.19 0.08 0.19 0.07 School Head's satisfaction with living accommodation 0.02 0.05 0.03 0.05 Type of school 0.03 0.02 0.03 0.02 Age of School 0.02 -0.06 0.02 -0.06 Average distance to public services -0.18 -0.06 -0.16 -0.05 Urban/rural (to be scaled) 0.22 0.12 0.19 0.11 Percentage female teachers 0.24 0.07 0.25 0.1 Staff education 0.04 0.04 0.04 0.03 Staff teacher training 0.23 0.17 0.26 0.17 Pupil/Staff ratio 0.04 -0.01 0.01 -0.06 Total enrolment 0.07 0.06 0.05 0.01 Percentage boys in Grade 5 -0.01 0.06 -0.01 0.06 N.classes in Grade 5 0.06 0.03 0.04 -0.01 Percentage shifts 1 pupil -0.19 -0.06 -0.19 -0.08 Percentage shifts 2 pupil -0.15 -0.08 -0.15 -0.06 Percentage shifts 3 pupil -0.02 -0.07 -0.04 -0.08 Percentage shifts 4 pupil 0.21 0.08 0.22 0.09 With satellites -0.17 -0.07 -0.16 -0.02 Time inspected since 1999 0.05 0 0.03 -0.01 School heads' activities (sum) 0 0.05 -0.01 0.01 Production of school magazine 0.09 0.05 0.11 0.08 Story telling competition 0.11 -0.02 0.11 0.01 General knowledge quizzes 0.08 0 0.06 -0.03 Camping expeditions 0.02 0.04 0.03 0.03 School day celebrations 0.02 -0.01 0.03 -0.03 Open meeting of sch.Council 0.01 0.01 0.02 0.05 Pupil problems (total) -0.11 -0.1 -0.13 -0.12 Pupil frequent problems -0.12 -0.1 -0.14 -0.13 Pupil non-frequent problems -0.08 -0.07 -0.09 -0.09 Teacher problems -0.11 -0.09 -0.12 -0.1 Teacher frequent problems -0.12 -0.09 -0.12 -0.1 Teacher non-frequent problems -0.08 -0.07 -0.09 -0.08 Days lost per year -0.12 -0.02 -0.12 -0.06 Square metres per pupil 0.12 0.07 0.12 0.08 Condition of school buildings 0.08 0.05 0.08 0.06 School resources 0.26 0.14 0.25 0.12 Community's contributions 0.18 0.09 0.18 0.08 Age of school 0.02 0.02 0.02 0.02 Total enrolment 0.07 0.01 0.05 -0.02 Pupil-staff ratio 0.04 -0.04 0.01 -0.07 194 Vietnam Reading and Mathematics Assessment Study more equitable allocation of heads to schools in terms of their education, The school variables teacher training and attendance at management courses. that were related to Small studies should be undertaken on the social development of children pupil achievement after and examine the relations ship of such development to many of the factors taking account of the taken up in the present study. A special study should be undertaken to pupils' home discover the nature of the health problem that heads said featured so backgrounds were frequently among pupils and teachers and which were related to absenteeism identified. and drop out. Health was clearly a matter of concern to the heads. But there were also other behavioural problems and these should be examined in more detail. A small study should be initiated in satellite schools in isolated areas to identify the special needs of the satellite schools, easy access to them for the pupils, and in general to identify any problems associated with travel to school and reasons for the number of days lost in a school year. At the same time, the actual condition of school buildings should be described by the Public Works department (the department responsible for school buildings) because it was clear that heads thought that there many deficiencies in the present state of the buildings. How the inspectors carried out their tasks seem to be very varied and it was suggested that the MOET should review the current roles of the inspectorate and then state clearly what the roles are and make these clear to all concerned. A small monitoring unit to ensure high quality inspections would be desirable. The Curriculum Centre should decide on the most desirable reading activities that schools can carry out as extra curricular activities or even as school activities in order to create a culture of reading within each school and each school community. These ideas should be included in all teacher training programs and in the management course for teachers. The inspectorate too should encourage this as much as possible. The book flood program mentioned in Chapter 3 would also be of use. Related to this suggestion is that of the overall relationships between schools and homes in order to foster good education. This is particularly relevant in those schools where the head perceived there to be several behavioural problems. Many of these problems require home and community action as well as school action. It is here that the notion of school intervention programs mentioned in the previous chapters can be so fruitful. Finally, it was seen that there were great disparities in the resources that schools had. Schools with pupils with wealthier parents had more resources but even when home background had been accounted for there was still a fairly strong relationship between school resources and achievement. The list of resources used in this study was developed by the study's questionnaire committee and it may well be that others are of a different opinion as to which group of resource items forms what must be considered as essential. Any combination of resource items (but only from the items for which data are available) can be examined and should be. But, it still remains a fact that there are disparities among schools for resources. It is the isolated schools that for the most part are most in need. The MOET could develop a plan to identify the missing resources and then develop a timetable to ensure that the resource items are provided to the needy schools first and eventually to all schools. 195 Vietnam Reading and Mathematics Assessment Study Appendix 5.1: Means and sampling errors of school resources by region and school location hall school room room ground ater for kit Region School recorder /staff room aid medical location community office head water/W apeT or Store First area/Play eacherT Piped tank/Borehole/Spring School Separate Separate Sports % SE % SE % SE % SE % SE % SE % SE % SE % SE Isolated 85 19 100 0 28 35 43 47 . . 43 47 85 19 100 0 58 40 Red River Rural 67 2 94 1 29 2 81 2 14 1 39 2 88 2 78 2 79 2 Delta Urban 80 5 88 4 31 5 70 4 26 5 73 5 95 2 65 5 94 3 Total 70 2 93 1 29 2 78 1 16 2 44 2 89 2 76 2 81 1 Isolated 42 5 82 3 62 5 53 4 9 3 10 4 27 4 48 5 44 5 Rural 54 3 85 1 40 3 72 2 12 2 23 2 60 2 70 2 54 3 Northeast Urban 65 5 85 4 37 6 77 3 23 4 49 5 87 4 68 4 82 4 Total 54 2 85 1 43 2 70 2 13 1 25 2 59 2 66 2 57 2 Isolated 22 6 62 8 68 7 15 4 2 2 . . 9 4 63 8 35 7 Rural 49 6 76 5 19 4 36 5 4 2 11 5 26 5 66 5 47 5 Northwest Urban 79 9 80 10 35 9 78 9 12 6 44 10 70 9 81 7 82 8 Total 44 4 72 4 39 4 35 4 4 2 12 3 27 4 67 4 48 4 Isolated 73 8 93 3 59 9 50 9 9 5 4 3 15 6 60 10 62 8 Rural 74 3 93 2 30 4 74 4 22 4 17 3 69 4 78 3 80 3 North Central Urban 92 3 87 4 61 8 91 5 27 7 55 8 98 1 90 4 99 1 Total 76 3 92 2 37 3 74 3 22 3 21 3 68 3 78 3 81 2 Isolated 47 9 57 10 18 6 36 8 5 3 7 5 50 9 23 6 51 9 Rural 57 3 57 4 23 3 49 3 17 3 15 2 77 3 56 4 72 3 Central Coast Urban 90 3 73 5 44 5 75 4 34 5 65 6 97 2 66 5 94 2 Total 64 3 61 3 28 2 54 3 20 2 27 3 79 2 55 3 75 3 Isolated 33 9 52 10 18 7 14 4 . . . . 4 4 28 9 28 9 Central Rural 64 7 65 6 21 5 33 6 7 4 7 4 28 6 50 7 55 7 Highlands Urban 80 7 73 7 21 7 73 7 24 8 27 8 69 8 77 7 79 7 Total 62 5 65 5 20 4 40 4 10 3 11 3 34 4 53 5 55 4 Isolated 69 7 58 6 10 5 20 6 3 3 3 3 64 5 39 8 54 6 Rural 89 2 46 3 19 3 27 3 9 3 10 2 87 2 55 3 72 3 Southeast Urban 94 2 54 4 51 5 51 5 26 4 54 4 96 1 58 4 94 1 Total 89 1 51 2 31 3 36 2 15 2 27 2 88 1 54 3 79 2 Isolated 62 5 24 4 16 3 13 3 3 2 2 1 37 4 29 4 66 4 Rural 73 2 29 2 13 1 15 1 7 1 8 1 51 2 44 3 72 2 Mekong Delta Urban 93 2 44 5 41 4 35 5 12 2 41 5 82 4 62 4 95 2 Total 75 2 31 2 18 2 18 1 7 1 12 2 54 2 45 2 75 1 Isolated 54 3 59 2 35 2 30 2 5 1 5 1 33 2 42 3 52 2 Rural 68 1 70 1 26 1 54 1 13 1 19 1 68 1 64 1 71 1 Vietnam Urban 87 1 68 2 43 2 63 2 24 2 53 2 91 1 67 2 92 1 Total 70 1 68 1 30 1 53 1 14 1 25 1 69 1 62 1 73 1 196 Vietnam Reading and Mathematics Assessment Study or water (mains Region School garden machine recorder location generator) elephoneT Drinking Fax ypewriterT Duplicator equipment School apeT Electricity Microphone/amplifier % SE % SE % SE % SE % SE % SE % SE % SE % SE Isolated 85 19 100 0 100 0 . . 43 47 . . . . 100 0 . . Rural 74 2 95 1 86 1 1 1 60 2 1 1 3 1 90 2 13 2 Red River Delta Urban 95 2 91 3 97 2 5 2 48 6 5 2 4 2 87 5 18 4 Total 78 2 94 1 88 1 2 1 58 2 2 1 3 1 89 1 13 1 Isolated 41 5 46 5 19 4 . . 44 4 1 1 1 1 41 5 1 1 Rural 58 2 81 2 58 2 0 0 50 3 1 1 2 1 75 2 2 1 Northeast Urban 80 4 97 2 95 2 1 1 36 5 3 2 3 2 88 3 0 0 Total 59 2 78 1 58 2 0 0 47 2 1 0 2 1 72 2 2 1 Isolated 51 8 31 7 6 3 . . 25 6 5 3 . . 27 7 . . Rural 35 6 58 5 33 6 . . 45 6 . . . . 60 6 7 3 Northwest Urban 78 8 96 4 90 7 . . 40 9 5 4 6 6 88 6 4 4 Total 47 5 54 4 32 4 . . 37 4 2 1 1 1 53 4 4 2 Isolated 54 9 49 10 24 6 . . 67 8 2 2 4 2 40 10 13 7 Rural 72 4 94 2 67 4 . . 64 3 4 1 2 1 79 3 11 2 North Central Urban 89 5 100 0 100 0 6 5 58 8 16 6 3 3 95 2 22 8 Total 73 3 91 2 67 3 1 1 63 3 6 1 2 1 78 3 13 2 Isolated 52 10 75 8 82 6 . . 17 6 57 8 3 3 51 8 15 6 Rural 62 4 87 2 90 2 0 0 25 4 66 3 0 0 77 3 19 3 Central Coast Urban 84 4 97 2 98 2 1 1 20 4 69 4 2 1 78 6 24 5 Total 67 3 88 2 91 2 0 0 23 2 66 2 1 1 74 2 20 3 Isolated 23 8 43 10 21 8 . . 31 8 2 1 . . 52 10 3 2 Rural 60 7 70 6 70 7 . . 27 6 12 5 1 1 53 6 10 4 Central Highlands Urban 74 7 95 4 97 2 . . 24 7 14 5 6 4 66 8 15 6 Total 56 4 71 4 66 4 . . 27 4 11 3 2 1 56 4 10 3 Isolated 59 7 80 5 67 7 1 1 10 3 52 7 4 4 58 7 10 4 Rural 60 4 91 2 91 2 1 1 19 3 62 3 1 1 56 3 19 3 Southeast Urban 85 3 94 3 99 1 3 1 18 4 79 3 6 2 67 4 35 4 Total 70 3 91 1 92 1 2 1 18 2 68 2 4 1 61 2 25 2 Isolated 55 4 79 4 78 4 . . 1 1 75 4 . . 24 5 3 2 Rural 63 2 77 2 81 2 . . 7 1 67 2 0 0 35 2 14 2 Mekong Delta Urban 78 5 95 2 100 0 . . 13 3 91 2 5 2 53 5 32 4 Total 64 2 80 2 84 1 . . 7 1 72 2 1 0 37 2 16 2 Isolated 50 2 62 3 47 3 0 0 26 2 33 2 2 1 41 3 6 1 Rural 65 1 86 1 76 1 0 0 40 1 27 1 1 0 68 1 12 1 Vietnam Urban 84 1 95 1 98 1 2 1 29 2 48 2 5 1 74 2 24 1 Total 67 1 85 1 77 1 1 0 36 1 32 1 2 0 66 1 14 1 197 Vietnam Reading and Mathematics Assessment Study school recorder Projector around Region School set equipment equipment location TV cassette Photocopier Computer borders hedge Cafeteria Overhead or Music ideoV Sports Fence % SE % SE % SE % SE % SE % SE % SE % SE % SE Isolated . . 85 19 85 19 . . 43 47 58 40 43 47 43 47 70 30 Rural 9 1 41 2 33 2 2 1 7 1 81 2 4 1 53 2 62 2 Red River Delta Urban 42 5 77 5 67 6 23 5 55 5 94 3 29 5 92 3 78 5 Total 14 1 48 2 39 2 5 1 16 1 83 2 9 1 60 2 65 2 Isolated . . 39 4 7 3 . . . . 57 6 . . 17 4 36 5 Rural 2 1 30 2 10 1 0 0 0 0 71 2 2 1 28 2 47 2 Northeast Urban 9 3 44 5 19 4 2 1 5 2 86 3 21 4 61 4 54 4 Total 2 1 34 2 11 1 0 0 1 0 71 2 5 1 31 2 46 2 Isolated . . 20 7 6 3 . . . . 54 8 . . 7 3 30 7 Rural . . 26 5 5 2 2 2 3 2 68 6 . . 21 5 46 6 Northwest Urban 11 6 65 10 34 10 3 3 21 9 87 7 35 11 64 10 73 9 Total 2 1 30 4 10 2 1 1 4 2 66 4 5 2 23 3 45 4 Isolated . . 37 9 9 5 . . . . 72 9 8 6 18 6 35 9 Rural 3 2 30 4 8 2 . . 3 1 81 3 1 1 32 4 61 4 North Central Urban 9 4 48 9 29 8 . . 20 6 90 5 26 7 66 8 88 5 Total 4 1 33 3 11 2 . . 5 1 82 2 5 1 36 3 63 3 Isolated . . 35 9 24 9 . . . . 40 7 4 3 8 4 25 7 Rural 2 1 32 4 26 4 0 0 10 2 65 4 3 1 20 3 38 4 Central Coast Urban 11 4 76 5 70 5 12 3 27 3 90 4 30 5 64 6 63 6 Total 4 1 44 3 37 3 3 1 13 2 68 3 10 1 29 3 43 3 Isolated . . 24 8 6 3 . . . . 30 10 . . 4 2 11 5 Rural . . 35 6 7 3 5 3 10 4 66 6 4 3 14 4 38 7 Central Highlands Urban 1 1 52 8 33 8 4 4 22 7 81 6 9 5 34 8 48 9 Total 0 0 37 4 14 3 4 2 11 3 62 4 5 2 17 4 35 5 Isolated 18 6 45 7 31 7 4 4 27 8 68 5 11 5 15 6 26 7 Rural 17 2 46 3 35 3 6 2 37 3 74 3 17 3 33 3 35 4 Southeast Urban 34 5 84 2 80 3 54 4 70 4 94 2 60 3 63 4 61 4 Total 24 2 62 2 53 2 26 2 49 2 82 2 34 2 44 2 45 2 Isolated . . 24 5 5 2 . . 6 3 43 4 9 3 7 2 9 3 Rural 1 0 10 1 6 1 0 0 5 1 60 2 12 2 13 2 20 2 Mekong Delta Urban 9 2 50 5 40 6 5 2 31 4 85 3 50 4 46 5 41 5 Total 2 1 19 1 12 1 1 0 10 1 62 2 18 2 18 1 22 2 Isolated 3 1 34 3 13 2 1 1 6 2 53 3 6 2 12 2 25 2 Rural 5 0 30 1 18 1 1 0 8 1 72 1 6 1 30 1 45 1 Vietnam Urban 21 2 66 2 54 2 22 2 40 2 90 1 39 2 63 2 62 2 Total 8 0 38 1 25 1 6 0 15 1 74 1 13 1 35 1 46 1 198 Vietnam Reading and Mathematics Assessment Study Chapter 6 HOW DO THE CONDITIONS OF SCHOOLING IN VIETNAM COMPARE WITH THE MINISTRY'S OWN BENCHMARK STANDARDS? Introduction I n this chapter the discussion of schooling conditions has been extended Benchmarks are beyond the descriptions given in Chapters 4 and 5, to a comparative standards set by an analysis in which the conditions have been compared with the external body that benchmark standards laid down by the Vietnam Ministry of Education schools are meant to and training. This comparative analysis has permitted judgements to be made attain. In this study, two about key aspects of the educational environment in relation to minimal sets of benchmarks were levels of provision that the Ministry has established for the proper used: those set by functioning of the primary schools. In some cases, it was difficult to establish MOET and those set by the ministry benchmarks. In this case, the researchers used either the the new fundamental benchmarks in the Fundamental School Quality Levels document from the school quality level 'Disadvantaged children project', or commonly accepted benchmarks were body. used. Some of the benchmark standards were established long ago. At the same In some cases, the time, there were no benchmark standards available for some key aspects of ministry may wish to schooling such as maximum and minimum total enrolments or per pupil revisit and revise some squared meter in the classroom. In other cases, information was available but benchmarks. no justification for the standards could be found. There would appear to be a need for the Ministry to re-visit all benchmarks standards in order to check their relevance for the first decade of the 21st century, and also to take action to fill those gaps where standards do not yet exist. This list should be readily available for all, including researchers. Policy suggestion 6.1: The Ministry might wish to review and, if necessary, to revise benchmark standards for the educational environment that are deemed to be reasonable for the proper functioning of primary schools. 199 Vietnam Reading and Mathematics Assessment Study Basic infrastructure Ten variables were selected for features of the basic infrastructure of the school. The results have been presented in Table 6.1. All classrooms should (a) Blackboard: It is clear that without a useable blackboard that it is have a blackboard, but difficult to teach pupils. Information was sought about whether a 6% of pupils were in blackboard was in the classroom or not but there was no information on classroom without a the quality of the blackboard. Some may have been shiny or very worn. useable blackboard. From Table 6.1 it can be seen that nearly six percent of pupils were in classrooms without a useable blackboard. The percentage was somewhat higher for pupils in isolated areas. Most pupils were in (b) Cupboard: In the basic list of fundamental school quality levels classrooms without document it is stated that there should be storage lockers or boxes for cupboard. instructional materials. No question was asked about such materials but a question was asked about the existence of a cupboard. It can be seen that 93 percent of pupils were in classrooms without cupboards but that in urban areas this figures was only 83 percent. In urban schools in the Red River Delta only 46 percent of pupils were in schools where there were cupboards in the classrooms and in the Central Coast region the figure was 78 percent. Only 2.8% of pupils (c) School resources: In Chapter 5, it was seen that a list of 27 resources were in schools that was used to establish which resources existed. Based on the information possessed 12 prescribed in the Document of current law on education and training vol.1: items. regulations about schools1 there were 12 items out of the 27 that were obligatory in all schools and these therefore constituted the benchmark. The items were school library, school hall, staff room, office for school head, store room, first aid kit, medical room, playground, piped water, fence around school, music equipment, and sports equipment. Only 2.8 percent of pupils were in schools with this minimum number of items. In urban areas the figure was 7.2 percent. It is to be noted that in some school locations in some regions there were zero pupils in schools with this minimum. This is a somewhat alarming state of affairs and it is suggested that the MOET undertake an audit in all schools but starting in those areas where there were no pupils having the minimum. ------------------------ 1 MOET, 2001 Cac van ban phap luat hien hanh ve Giao duc-Dao Tao 1. Cac quy dinh ve nha truong 200 Vietnam Reading and Mathematics Assessment Study (d) Classroom resources: In Chapter 4 information was presented on the 80% of pupils were in number of items of teaching materials (supplies) and furniture in the classrooms meeting the classroom. In all there were 16 items. Again, according to the Document ministry's benchmarks of current law on education and training, eight items were compulsory for classroom resources. in all classrooms and these items constituted the benchmark. These eight items were writing board, chalk, teacher table, teacher chair, geometrical instruments, teachers' guide (Vietnamese), teachers' guide (math), and pictures as teaching illustrations. It can be seen that 80 percent of pupils in Vietnam were in the classrooms with these minimum resources. Pupils in the isolated areas tended to be in the classrooms which did not satisfy the minimum standard of resources. Especially, in the Northwest, only 41 percent of pupils in isolated areas were in the classroom with minimum resources. (e) Sitting/writing place: It is reasonable to expect that each child should Only 4% of pupils in have a sitting and writing place at a desk. The benchmark was therefore classrooms with one sitting/writing place per pupil. It can be seen that four percent of insufficient sitting and pupils did not have such a place but this figure was over six percent in writing places. urban areas. It would be wise to make the provision of a proper sitting and writing place for all a priority issue. (f) Classroom space: One often-used benchmark for the over- 85% of pupils were in crowdedness of a classroom is 1.5 metres space per pupil. This was not classrooms where they found in the Vietnamese 'rule books' but has been used. In Table 6.1 it is did not have 1.5 square the percentage of pupils in classrooms with less than 1.5 metres that has meters per pupil. been reported. It can be seen that 85 percent of pupils were in the over crowded classrooms. Pupils in urban were more likely in the crowded classrooms. (g) Class size in all primary classes: From the school questionnaire, it 21% of pupils, mostly was possible to derive, somewhat roughly, the average class size of all coming from urban classes in the school. This was done by dividing the number of classes in areas, were in the school into the total enrolment. The bench mark of not more than 35 classrooms with more pupils per class was used. It can be seen that 21 percent of pupils were than 35 pupils. in classrooms that were larger than the benchmark. For the different school locations, the figures were eight percent for isolated areas, 13 percent for rural areas and 53 percent for urban areas. The urban areas in the regions of Hanoi, Central Coast and the Southeast had very high class sizes and there more than 70 percent in classes with more than 35 pupils. This seems to be very high but it must be recalled that the enrolment was probably at its highest at the beginning of the 21st century and with the declining birth-rate it is expected that the problem will become less urgent. 201 Vietnam Reading and Mathematics Assessment Study It would seem to be desirable to check in the most crowded areas (in terms of the percentage of pupils in classrooms with less than 1.5 square metres per classroom) and these are urban areas to see what might be done to alleviate this situation. And, the figure was 33% (h) Class size Grade 5: The information for the grade 5 class size is for Grade 5 classrooms. accurate and comes from the teacher questionnaire. Again the same benchmark was used of 35 pupils per class. In Table 6.1 it can be seen that 33 percent of pupils were in classrooms with more than 35 pupils per classroom. The figures for isolated, rural and urban areas were 17, 31, and 53 percent respectively. The same comments apply as were given in (g) above. 25% of pupils were in (i) Total number of class groups: The government set the benchmark of schools with more than 30 as maximum numbers of class groups per school. As seen in Table 6.1 30 class groups. below, 25 percent of pupils were in the school with too many class groups. The percentage was higher in urban areas. The ministry allows certain schools to have more than 30 class groups where there is urgent need to enrol more pupils. It is therefore to be expected that some school will exceed the benchmark. Using 35 pupils per (j) Total enrolment: No benchmarks were found for the total enrolment class * 30 class groups of a school. Based on the maximum number of pupils per class (35 as the maximum pupils per class) and the maximum numbers of classes per school (30 enrolment benchmark, classes per school) which MOET set as the standard, the benchmark of 21% of pupils were in the total enrolment of a school could be calculated. The total enrolment schools with more than 1050 pupils. of a school, therefore, should not exceed 1050. It can be seen that 21 percent pupils were in schools which exceeded the benchmark. It was particularly in urban areas that the classrooms were overcrowded. Again, since the ministry allows certain schools to exceed the benchmark, it is not surprising that some schools had high enrolment. 25% of pupils were in (k) Pupil-staff ratio: The pupil/staff ratio (as opposed to the pupil-teacher schools where the pupil- ratio) has been described in Chapter 5. It is the total enrolment of the staff ratio was higher school divided by the total number of full-time staff. The benchmark was than 30. set, again some what arbitrarily, at 30. In Table 6.1 it can be seen that 25 percent of pupils were in schools with a pupil-staff ratio of more than 30 pupils. Again the high figures for Central Coast and Southeast urban areas should be noted. 202 Vietnam Reading and Mathematics Assessment Study Table 6.1: Basic infrastructure Black School Classroom Sitting/wriClassroom G5 class Total class Total pupil-staff board Cupboad resource resource ting place space per Class size group per pupil size school enrolment ratio Region School location Do not Do not Have Have <1.5 >= 35 > 35 > 30 >30 have have minimum minimum 8 < 1 seat squared class >1050 pupils 12 items items mater pupils pupils groups pupils % SE % SE % SE % SE % SE % SE % SE % SE % SE % SE % SE Isolated 0 0 100 0 0 0 20 47 17 18 100 0 43 47 69 70 42 40 42 40 . . Red River Rural 3 1 90 2 4 1 87 1 5 1 89 2 18 2 46 2 16 2 13 2 13 1 Delta Urban 1 1 54 6 7 2 86 4 9 5 90 4 74 4 77 4 36 5 50 6 28 5 Total 3 1 84 2 5 1 86 2 6 1 89 2 28 2 52 2 20 2 20 2 15 1 Isolated 6 2 99 1 2 2 68 4 1 1 59 6 1 1 11 3 25 4 3 2 4 2 Rural 6 1 96 1 2 1 80 2 2 1 79 2 6 1 21 2 23 2 11 1 5 1 Northeast Urban 4 2 94 3 6 3 86 4 4 2 85 4 38 5 56 6 13 3 20 4 20 4 Total 6 1 96 1 3 1 79 2 2 1 76 2 10 1 25 1 22 2 11 1 7 1 Isolated 14 6 99 1 0 0 41 6 12 5 47 10 0 0 16 5 34 7 10 4 6 3 Rural 7 3 100 0 0 0 70 5 2 1 72 6 0 0 9 3 35 5 12 3 . . Northwest Urban 7 5 99 2 4 4 82 8 5 3 59 12 9 7 14 6 7 5 12 7 . . Total 10 3 99 0 1 1 61 4 6 2 62 4 1 1 13 3 30 4 12 2 2 1 Isolated 5 4 100 0 0 0 71 8 1 1 74 11 5 3 12 4 16 7 2 2 9 5 Rural 8 2 93 2 2 1 78 3 2 1 92 3 18 3 35 3 15 3 10 2 33 4 North Central Urban 2 1 86 4 15 6 87 4 2 1 84 9 40 7 50 7 18 7 34 8 15 4 Total 7 2 93 1 4 1 79 2 2 1 90 3 20 2 34 3 16 3 13 2 28 3 Isolated 4 2 99 1 0 0 69 9 8 4 77 9 17 6 21 6 15 7 16 6 23 7 Rural 6 1 98 1 2 1 86 2 4 1 86 4 25 2 43 4 28 3 26 3 39 3 Central Coast Urban 9 3 78 4 12 3 87 4 7 3 89 5 56 6 58 6 43 7 47 7 53 6 Total 7 1 94 1 4 1 83 2 5 1 85 3 31 3 43 3 30 3 30 3 41 2 Isolated 5 2 100 0 0 0 56 9 2 2 80 9 4 3 13 5 54 10 40 11 17 8 Central Rural 7 4 99 1 0 0 73 5 3 2 96 2 6 3 21 4 32 6 18 6 23 5 Highlands Urban 7 4 94 5 1 1 82 5 8 5 99 1 32 8 42 7 37 8 35 8 43 8 Total 7 2 98 1 0 0 70 4 3 1 94 2 13 3 23 3 38 4 27 4 27 4 Isolated 10 3 100 0 0 0 77 5 4 2 79 6 14 4 23 5 31 5 22 5 35 6 Rural 6 1 94 2 1 1 82 2 4 1 88 3 16 3 24 3 31 3 28 3 35 3 Southeast Urban 7 2 83 3 7 2 86 3 8 3 92 3 71 4 54 4 44 5 55 4 63 4 Total 7 1 91 1 3 1 83 2 5 1 89 2 38 3 35 2 36 2 39 2 47 3 Isolated 6 2 98 1 0 0 78 4 4 1 88 4 10 3 18 3 27 5 18 4 33 5 Rural 5 1 99 0 0 0 80 2 4 1 82 3 5 1 21 2 23 2 15 2 19 2 Mekong Delta Urban 7 2 95 2 4 2 82 3 5 2 80 4 31 5 37 5 45 5 43 5 45 5 Total 6 1 98 0 1 0 80 2 4 1 82 2 10 1 23 1 27 2 20 2 26 2 Isolated 7 1 99 0 0 0 68 2 4 1 74 3 8 2 17 2 28 2 15 2 19 2 Rural 6 1 95 1 2 0 81 1 3 0 86 1 13 1 31 1 22 1 15 1 21 1 Vietnam Urban 5 1 83 2 7 1 85 1 6 1 88 1 53 2 53 2 35 2 43 2 41 2 Total 6 1 93 1 3 0 80 1 4 0 85 1 21 1 33 1 25 1 21 1 25 1 203 Vietnam Reading and Mathematics Assessment Study School head and teacher qualifications In this section, the term "staff" has been used to describe all of the teachers in the school. Information about all of the teachers was provided by the school heads in the school questionnaire. Next to the variable name in the accompanying tables, "from Shquest" has been written. The word "teacher" in the variable headings has been used to refer to the Grade 5 teachers who completed the teacher questionnaire. The word " from Tbooklet" has been written next to the name of the variable. The result for school heads and teachers who had 12+2 years of training have been presented in Table 6.2. The results for school heads and teachers who had 9+3 years of training have been presented in Table 6.3. When reviewing the two types of teacher training, it must be recalled that until 1983 nearly all teachers were trained under the 9+3 system. After 1983, teachers were trained under the 12+2 system. Hence, it is the older teachers who will have been under the 9+3 system. 10% of pupils were in (a) School heads' academic education: There is no benchmark for the years schools where the school of academic education for school heads. However, all school heads much heads were not have been teachers, and the benchmark for teachers is that they should have completed upper completed at least upper secondary school. It can be seen from Table 6.2 that secondary. 10 percent of pupils were in schools with school heads who had not completed upper secondary school. However, according to the initial version of FSQL, minimum academic education all heads should have completed at least lower secondary school. The percentage of pupils in schools where the heads did not have this qualification have been presented in Table 6.3 and it can be seen that there were zero percent meaning that all pupils were in schools where heads satisfied the benchmark. More than 90% of pupils (b) School heads' teacher training: The MOET set the benchmark that school were in schools where heads in isolated areas should have completed 9+3 and school heads in rural the school heads did and urban should have completed 12+2. It can be seen from Table 6.2 that 12 meet the benchmark of percent of pupils in urban and 18 percent of pupils in rural areas were in completing 9+3 teacher schools with school heads who had not completed 12+3. It can also be seen training. teacher from Table 6.3 that seven percent of pupils in isolated were in schools with training. school heads who had not completed the 9+3 training. On the other hand, according to FSQL, all heads should have completed the teacher training known as 9+3. In Table 6.3, it is the percentage of pupils in schools where the heads did not satisfy the benchmark that have been presented. Just over three percent of pupils were in schools where the heads had not satisfied this particular benchmark. However, isolated schools in the Northeast, Northwest and Mekong Delta as well as rural schools in the Central Highlands give rise for concern. In time this problem will be resolved. 204 Vietnam Reading and Mathematics Assessment Study (c) Grade 5 teacher education: According to the educational regulation in 10% of pupils were in Vietnam, all teachers should have completed at least upper secondary school where Grade 5 school. It can be seen that almost 10 percent of pupils were with teachers teachers did not meet who had not completed secondary school. Pupils in isolated schools the benchmark for teacher education. tended to have teachers who had not completed secondary school compared with those in rural and urban areas. Yet again, FSQL set the benchmark that all teachers should have completed at least lower secondary school. As can be seen from Table 6.3 that there were very few pupils in schools where the teachers had not done this. (d) Teaching staff education: The percentages of pupils have been 18% of pupils were in presented in schools where the head said that at least one member of the schools where at least teaching staff had not completed lower secondary education. This figure one member of the was larger than for the Grade 5 teachers only. The overall figure for teaching staff did not Vietnam was 18 percent and in isolated areas it was 25 percent. The largest meet the ministry's benchmark for teacher percentages were in the Northwest, especially in the isolated areas. If the education. benchmark were to be set as having completed upper secondary school, then the percentage of pupils who were in schools with school at least one member of staff had not completed upper secondary would be 72. (e) Grade 5 teacher training: Similarly to the benchmark for school heads, 21% of pupils were in the MOET set the benchmark that teachers in isolated areas should have school where Grade 5 completed 9+3 and that teachers in rural and urban areas should have teachers did not meet completed 12+2. As can be seen from Table 6.2, 19 percent of pupils in the ministry's benchmark urban areas and 20 percent of pupils in rural areas were in schools with for teacher training, school heads who had not completed 12+3. It can be seen from Table 6.3 that 5 percent of pupils in isolated were in schools with school heads who had not completed 9+3. Again, FSQL set a different benchmark, namely that all teachers should have had the 9+3 teacher training. Only 3.5 percent of pupils were in schools where Grade 5 teachers had not had at least this kind of training. The Central Highlands was somewhat high and it might be worth some further investigation. (f) Primary school teaching staff teacher training: The same benchmark 50% of pupils were in was used as for (e) above. However, this time it was the percentages of schools where at least pupils in schools where the school head had said that at least one member member of the teaching of teaching staff had not received this training that have been reported. staff did not meet the When the benchmark was set at 9+3, 50 percent of pupils were in schools ministry's benchmark for teacher training. where at least one member of staff did not reach the benchmark. Again there was a small problem in the Central Highlands. When the benchmark was set at 12+2, the percentage went up to 80. 205 Vietnam Reading and Mathematics Assessment Study Table 6.2: School head and teacher qualifications (12+2/Upper Secondary) School with at least School with at least one teaching one teaching staff SH academic SH teacher Grade 5 teacher Grade 5 teacher academic education staff member member not having not having 12+2 professional training prescribed Region School education training (from Tbooklet) (from Tbooklet) location professional training academic education (from Shquest) (from Shquest) < upper secondary < 12+2 < upper secondary < 12+2 < 12+2 < upper secondary % SE % SE % SE % SE % SE % SE Isolated 70.5 30 43 46.7 . . 100 0 . . 100 0 Red River Rural 10.1 1.5 9.6 1.5 3.4 0.7 66.4 2.4 2.9 0.7 64.7 2.2 Delta Urban 7.3 3 3.5 2.2 6.2 2.7 51 5.6 4.1 1.7 61.6 5.1 Total 10.3 1.3 9 1.4 3.8 0.7 64.2 2.4 3 0.7 64.6 2 Isolated 26.2 4.4 24.5 3.7 14.7 2.4 87 3.7 20.7 2.5 86.2 4.7 Rural 21.3 2.1 12.6 1.5 12.6 1.2 64.4 2.4 11 1 71.7 2.4 Northeast Urban 13.8 3.3 6.1 2.2 16.7 3.2 47.2 5.2 8.9 2 57.7 5.8 Total 20.9 1.4 13.5 1.2 13.7 1 65.3 1.9 12.5 0.9 71.8 1.8 Isolated 33.3 7.4 30.8 7.1 42.2 6.6 100 0 48.4 5.8 97.9 2.2 Rural 29.6 5.5 25.3 5.3 23.8 3.1 85.5 3.9 23.1 3.2 82.7 4.5 Northwest Urban 20.1 6.8 15.6 5.8 22.7 6.9 77.7 8 21.9 6.9 70.3 9.2 Total 29.5 3.7 25.8 3.5 30.7 2.8 89.4 2.5 32.7 2.8 86.2 2.7 Isolated 12.1 6.7 4.9 3.9 7.8 4.1 90.3 5 16.9 4.5 87.5 4.4 North Rural 14 2.7 10.7 2.3 5.5 1.5 79.1 2.9 6.9 1.6 74.7 3 Central Urban 7 4.9 5.8 3.9 1.3 1.4 35.7 7.6 4.1 2.1 47.1 9.2 Total 12.8 2.1 9.4 1.9 5.2 1.2 73.9 2.7 7.7 1.4 71.9 2.6 Isolated 4.5 4.6 17.9 6.8 11.5 4.3 82.2 6.1 18.8 4.9 57 8.5 Central Rural 0.2 0.2 8.3 2 2.1 0.7 72.7 2.6 12 1.7 58.7 3.9 Coast Urban 1.6 1.7 6.2 2.7 3.4 1.6 68.3 5.1 10.2 2.5 49.2 5.3 Total 1.1 0.7 9 1.6 3.8 1 72.8 1.6 12.7 1.3 56.1 2.7 Isolated 8.2 3.4 46.1 9.9 21.8 5.9 100 0 45.1 7.3 86.2 7.2 Central Rural 9.3 3.7 41.5 6.7 9.9 3 99.1 0.9 46.7 4.6 90.9 3.3 Highlands Urban . . 13.9 6.2 11.2 3.6 100 0 37.9 8.4 84.3 5.7 Total 6.5 2 34.9 4.3 13.3 2.3 99.6 0.5 44.4 3.4 88.1 2.7 Isolated 13.5 4.6 48.2 7.6 15.3 3.8 97.4 2.6 37.7 6.2 80.8 6.1 Rural 3 0.9 28.7 3.3 7 1 98.3 0.9 36 2.2 73 3.1 Southeast Urban 0.4 0.4 19.4 3.9 3.1 0.9 95.2 2 29 3.3 68.8 4.4 Total 3.2 0.7 27.3 2.6 6.9 0.8 96.9 0.9 33.8 1.6 72.2 2.4 Isolated 11.2 2.9 47.2 5.6 18.5 3 96.5 2 44.8 4.2 88.6 2.9 Mekong Rural 6.1 1 33.3 2.1 14.5 1.6 92.2 0.9 42.5 2 77.6 1.8 Delta Urban 3.5 1.7 14 2.8 11 2.5 92.3 2.2 36.2 4.3 76.5 3.2 Total 6.4 0.9 32.1 1.7 14.6 1.2 92.9 0.7 42 1.6 79.1 1.5 Isolated 16.8 1.9 33.5 2.5 17.2 1.4 93.4 1.2 32.2 2 84.9 2 Rural 10.6 0.7 18.4 0.8 8.6 0.5 78.9 1 19.5 0.8 71.9 1 Vietnam Urban 4.6 0.8 11.5 1.4 7.5 1 73.6 1.8 19.4 1.4 64.4 2 Total 10.1 0.5 18.8 0.7 9.6 0.4 79.5 0.8 21.3 0.6 71.9 0.8 206 Vietnam Reading and Mathematics Assessment Study Table 6.3: School head and teacher qualifications (9+3/Upper Secondary) School with at least School with at least Grade 5 teacher one teaching Grade 5 teacher one teaching SH academic SH teacher academic staff member professional staff member education training education not having prescribed training (from not having 9+3 academic education Tbooklet) professional training Region School (from Tbooklet) location (from Shquest) (from Shquest) < lower Teacher < lower School with teacher School with secondary < 9+3 secondary < lower secondary Teacher < 9+3 teacher < 9+3 % SE % SE % SE % SE % SE % SE Isolated 0.0 0.0 0.0 0.0 0.0 0.0 N.A. N.A. 0.0 0.0 N.A. N.A. Red River Rural 0.0 0.0 1.5 0.7 0.1 0.1 21.0 1.9 0.7 0.3 39.4 2.5 Delta Urban 0.0 0.0 2.7 2.0 1.1 1.1 27.4 5.0 1.7 1.2 22.2 4.6 Total 0.0 0.0 1.7 0.6 0.2 0.2 22.9 2.0 0.8 0.3 37.0 2.4 Isolated 0.0 0.0 8.2 2.4 0.0 0.0 38.1 3.8 5.5 1.7 66.3 5.3 Rural 0.0 0.0 1.9 0.6 0.0 0.0 23.4 1.8 1.5 0.4 33.9 2.1 Northeast Urban 0.0 0.0 2.0 1.2 0.5 0.5 19.2 4.7 1.7 1.1 24.4 4.6 Total 0.0 0.0 2.9 0.5 0.1 0.1 25.1 1.4 2.3 0.4 37.6 1.6 Isolated 0.0 0.0 9.2 5.3 0.5 0.5 60.4 6.8 5.4 2.2 80.7 5.2 Rural 0.0 0.0 4.2 1.9 0.6 0.6 39.6 5.9 2.1 1.6 52.4 5.3 Northwest Urban 0.0 0.0 0.0 0.0 0.0 0.0 24.2 9.4 0.0 0.0 30.6 9.2 Total 0.0 0.0 5.3 2.1 0.5 0.3 44.6 4.2 3.0 1.2 59.1 3.3 Isolated 0.0 0.0 1.2 1.2 0.0 0.0 18.6 8.1 0.6 0.6 42.3 8.8 North Rural 0.0 0.0 0.0 0.0 0.1 0.1 22.3 2.8 1.7 0.7 42.5 3.4 Central Urban 0.0 0.0 0.0 0.0 0.0 0.0 21.8 8.0 0.0 0.0 5.4 2.4 Total 0.0 0.0 0.1 0.1 0.1 0.1 21.9 2.5 1.3 0.6 37.1 3.0 Isolated 0.0 0.0 6.3 4.8 0.0 0.0 10.2 6.5 5.0 3.0 54.2 8.1 Central Rural 0.0 0.0 3.3 1.3 0.0 0.0 14.9 3.0 3.2 0.9 44.1 3.5 Coast Urban 0.0 0.0 2.8 2.0 0.0 0.0 11.3 3.2 2.7 1.1 46.1 5.7 Total 0.0 0.0 3.5 1.2 0.0 0.0 13.4 2.1 3.4 0.8 45.8 2.5 Isolated 0.0 0.0 13.1 7.9 0.0 0.0 38.4 9.6 13.6 5.6 90.2 3.9 Central Rural 0.0 0.0 28.0 6.3 0.0 0.0 21.1 5.3 22.9 4.8 85.7 3.5 Highlands Urban 0.0 0.0 7.9 5.5 0.0 0.0 25.7 7.0 21.9 7.2 73.7 6.5 Total 0.0 0.0 19.3 3.9 0.0 0.0 26.1 3.8 20.2 3.2 83.4 2.5 Isolated 0.0 0.0 2.4 1.8 0.0 0.0 15.2 4.7 2.9 2.6 60.4 6.6 Rural 0.0 0.0 4.1 1.4 0.0 0.0 7.4 2.2 2.6 0.7 54.8 3.0 Southeast Urban 0.0 0.0 1.7 0.8 0.0 0.0 8.4 2.1 5.1 1.7 61.6 4.0 Total 0.0 0.0 2.9 0.8 0.0 0.0 8.8 1.5 3.6 0.8 58.3 2.2 Isolated 0.0 0.0 10.5 3.4 0.0 0.0 6.0 2.2 4.7 1.5 63.4 5.0 Mekong Rural 0.0 0.0 3.4 0.8 0.0 0.0 10.5 1.6 5.1 0.8 64.8 2.1 Delta Urban 0.0 0.0 0.6 0.6 0.0 0.0 6.9 2.7 1.4 0.8 71.8 3.7 Total 0.0 0.0 4.0 0.7 0.0 0.0 9.2 1.1 4.5 0.6 65.8 1.6 Isolated 0.0 0.0 7.2 1.4 0.0 0.0 24.6 2.3 4.9 0.9 64.3 2.3 Rural 0.0 0.0 3.1 0.4 0.1 0.0 18.0 0.8 3.2 0.3 48.1 1.0 Vietnam Urban 0.0 0.0 2.0 0.6 0.2 0.2 15.3 1.4 3.6 0.7 45.7 2.0 Total 0.0 0.0 3.4 0.3 0.1 0.0 18.3 0.7 3.5 0.3 49.6 0.8 N.A.=data not available or too few for stable estimate 207 Vietnam Reading and Mathematics Assessment Study Lack of pupil and teacher materials (do not have) It goes without saying that teachers need the teacher guides for what they are teaching and that the pupils need certain materials. The government set the benchmark that all pupils should have the Vietnamese, Math, Science, and History and geography textbooks. According to FSQL, a notebook and a pen were the minimum materials that all pupils should have. Only 1 % of pupils did (a) Notebook for writing: Parents are expected to provide a notebook for not have notebook for writing for their children. It can be seen from Table 6.4 that only one percent writing. of Grade 5 pupils had no notebooks for writing. 7% did not have black (b) Black pencil: Again the parents are meant to ensure that their children pencil. who go to school have a pencil. There were nearly seven percent of pupils who did not have a pencil. In isolated areas in the Northeast, Northwest, North Central, and the Central Coast regions there were quite high percentages of pupils without a pencil. One wonders why this is so and if it is a matter of poverty, aid agencies might be persuaded to supply pencils to the isolated areas in these regions for a period of time. 1% did not have (c) Vietnamese textbook 5, Volume 1: Only one percent of pupils did not Vietnamese textbook. have this textbook but in isolated schools in the Northwest this was four percent. The authorities are meant to provide textbooks to needy children. Vietnamese textbook Vol.1 Vietnamese textbook Vol.2 Math textbook Province Do not have Province Do not have Province Do not have % SE % SE % SE Ha Giang 3.9 1.1 Hoa Binh 2.7 0.6 Ha Giang 2.7 0.8 Son La 2.7 0.9 Ha Giang 2.6 0.9 Gia Lai 2.0 0.8 Hoa Binh 2.7 0.7 Gia Lai 2.4 0.9 Ha Tay 1.7 0.3 Soc Trang 2.2 0.6 Kien Giang 2.0 0.5 Yen Bai 1.6 0.8 Yen Bai 2.0 1.0 An Giang 1.9 0.6 Cao Bang 1.5 0.6 (d)Vietnamese textbook 5, Volume 2: Only in the isolated schools in the Central Highlands was there a small problem. 1% did not have (e) Math textbook: Nearly all pupils had this textbook and again it was in the mathematics textbook isolated schools in the Central Highlands that there was a small problem. But, in Science, the (f) Science textbook: Almost 6 percent of pupils did not have science figure was 6%. textbooks. Pupils in isolated areas tended not to have the textbooks. Especially, in isolated Central Highlands and isolated Northwest are problematic. And for History and (g) History and geography textbook: Eleven percent pupils did not have geography, it was 11%.. history and geography textbooks. Again, in isolated Central Highlands and Northwest, more than 25 percent of pupils did not have the textbooks. 208 Vietnam Reading and Mathematics Assessment Study (h) Teacher Guide for teaching Vietnamese: Nearly five percent of pupils 5% of pupils were in were in schools where the teacher stated that he or she did not have access to schools without teacher a Teacher Guide in Vietnamese. Again it was in the isolated schools in the guides for Vietnamese Northwest, North Central and Central Highlands (also rural areas in the and mathematics. Central Highlands) that there were problems. It is the local authorities that are responsible for delivering these books to the schools and one wonders what must have gone wrong that the teachers either did not have access to a guide or did not know that they had such access. (i) Teacher Guide for teaching Mathematics: Exactly the same comments apply for the math teacher guides as for the Vietnamese teacher guides. Table 6.4: Pupil and teacher materials Notebook for Vietnamese Vietnamese Science History and Teachers Teachers writing Black penciltextbook Vol.1 textbook Vol.2Math textbook textbook geography guide (VN) guide (Math) Region School textbook location Do not have Do not have Do not have Do not have Do not have Do not have Do not have Do not have Do not have % SE % SE % SE % SE % SE % SE % SE % SE % SE Isolated 0 0 10.4 8.8 0 0 2.2 2.3 2.2 2.3 9.0 10.5 15.4 12.4 34.5 35 0 0 Red River Rural 0.5 0.1 4.2 0.4 0.8 0.1 0.8 0.1 0.8 0.1 4.5 0.2 8.1 0.4 2.8 0.7 2.9 0.7 Delta Urban 0.4 0.1 1.6 0.4 0.5 0.2 0.3 0.1 0.7 0.3 1.8 0.3 3.8 0.5 6.7 3 4.5 2 Total 0.5 0.1 3.8 0.3 0.7 0.1 0.7 0.1 0.8 0.1 4.1 0.2 7.5 0.3 4 1.1 3.1 0.6 Isolated 0.6 0.2 17.1 2.2 2.1 0.4 1.5 0.3 2 0.3 13.4 1.7 20.3 1.9 4.8 1.9 5.3 1.9 Rural 0.4 0.1 6.3 0.4 0.9 0.1 0.9 0.1 0.7 0.1 7.1 0.4 12.7 0.5 4.1 0.8 3.6 0.7 Northeast Urban 0.3 0.2 1.5 0.4 0.8 0.2 0.6 0.2 0.5 0.2 2.4 0.6 5.4 0.7 4.7 2.1 4.2 1.7 Total 0.4 0.1 7.3 0.5 1.1 0.1 1 0.1 0.9 0.1 7.4 0.4 12.8 0.5 4.3 0.8 4 0.7 Isolated 2.1 0.7 25.2 3.5 3.9 1.1 3.3 0.7 1.9 0.6 18.9 2.8 26.5 3.0 12.7 4.7 13.2 4.7 Rural 0.7 0.3 12.7 2.1 1.9 0.5 1.8 0.4 0.7 0.2 13.5 1.9 21.3 2.0 6.5 2.7 7.8 2.9 Northwest Urban 0.2 0.2 2.4 0.8 0.4 0.3 0.4 0.2 0.4 0.3 3.9 0.7 6.3 1.3 3.9 4 0 0 Total 1.1 0.3 15.6 1.6 2.4 0.5 2.1 0.3 1.1 0.2 14.0 1.2 20.9 1.4 8.5 2.3 8.6 2.3 Isolated 0.8 0.5 22 3.6 2.7 1 0.8 0.4 0.6 0.3 14.6 2.3 21.2 2.8 15.6 6.8 16.3 6.9 North Rural 0.8 0.2 8.7 1 1.4 0.3 0.5 0.1 0.4 0.2 6.9 0.9 13.2 0.7 4.5 1.5 4.6 1.5 Central Urban 0.9 0.3 2.4 0.8 0.7 0.3 0.1 0.1 0.3 0.2 3.6 0.9 6.9 0.9 3.6 2.5 4.6 2.7 Total 0.8 0.1 9.1 0.9 1.4 0.3 0.5 0.1 0.4 0.2 7.2 0.7 13.1 0.6 5.7 1.5 6 1.5 Isolated 2 0.6 16.9 4.5 1.4 0.5 1 0.4 1.1 0.5 8.9 2.4 17.7 2.8 4 2.7 4 2.7 Central Rural 2.1 0.3 4.4 0.6 0.7 0.2 0.4 0.1 0.2 0.1 4.5 0.4 10.7 0.5 2 1.2 2 1.2 Coast Urban 2.1 0.4 3.5 0.5 0.3 0.1 0.4 0.2 0.4 0.2 2.6 0.4 8.1 1.0 0.4 0.3 1.8 1.7 Total 2.1 0.2 5.7 0.7 0.7 0.1 0.5 0.1 0.3 0.1 4.5 0.4 10.9 0.6 2 1 2.3 1.1 Isolated 1.2 0.5 26.3 4 2.3 0.7 2.6 1.2 2.7 1.1 19.2 3.3 31.3 3.7 12.5 6.1 15.6 6.9 Central Rural 1 0.3 13.8 1.8 0.6 0.2 0.6 0.2 0.6 0.3 8.2 1.0 15.0 1.2 11.5 3.6 11.9 3.4 Highlands Urban 0.9 0.3 4.3 0.9 0.7 0.3 0.1 0.1 0.5 0.3 4.0 0.8 8.7 1.1 2.1 1.2 3.6 2 Total 1 0.2 13.9 1.2 1 0.2 0.9 0.3 1 0.3 9.5 0.8 16.8 1.1 9.8 2.4 11.1 2.4 Isolated 1.2 0.4 8.8 1.4 1.5 0.4 0.5 0.2 0.6 0.2 5.1 1.0 11.4 1.1 3.3 1.5 3.1 1.5 Rural 1.1 0.2 4.9 0.4 0.9 0.2 0.5 0.1 0.5 0.1 4.2 0.3 9.5 0.4 5.1 1.3 6 1.5 Southeast Urban 1.5 0.3 2.1 0.3 0.6 0.1 0.3 0.1 0.2 0.1 2.1 0.3 6.5 0.6 2.4 2 3 2 Total 1.3 0.1 4.2 0.3 0.9 0.1 0.4 0.1 0.4 0.1 3.4 0.2 8.5 0.4 3.8 1 4.5 1 Isolated 2.6 0.4 8.5 1 0.9 0.2 1.2 0.3 0.3 0.1 6.5 0.7 13.8 1.0 3.8 1.8 3.8 1.8 Mekong Rural 2.5 0.2 7.1 0.4 0.9 0.1 0.9 0.1 0.4 0.1 5.8 0.3 12.0 0.5 3.6 0.9 4.7 0.9 Delta Urban 2 0.4 4.1 0.7 0.6 0.2 0.3 0.1 0.3 0.1 4.1 0.5 8.7 0.6 5.3 2.3 5.3 2.3 Total 2.4 0.1 6.8 0.3 0.9 0.1 0.9 0.1 0.4 0.1 5.6 0.2 11.7 0.4 3.9 0.8 4.6 0.8 Isolated 1.5 0.1 15.6 0.9 1.8 0.2 1.4 0.2 1.2 0.1 11.1 0.7 18.6 0.7 7.6 1.5 7.4 1.2 Rural 1.1 0.1 6.5 0.2 1 0.1 0.7 0.1 0.6 0.1 5.9 0.2 11.4 0.2 4.1 0.5 4.4 0.5 Vietnam Urban 1.2 0.1 2.6 0.2 0.6 0.1 0.3 0 0.4 0.1 2.8 0.2 6.7 0.3 3.8 0.8 3.8 0.8 Total 1.2 0.1 6.8 0.2 1 0.1 0.7 0 0.6 0 5.9 0.1 11.3 0.2 4.6 0.4 4.7 0.4 209 Vietnam Reading and Mathematics Assessment Study It was satisfying to note that for the most part the benchmarks were met. This must be very gratifying for the MOET. There were some instances where pupils did not have the required textbooks and materials and in this case action should be taken to ensure that pupils acquire them. There were other instances where the benchmarks for teacher training had not been met fully and there was reason to believe that some further action is needed. Policy suggestions for these last two points have already been made in earlier chapters. Above all there were two sets of benchmarks and some were even further subdivided into isolated vs. rural and urban schools. There were several instances where benchmarks did not exist when it would have been desirable to have had them. Once a new set of benchmarks has been arrived at, these should be printed into one document and made known to all. Conclusion It would be desirable if the MOET were to re-examine its sets of benchmarks and decide on possible revisions to them. This would involve not only revising existing benchmarks but also deciding on further benchmarks that might be useful. For the most part the existing benchmarks were met and where there was a lack of textbooks and materials in a few schools or where there was a dearth of teachers meeting the teacher training standards, then appropriate action should be take to have the pupils acquire the textbooks and materials, and the teachers should be upgraded. 210 Vietnam Reading and Mathematics Assessment Study Chapter 7 HOW EQUITABLY ALLOCATED WERE EDUCATIONAL RESOURCES TO PRIMARY SCHOOLS? HOW EQUITABLY DISTRIBUTED WAS PUPIL ACHIEVEMENT BETWEEN SCHOOLS? Introduction T he educational goal of giving equality of educational opportunity to all pupils is an issue that goes far beyond giving every child access to school education. It also involves giving all of those pupils who are in school an equal opportunity to learn. One way to do this is to ensure that there is equitable distribution of educational resources among all schools. In all school systems where the government wishes to take action to address Not only are the levels of the issue of equity, it is important to know the `location' of differences or inputs to schools variations in resource inputs to schools. For example, it is important to know important (see previous whether variations in resource inputs are more pronounced among provinces, chapters) but how or whether they are larger among schools within provinces. An answer to this equitable the inputs were type of question provides guidance concerning which resources are allocated among schools distributed evenly or unevenly, and at the same time suggests the level at is also an important which decisions must be taken (national or provincial) in order to address any feature of any major inequities that are observed. educational system. In exploring questions of equity, it must be recognised that there is a need to examine allocation patterns in association with the actual levels of provision that were given in chapter 4 and 5. Such information is necessary because it enables policy-makers to identify which resources require attention, and also to have some feeling for the amounts of supplementary resources that may be needed in order to achieve a more equitable distribution. 211 Vietnam Reading and Mathematics Assessment Study This chapter concentrates mainly on an examination of the distribution of educational resources and not upon absolute resource levels. Consequently the results presented in the chapter should be examined in conjunction with the results from chapters 4, 5, and 6 which examined baseline data and benchmark allocations. Two approaches to the measurement of equity in material and human resource inputs (i) Variation among schools within regions compared with overall variation Two measures of equity It is also possible to quantify the differences among schools within a were employed. The firs particular region by making a comparison with the variation among schools was variation among at the national level. This can be achieved by using the formula below: schools within regions. The variation among Standard deviation for schools in a region schools within regions X 100 was compared with the Standard deviation for schools in the nation variation among schools for the whole country. The standard deviation of an indicator for a particular region measures the amount of variation among schools within the region, whereas the standard deviation for the whole country measures the variation among schools for the whole nation. The ratio of the standard deviation of an indicator for a region to the standard deviation for the nation, expressed as a percentage, provides a measure of the degree of equity within a region compared with the national picture. To illustrate the interpretation of these ratio values it is helpful to consider two hypothetical regions: Region A and Region B. Assume that the levels of a resource are measured by an indicator that has a ratio value of 50 percent for Region A and 150 percent for Region B. This figure would mean that the variation in resource levels among schools in Region A is 50 percent less than the variation in resource levels for the nation as a whole: and the variation in Region B is 50 percent higher than for the nation as a whole. From these ratio values it can be seen that, compared with the national picture, there has been a more equitable allocation of that resource in Region A while in Region B there are major inequities among schools compared with the nation as a whole. 212 Vietnam Reading and Mathematics Assessment Study (ii) Variation among provinces A statistic called the coefficient of intra-class correlation (rho) may be used The second was the to divide the variation in resource inputs into two components: (a) among proportion of total provinces, and (b) among schools within provinces. Rho can range from zero variation that was to 1.00. When used in this way, rho is a ratio that measures the percentage of attributable to variation total variation among schools that can be attributed to variation among among province as provinces. compared with among schools within To appreciate the meaning of rho, it is useful to consider two hypothetical provinces. school systems: System A and System B. In school system A, resources are allocated equally, or nearly equally, to all schools and therefore when one calculates average resource levels for regions or provinces in the system one finds that these are more or less the same - except perhaps for some minor chance deviations. For such a school system, the value of rho would be close to zero because of the small variation among regions or provinces. In this situation most of the variation would be among schools within regions or provinces. On the other hand, consider school system B where, because of administrative decisions, historical factors, or geographical differentiation of social class groups, etc., there are large variations among the provinces. In this case the value of rho would be close to unity. The majority of the variation in this case would be due to variations among regions or provinces and there would be little variation among schools within regions or provinces The above examples are extremes that serve to illustrate the interpretation of rho. In using rho for policy discussion it is common practice to multiply the values of rho by 100 in order to present a more readable discussion about the `percentage of variance'. For example, a rho of, say, 0.20 means that 80 percent of the variation is among schools within regions or provinces and 20 percent among regions or provinces. In contrast, a rho of 0.8 would indicate that 80 percent of the variation is among regions or provinces and 20 percent among schools within regions or provinces. Equity calculations for material resource inputs Two sets of equity measures for material resources have been presented in Table 7.1. The first measure was how the variation among schools within a region compared with the national variation among schools. For each region the value for the comparison of the standard deviation for a resource in the region with the national standard deviation has been made. The value of 101.4 for classroom furniture in the Red River Delta means that the variation among schools in that region was slightly higher than for the country as a whole. On the other hand the variation for classroom materials among schools within the region was less than for the country as a whole. 213 Vietnam Reading and Mathematics Assessment Study The second measure of equity was the proportion of total variation for a resource that was attributable to variation among provinces as compared with between schools within provinces. In the bottom row of Table 7.1 values of rho have been presented. It can be seen, for example, that in the final column for `School resources' the rho was 0.34. This means that 34 percent of the variation for school resources was attributable to differences among provinces and 66 percent of the variation was attributable to differences among schools within provinces. The variation tells us nothing about the levels of inputs. The levels were presented in Chapters 4 and 5. It is important to interpret the variation statistics bearing in mind the absolute levels of provision that were reported in the previous chapters. Table 7.1: Equity of classroom resources and school resources within regions and among provinces Classroom Classroom Classroom Classroom Classroom School furniture index materials index resources library or space per resources (max=10) (max=6) (furniture + supplies) index bookcase pupil index Red River Delta 101.4 76.0 87.0 117.4 99.5 76.7 Northeast 93.5 103.7 97.5 72.4 81.1 92.4 Northwest 82.8 120.0 99.6 36.4 272.9 92.0 North Central 105.1 104.5 108.6 148.4 79.9 84.4 Central Coast 84.9 84.4 85.9 126.8 63.3 94.0 Central Highlands 81.3 124.7 97.7 17.0 95.9 101.0 Southeast 93.3 93.2 90.2 89.3 72.1 108.2 Mekong Delta 73.7 98.6 80.8 53.2 104.0 85.8 Variation among provinces 0.27 0.14 0.27 0.04 0.04 0.34 There was considerable In terms of disparities between schools within regions compared with the variance among national disparity it will be seen that for classroom libraries (bookcases) the provinces, for classroom Red River Delta, North Central and Central Coast regions had more variation furniture and school than the nation whereas the Central Highlands had very low variation resources. compared with the national variation. Taking one or two examples might prove useful. The relevant tables in Chapter 4 and 5 are: Tables Classroom furniture and teaching materials indices 4.50 Classroom library or book case 4.40 Classroom space per pupil 5.13 School resources 5.16 The first index for classroom furniture is for the Red River Delta and is 101.4. The calculation to arrive at 101.4 was; 214 Vietnam Reading and Mathematics Assessment Study Standard deviation for Red River Delta = 1.4919 *100=101.4 Standard deviation for Vietnam = 1.4707 If the two standard deviations had been identical, then the resultant index would have been 100. Hence, it can be seen that the variation in the Red River Delta was about the same as for Vietnam as a whole. The level (see Table 4.5) for furniture in the Red River Delta was 6.5 out of 10 items. This was above the mean of the country, which was 5.5. It can be said that the Red River Delta had relatively good level of furniture and about the same variation among schools as the country as a whole. The lowest index in the table is 17.0 for classroom library or bookcase in the Within regions the Central Highlands. The index of 17.0 shows that there was very little Northwest had high variation among schools in this region. In Table 4.4, it can be seen that the variation among schools mean for the region was one book per classroom. The mean for Vietnam was for classroom space and 7.8 books per classroom. Thus, the level was extremely low for the Central Highlands and there was very little variation among schools. That is, all the North Central for schools had virtually no books in their classroom library or bookcases. classroom library. There was low variation in For classroom space the variation in the Northwest region was very large many provinces for compared with the nation as a whole. classroom library where no school had them. For the other variables there was not very much difference in variation within regions as for the nation altogether. For the rho values it can be seen that there were three values over 0.20. These were classroom furniture, total classroom resources and school resources. In these cases the national authorities need to consider what can be done together with the provincial authorities. Policy suggestion 7.1: The MOET may wish to consider equalizing classroom resources and school resources between provinces. Equity calculations for human resource inputs Educational planners are not only responsible for the allocation of material There was an equitable resources to schools but also for the allocation of human resources. If all allocation of human pupils are to have an equal access to education, it is important that they are in resources, except for schools that have the basic resources in terms of materials and also teachers. teacher qualifications in the Northern regions. The statistics for the variation among schools within regions as well as the rho between regions have been presented in Table 7.2. It can be seen that the rho values for all resources among regions were below the 20 percent limit. For teacher academic qualifications there was slightly more variation in the northern regions than in the country as a whole. For teacher professional 215 Vietnam Reading and Mathematics Assessment Study qualifications the region with the largest variation was the Northwest but it was not very high. However the only two figures worthy of note were for school head's professional qualifications where the variation was high in the Central Highlands and for inspectors visits in the Southeast. Table 7.2: Equity resource distribution of human resources to schools within regions and among provinces Teacher Teacher School head School head Academic professional Teacher academic professional School head Inspectors Pupil-teacher Qualification qualificationexperience qualifications qualificationsexperience visits ratio Red River Delta 110.2 78.7 97.4 106.9 88.8 98.4 75.3 81.2 Northeast 113.8 90.3 104.8 100.2 87.4 101.9 79.1 82.7 Northwest 111.2 109.4 96.3 107.4 82.6 110.8 42.7 70.1 North Central 111.5 82.9 107.0 102.0 91.9 108.4 68.8 95.1 Central Coast 97.4 95.7 101.3 84.7 95.0 61.7 85.8 94.8 Central Highlands 97.3 99.2 88.2 98.0 125.2 80.4 56.6 77.9 Southeast 78.0 70.4 97.3 86.7 109.4 91.3 136.9 109.1 Mekong Delta 73.6 80.6 92.6 93.4 107.7 83.8 121.8 90.3 Variation among provinces 0.18 0.20 0.06 0.14 0.11 0.04 0.22 0.38 Policy suggestion 7.2: a) The national authorities and provincial authorities need to consider what can be done to reduce the disparities among provinces for classroom and school resources; and, b) The provincial authorities in the regions of the Central Highlands and the Southeast and Mekong Delta should take action to decrease the variation in the school heads' teaching experience (Central Highlands) and the number of inspectors visits to schools (Southeast and Mekong Delta). How much variation was there in achievement? For achievement, the coefficient of intra-class correlation (rho) was used to measure equity. Two different rho values were calculated. One was to measure the variation in achievement scores into two components: (a) among provinces, and (b) among pupils within provinces. When used in this way, rho is a ratio that measures the percentage of total variation among pupils that can be attributed to variation among provinces. The other was to measure the variation in achievement scores into two components: (a) among schools, and (b) among pupils within schools. When used in this way, rho is a ratio that measures the percentage of total variation among pupils that can be attributed to variation among schools. In Vietnam it was the variation among schools in achievement that was remarkable. From Table 7.3 it can be seen that whereas the variation among provinces was relatively low it was the variation among schools that was high. 216 Vietnam Reading and Mathematics Assessment Study Table 7.3: Variation in achievement scores among provinces and Pupil Among provinces, Reading score Math score variation was not large Variation among provinces 0.11 0.14 but among schools it Variation among schools 0.56 0.64 was very large indicating great inequity among schools Variation among schools is usually a result of both the social stratification of of achievement of the society where the communes can be very different and the allocation of different schools. materials and resources to schools. It is reasonable to assume that children are not very different in intelligence according to where they are born. If the schools are more or less the same then the chances of receiving the same education are high. But where the schools are very different then the chances are very unequal. A fair society is one that strives to offer all children a good education wherever they happen to live. This is explicated further below. - How much variation was there among schools within regions and provinces in achievement? Among provinces, Within each region, rho was calculated in order to measure the variation in variation was not large achievement scores into two components: (a) among schools, and (b) among but among schools it pupils within schools. When used in this way, rho is a ratio that measures the was very large percentage of total variation among pupils that can be attributed to variation among schools. This has been presented in Table 7.4. The rho for the Red indicating great inequity River Delta was 0.51, meaning that in the Red River Delta, 51 percent of the of achievement of variation in pupils' reading achievement scores was among schools and 49 different schools. percent of the variation was among pupils within schools. As seen in Table 7.4, there was large variation in pupils' achievement scores among schools in the Northeast, Northwest, and Central Highlands regions. Table 7.4: The intra-class correlations (rho) for reading and mathematics by region Pupil The differences among Region Reading score Math score schools were high in the Red River Delta 0.51 0.61 Northeast, Northwest, Northeast 0.69 0.76 and Central Highlands. Northwest 0.67 0.77 North Central 0.55 0.58 Central Coast 0.44 0.48 Central Highlands 0.68 0.70 Southeast 0.41 0.50 Mekong Delta 0.53 0.60 217 Vietnam Reading and Mathematics Assessment Study Similarly, in Table 7.5, within each province the differences among schools were assessed by calculating the intra-class correlation for each subject area. Within provinces there Province Rho: Reading Rho: math were also large Dong Nai 0.15 0.23 Khanh Hoa 0.25 0.29 differences among Ba Ria - Vung Tau 0.29 0.32 schools. Binh Phuoc 0.42 0.37 Binh Thuan 0.34 0.39 Quang Nam 0.44 0.40 Phu Yen 0.27 0.41 Ho Chi Minh 0.35 0.42 Da Nang 0.34 0.42 Ben Tre 0.32 0.45 Soc Trang 0.47 0.46 Ha Nam 0.36 0.46 Quang Tri 0.38 0.47 The equity of Ha Noi 0.35 0.47 achievement scores Bac Giang 0.40 0.47 among schools within Ninh Thuan 0.35 0.48 Tay Ninh 0.37 0.48 provinces was examined. Ninh Binh 0.32 0.49 The differences among Can Tho 0.57 0.49 Tra Vinh 0.48 0.49 schools were dramatic. Binh Duong 0.38 0.50 Among regions and Long An 0.38 0.50 provinces, it was not Vinh Long 0.52 0.50 Quang Ngai 0.48 0.50 large but among schools Hai Phong 0.44 0.51 it was very large Ha Tinh 0.57 0.51 Kien Giang 0.47 0.54 indicating great inequity Thanh Hoa 0.48 0.55 of achievement of Thua Thien - Huu 0.58 0.57 different schools. Nam Dinh 0.40 0.59 An Giang 0.54 0.59 Binh Dinh 0.56 0.60 Dak Lak 0.57 0.60 Thai Binh 0.59 0.61 Hung Yen 0.54 0.63 Lang Son 0.61 0.63 Lam Dong 0.52 0.64 Phu Tho 0.59 0.64 Vinh Phuc 0.51 0.65 Nghe An 0.62 0.65 Tien Giang 0.54 0.65 Lao Cai 0.52 0.65 Gia Lai 0.65 0.66 Hai Duong 0.59 0.66 Dong Thap 0.54 0.66 Ha Tay 0.57 0.68 Quang Binh 0.59 0.68 Ca Mau 0.57 0.69 Lai Chau 0.57 0.70 Tuyen Quang 0.58 0.70 Ha Giang 0.60 0.72 Bac Lieu 0.65 0.73 Quang Ninh 0.67 0.73 Thai Nguyen 0.68 0.75 Kon Tum 0.73 0.76 Yen Bai 0.75 0.76 Cao Bang 0.76 0.77 Bac Ninh 0.68 0.77 Hoa Binh 0.65 0.79 Son La 0.73 0.79 Bac Kan 0.70 0.80 218 Vietnam Reading and Mathematics Assessment Study Table 7.5: The intra-class correlations for reading and mathematics by province The rhos have been presented in ascending order. The first province (Dong Nai) has the lowest rho and the last province (Bac Kan) has the highest. As described at the beginning of this section, the rho means that in Dong Nai there were relatively small differences between schools (23 percent of all variation in scores) and therefore 77 percent of the variation was among pupils within schools. On the other hand, a rho of .8 is very high and means that 80 percent of the variation was between schools and only 20 percent of the variation was within schools. The differences among schools were much In general the observed rhos in Vietnam are very high. In Japan and higher than many other Scandinavian countries the rho are below 0.10 meaning that it does not countries. very much matter where a child is born. He or she will have more or less equal chances of having the same education. In Europe at the primary school level the rho is between 0.15 and about 0.20. In many developing countries it is between 0.30 and 0.40. But given the value of the observed rhos in Vietnam, the chance of a child in one small area of having the same education as a child in another area is small. Three figures have been presented below for the three provinces of Dong Nai (a relatively low rho and the lowest in Vietnam), Hanoi (with a rho of 0.47) and Son La (with a rho of 0.79). The blue line represents the mean and the spread presented for each school is the standard deviation of scores for the school. The scores are for mathematics. Although Dong Nai has a low rho, it can also be seen that the spread of mean scores is low ranging from 400 to 500. Thus the levels were not too different but also low. Hanoi ranged from 400 to 700 and Son La from 250 to nearly 700. Son La showed a great deal of differences between schools and the low schools in Son La certainly need help. It can be seen that in the province of Son La it made a great deal of difference where a child was born. If he or she was born in a school near the left hand side of the figure then his or her chances of a good education was poor. If on the other hand he or she was born near a school on the right hand side of the figure the chances were good. Indeed, there was a difference of about two and half pupil standard deviations between the worst and best school. It is often said that about three-quarters of a 219 Vietnam Reading and Mathematics Assessment Study standard deviation is the same as one year of schooling. Three pictorial examples of differences among schools within provinces for mathematics achievement were given as examples. Dong Nai, where the differences were not great... Figure 7.1: Differences in math scores between schools in Dong Nai Province Hanoi where the differences were larger... Figure 7.2: Differences among schools in mathematics achievement in Hanoi province 220 Vietnam Reading and Mathematics Assessment Study and Son La where the differences were very large. The disparity in pupil achievement between schools is a major problem. Figure 7.3: Differences in math score among schools in Son La Province It is clear from the above that the variation between schools within provinces is extremely high. It is also known that some of the lower performing schools have poor human and material resources. Suggestions had been made in earlier chapters about the urgent need to increase such resources in needy schools. Conclusion This chapter has been concerned with the equity of provision to schools and achievement in provinces and regions. It was found that there was reasonable equity on many of the resources but that the national authorities should take action to correct the imbalance between provinces for classroom and school resources as well as for staff teacher training and pupil meals per day. Of particular concern was disparity between schools within provinces and for the country as a whole. Although action should and can be taken by the authorities to equalise the allocation of human and material resources to schools the question must be raised about providing a better infrastructure to communities, having them sending their children to school and maybe giving more resources to the most needy schools. 221 Vietnam Reading and Mathematics Assessment Study The differences among schools overall in the country and also within provinces was high compared with other countries. This kind of inequity at the primary education level is unacceptable and much effort will be needed to reduce these differences. As has been seen in this chapter and also in subsequent chapters it is a number of variables that in conjunction are accounting for these achievement differences among schools. It is for the national and provincial authorities to examine the levels if provision for the variables associated with the achievement differences and decide on a plan of action. It must be recognized that in order to have equality of outcomes it is often necessary to have inequity of inputs and particularly processes. At the same time, there must be a level of provision below which no school should ever be. 222 Vietnam Reading and Mathematics Assessment Study Chapter 8 EFFECTIVE PRIMARY SCHOOLS IN VIETNAM1 Introduction I n Chapters 3, 4, and 5, those variables whose correlations and partial correlations were reasonably large were identified as being associated with differences in pupil achievement in the Vietnamese primary education system. However, it was considered desirable to analyse the data in a different way in order to see if similar results were obtained. Certainly, if similar results were obtained from a different way of 'looking at the data', this would be very reassuring. The 'other' way has been called an 'effective schools analysis' An effective school is one that achieves better than expected given its intake An effective school is of pupils. That is to say, on average the pupils achieve better than could be school whose pupils expected given the relationship between achievement and home background achieved better than for all Grade 5 pupils in the country. By definition, an ineffective school is expected given their therefore one that achieves less well than expected given its intake of pupils. What is meant by 'given its intake of pupils'? In general, pupils from better "home background". An homes have higher achievement scores than pupils from poorer homes. What ineffective school is one is meant by a 'good or poor home'? Good homes are those where the parents that achieved less well have a high level of education, where there are many possessions in the home than expected. because the parents have sufficient money to buy the possessions. A good home will therefore have many books in the home. It is true that a pupil from a home with many books has more opportunity to read than a pupil from a home where there are few books. In Vietnam, pupils from poorer homes tend to come from more isolated areas, "Home background" belong to ethnic groups other than the main ethnic group of Kinh and not to represents the extent to speak so much Vietnamese outside the school. They tend to speak their which parents were mother-tongue language that is very different from Vietnamese. All of these interested in and cared factors can be formed into a construct called 'home background'. In a sense, these variables are proxy measures for the interest that the parents show in for children's education. their children learning in school. 1This chapter was written by Miyako Ikeda 223 Vietnam Reading and Mathematics Assessment Study Variables with large There were 3654 schools in the sample. It was of interest to identify the differences between the schools that were most effective and the schools that were least effective and group of the most then identify those variables where there were large differences between the effective schools and the two groups of schools. groups of the least effective schools were identified. Method of analysis A home background factor (a principal component) was formed of the following variables with the following loadings: Home background Variable Loading consists of 5 variables. Speaking Vietnamese at home .797 Belonging to ethnic minority .794 No. of books at home .467 Years parental education .591 Wealth of home (total possessions) .662 Pupils with 'higher' The correlation between this home background factor and reading home backgrounds had achievement was 0.30 and with mathematics achievement it was 0.27. As higher achievement expected, pupils with higher values on the home background factor tended to scores in both obtain higher scores both in reading and in mathematics. The correlation mathematics and between the two achievement scores was 0.74. reading. An "expected" score for A simple bivariate regression line was established between the home each pupil was background factor and the pupil scores on a) the reading test, and b) the mathematics test. The pupils placed above the regression line were calculated. interpreted as having reading scores (or mathematics) scores that were better than could be expected - after taking the home background into account. Conversely, pupils placed below the line had scores that were worse than might be expected. The differences between The residual (actual minus predicted scores) scores were then averaged over the expected and schools so that a school with a very high mean residual score was identified achieved scores at the as a 'more effective' school because it had many pupils whose achievement pupil level were scores were much higher than expected. (This procedure also avoids calculated, then, aggregation effects that would have occurred if the procedure had been aggregated to the school carried out at the between school level only). level. Schools where the actual It is important to note that this definition of a 'more effective' school score was higher than made it possible for a school to be designated as effective even if it the expected score were had a relatively low raw mean score - perhaps even considerably lower than the average for all schools in the country. Similarly, a 'more effective.' Schools school with a relatively high mean score might after considering the where the actual score home background of its pupils be designated as a 'less effective' was lower than the school. As seen in Figure 8.1 below, School A is regarded as 'more expected score were 'less effective' school because it is above the regression line, even though effective.' the mean mathematics score of School A was 453, which was lower than the national average of 500. On the other hand, School B is 224 Vietnam Reading and Mathematics Assessment Study regarded as 'less effective' school because it is below the regression line, even though the mean mathematics score of School B was 514, which is higher than the national average. Figure 8.1: Comparison between math score and home background based on a regression line For each score (reading and mathematics) the schools were placed in rank Two groups of schools order from the most effective to the least effective school and then 100 were formed; 100 most schools at each extreme were taken in order to identify on which variables effective schools and the there were differences. 100 lest effective It should be pointed out that the aim of the analyses presented in this schools. article comparing the more and less effective schools had not been to establish precise measures of the effects of various variables on mean pupil reading scores as might be undertaken using complex and hierarchical causal modelling. Rather the aim is to identify a summary list of variables that will be of interest to planners and to others undertaking further analyses of the data. Construction of variables In all, there were more than 440 separate questions on the questionnaires. In New variables were also some cases, one question (e.g., sex of pupil) was used as an indicator. This created based on the type of indicator is known as a singleton. Sometimes, one question was original variables in the recoded to make it more meaningful for analysis purposes; for example, as seen in the question below, the original coding was from 1 to 5. But it was questionnaires. recoded into 0 through 2 for the purpose of differentiating pupils did not use 225 Vietnam Reading and Mathematics Assessment Study textbooks (0), from those who shared with others (1), from those who had their own textbooks. Do you have a Vietnamese textbook in your classroom during the lessons? (Please write a cross in only one box.) Original coding Recoded into; 1There are no Vietnamese textbooks. 0 2Only the teacher has a Vietnamese textbook. 0 3I share a Vietnamese textbook with two or more pupils. 1 4I share a Vietnamese textbook with one pupil. 1 5I use a Vietnamese textbook by myself. 2 In all, the differences In other cases, two or more variables were used to form an indicator; between the two groups for example, the total enrolment of the school divided by the number of schools were of class groups forms a ratio indicator know as the average class size. examined for 123 In yet other cases, a number of questions were combined to estimate, variables. for instance, the total school resources. Where questions are combined in some way, the resultant variable is known as a construct or composite. The list of indicators used for the analysis in this chapter and the type of construct they are have been presented in Appendix 8.1. Most and least effective 100 schools Differences worthy of The first comparison made was between the most effective 100 schools note have been and the least effective 100 schools in reading and mathematics. For presented. Large each variable the mean of the top 100 schools (the most effective differences have been schools) has been presented. This has been followed by the mean of the given and very 100 least effective schools and then the overall standard deviation. The large differences . standard deviations will be the same under the columns for reading and The first result presented math because it is, for example, the age or ratio female for all schools were those for the and they are the same for both subject areas. Finally, the difference has differences between the 100 most effective and been presented. Where the difference between the values for the most the 100 least effective and least effective schools was greater than one standard deviation schools. three diamonds ( ) have been entered, half a standard deviation then two diamonds ( ), and where the difference was greater than one quarter of a standard deviation but less than half of a standard deviation, one diamond ( ) has been entered. Where the difference was smaller than one quarter of a standard deviation, the difference cell has been left blank. 226 Vietnam Reading and Mathematics Assessment Study Table 8.1: Differences between most effective and least effective schools in reading and math Reading Math Variable Bottom 100 Top 100 All Bottom 100 Top 100 All Difference Difference Mean Mean SD Mean Mean SD Pupil characteristics Pupil age (in months) 139.97 137.58 6.58 -0.36 140.11 138.11 6.58 -0.3 Pupil sex (ratio female) 0.47 0.46 0.12 -0.05 0.51 0.47 0.12 -0.32 Family time (mins. Per day) 112.28 116.11 26.62 0.14 116.07 116.29 26.62 0.01 Private corner for study at home (ratio) 1.71 1.84 0.22 0.62 1.72 1.79 0.22 0.33 Meals per day 2.48 2.8 0.3 1.07 2.47 2.73 0.3 0.87 Travel time to school (mins.) 20.65 17.96 6.83 -0.39 22.1 18 6.83 -0.6 Days absent last month 1.1 0.62 0.62 -0.78 0.86 0.63 0.62 -0.36 Grade repetition (ratio repeating) 0.3 0.28 0.23 -0.12 0.33 0.29 0.23 -0.16 Borrow books 0.26 0.5 0.36 0.66 0.28 0.46 0.36 0.51 Materials 7.46 8.11 0.96 0.68 7.67 8.1 0.96 0.44 Textbooks 14.43 15.86 3.61 0.4 14.23 16.5 3.61 0.63 Homework and extra tuition Ensure homework done 1.34 1.55 0.36 0.58 1.35 1.5 0.36 0.42 Help with homework 0.81 0.91 0.39 0.25 0.79 0.9 0.39 0.26 Home pays attention to work at school 2.39 2.54 0.33 0.44 2.38 2.52 0.33 0.41 Get Math homework 3.39 3.87 0.38 1.26 3.48 3.86 0.38 0.99 Math homework corrected 3.82 4.14 0.45 0.72 3.84 4.16 0.45 0.71 Get Reading homework 3.67 3.97 0.21 1.43 3.76 3.96 0.21 0.96 Reading homework corrected 4.02 4.23 0.38 0.53 4.01 4.23 0.38 0.58 Hours extra tuition (Vietnamese) 3.62 3.82 1.86 0.1 3.42 4.12 1.86 0.37 Hours extra tuition (Math) 4.02 3.85 1.94 -0.09 3.34 4.13 1.94 0.41 Hours extra tuition V and M 6.8 7.71 3.44 0.27 6.18 8.69 3.44 0.73 Textbooks and subjects studied Have Vietnamese textbooks 4.55 4.96 1.01 0.41 4.51 5.04 1.01 0.52 Have Math textbooks 2.5 2.86 1.01 0.36 2.38 3.02 1.01 0.64 Have Other textbooks 7.38 8.04 2.03 0.32 7.34 8.44 2.03 0.54 Use own Vietnamese textbooks 1.95 1.98 0.05 0.72 1.96 1.99 0.05 0.6 Use own Math textbooks 1.94 1.99 0.06 0.74 1.96 1.99 0.06 0.61 Opportunity to learn 8.65 8.74 0.47 0.18 8.76 8.76 0.47 -0.01 School head's characteristics Head sex (ratio female) 0.2 0.42 0.48 0.46 0.23 0.38 0.48 0.32 Ethnic group of head (ratio non-Kinh) 1.85 1.82 0.31 -0.1 1.84 1.82 0.31 -0.06 Head Academic Education 12.15 12.31 2.12 0.08 12.06 12.08 2.12 0.01 Head TTR 2.23 2.49 0.91 0.29 2.18 2.54 0.91 0.39 Head teaching (hours per week) 1.3 2.03 1.74 0.42 1.31 2.03 1.74 0.41 Head total possessions 10.36 11.13 3.37 0.23 10.23 11.54 3.37 0.39 Characteristics of school Age of school 21.73 27.51 16.95 0.34 23.4 26.22 16.95 0.17 Distance from civilization 7.53 5.57 5.12 -0.38 6.32 5.72 5.12 -0.12 Permanent teachers 23.45 24.65 11.47 0.1 25.73 27.31 11.47 0.14 Temporary teachers 4.64 2.16 3.89 -0.64 4.42 1.62 3.89 -0.72 Staff education 10.33 10.41 2.14 0.03 10.13 10.44 2.14 0.14 Staff TTR 1.72 2.05 0.33 0.97 1.75 2.07 0.33 0.98 Total enrolment (school size) 715.64 620.81 371.5 -0.26 742.34 697.28 371.47 -0.12 Grade 5 enrolment (grade size) 125.18 118.55 77.94 -0.09 137.41 136.19 77.94 -0.02 Satellite campuses 3.32 2.6 2.42 -0.3 3.72 3.07 2.42 -0.27 Class size 29.14 28.66 5.66 -0.08 28.78 28.76 5.66 0 Grade 5 class size 30.04 29.68 6.45 -0.05 30.48 30.11 6.45 -0.06 Square meters per pupil 1.25 1.36 0.8 0.13 1.27 1.31 0.8 0.06 Condition of school building 2.26 2.56 1.23 0.25 2.43 2.51 1.23 0.06 School resources 7.16 9.91 4.8 0.57 7.45 9.31 4.8 0.39 Pupils can take books home 0.49 0.55 0.49 0.12 0.43 0.5 0.49 0.14 Importance of Head's activities Contacts with local community 2.63 2.49 0.5 -0.27 2.6 2.55 0.5 -0.1 Monitoring pupil's progress 2.57 2.5 0.5 -0.13 2.63 2.61 0.5 -0.04 Administrative tasks 2.65 2.84 0.46 0.41 2.7 2.75 0.46 0.11 Observe/discuss lessons with staff 2.66 2.55 0.5 -0.21 2.71 2.61 0.5 -0.2 Professional development of teachers 2.71 2.74 0.43 0.06 2.78 2.76 0.43 -0.05 Self development of Head 2.6 2.59 0.5 -0.02 2.66 2.55 0.5 -0.22 Organize extra curricular activities 2.08 2.12 0.35 0.11 2.16 2.14 0.35 -0.06 Improve school conditions 2.73 2.73 0.43 0.01 2.79 2.74 0.43 -0.12 Organize excellent competitions 2.44 2.56 0.51 0.23 2.54 2.56 0.51 0.04 School activities School magazine 1.2 1.37 0.63 0.27 1.17 1.35 0.63 0.29 227 Vietnam Reading and Mathematics Assessment Study Story telling competition 1.88 2.17 0.52 0.57 1.89 2.15 0.52 0.5 General knowledge quizzes 1.92 2.08 0.66 0.24 1.95 2.06 0.66 0.17 Camping expeditions 1.8 1.92 0.59 0.2 1.85 1.86 0.59 0.02 School day celebrations 1.57 1.8 0.72 0.31 1.52 1.71 0.72 0.27 Open meeting school council 2.21 2.32 0.51 0.21 2.31 2.34 0.51 0.06 Community contribution to school Help building 1.58 1.82 0.46 0.52 1.68 1.83 0.46 0.33 Maintenance of school facilities 1.69 1.81 0.42 0.29 1.67 1.75 0.42 0.19 Construction or repair of furniture 1.37 1.48 0.5 0.22 1.34 1.46 0.5 0.24 Donate books 1.17 1.28 0.42 0.26 1.2 1.29 0.42 0.21 Stationary 1.11 1.12 0.33 0.03 1.12 1.09 0.33 -0.09 Salary of additional teachers 1.06 1.06 0.28 0 1.03 1.11 0.28 0.28 Additional salary 1.05 1.13 0.31 0.26 1.01 1.16 0.31 0.48 Salary of non-teaching staff 1.09 1.25 0.39 0.41 1.12 1.29 0.39 0.43 Additional salary for NT staff 1.04 1.08 0.32 0.12 1.05 1.1 0.32 0.16 Extra curricula activities 1.12 1.2 0.45 0.18 1.07 1.19 0.45 0.27 Assistance of teachers 1.16 1.19 0.42 0.07 1.18 1.24 0.42 0.14 Provisions of school meal 1.02 1.04 0.17 0.12 1.03 1.05 0.17 0.12 Land for housing 1.34 1.37 0.43 0.07 1.34 1.38 0.43 0.09 Land for farming 1.05 1.1 0.24 0.21 1.04 1.08 0.24 0.17 Pupil and teacher behaviours seen to be problem by school head Total number of pupil problems 0.3 0.23 0.17 -0.43 0.3 0.25 0.17 -0.31 Total teacher problems 0.22 0.13 0.16 -0.55 0.22 0.14 0.16 -0.53 Teacher characteristics Teacher sex (ratio female) 0.5 0.76 0.35 0.75 0.49 0.78 0.35 0.84 Teacher ethnic (ratio Kinh) 1.86 1.82 0.27 -0.13 1.85 1.83 0.27 -0.06 Teacher education 11.51 11.71 1.14 0.17 11.47 11.88 1.14 0.36 Teacher TTR 1.82 2.13 0.46 0.66 1.87 2.19 0.46 0.67 Teacher experience (years) 10.29 11.27 6.24 0.16 9.86 11.28 6.24 0.23 Teacher excellent (ratio) 0.56 0.64 0.39 0.19 0.53 0.72 0.39 0.48 Teacher perceives in-service effect 3.23 3.19 0.52 -0.07 3.22 3.21 0.52 -0.02 Teacher total possessions 9.94 10.59 3.07 0.21 9.26 11.17 3.07 0.62 Hours extra work for money 1.71 1.45 1.74 -0.15 1.68 1.42 1.74 -0.15 Teaching resources Books in classroom bookcase 1.56 10.36 32.07 0.27 2.46 14.88 32.07 0.39 Classroom furniture (Max 10) 4.6 5.39 1.37 0.58 4.63 5.37 1.37 0.54 Classroom supplies (max 6) 5.01 5.07 1.02 0.05 4.9 5.36 1.02 0.45 Classroom resources (max 16) 9.62 10.46 1.98 0.42 9.54 10.72 1.98 0.6 Teachers' teaching Teacher hours teaching per week 16.25 16.39 2.36 0.06 16.37 16.82 2.36 0.19 teacher work per week 36.21 37.38 6.14 0.19 36.66 37.59 6.14 0.15 Meeting parents 2 2.09 0.35 0.25 2.03 2.11 0.35 0.24 Percent of parents met 75.53 82.45 16.5 0.42 76.26 85.43 16.5 0.56 Teacher testing reading 18.39 20.3 11.47 0.17 19.78 22.72 11.47 0.26 Teacher testing Math 18.29 20.35 8.9 0.23 18.85 23.06 8.9 0.47 Advice from Head 2.29 2.47 0.65 0.27 2.37 2.51 0.65 0.22 Times observed by colleagues 1.81 2.03 1.17 0.19 1.58 2.08 1.17 0.42 Teachers perception of inspectors Number of inspection since 1998 2.19 2.42 2.43 0.1 2.17 3.1 2.43 0.38 Advises me 1.56 1.69 0.44 0.28 1.6 1.73 0.44 0.29 Criticizes me 1.23 1.27 0.41 0.1 1.28 1.33 0.41 0.13 Suggests new ideas 1.43 1.56 0.45 0.29 1.46 1.59 0.45 0.29 Clarifies new objectives 1.46 1.58 0.45 0.27 1.45 1.65 0.45 0.44 Explains curriculum 1.38 1.45 0.43 0.16 1.41 1.54 0.43 0.28 Recommends new materials 1.38 1.46 0.43 0.19 1.38 1.46 0.43 0.19 Provides information for self improve 1.48 1.61 0.44 0.29 1.49 1.7 0.44 0.47 Contributes little to teaching 1.76 1.89 0.29 0.43 1.82 1.92 0.29 0.36 Suggests teaching methods 1.68 1.79 0.39 0.29 1.69 1.91 0.39 0.56 Encourages contacts with teachers 1.38 1.58 0.44 0.44 1.42 1.65 0.44 0.53 Provides in-service training 1.51 1.66 0.44 0.33 1.57 1.78 0.44 0.48 Finds faults and reports 1.31 1.41 0.43 0.23 1.38 1.5 0.43 0.26 Others and achievement scores School location 1.75 1.9 0.59 0.25 1.81 1.91 0.59 0.17 Pupil reading 500 score 360.34 701.38 77.9 4.38 386.7 632.78 77.9 3.16 Pupil Math 500 score 385.39 644.07 82.57 3.13 356.06 703.59 82.57 4.21 Teacher reading 500 score 591.59 655.8 66.09 0.97 565.41 654.37 66.09 1.35 Teacher Math 500 score 623.94 716.25 87.97 1.05 588.82 732.69 87.97 1.64 228 Vietnam Reading and Mathematics Assessment Study It can be seen that, for the most part, where there was a noticeable difference In general, variables on a variable for reading this difference could also be seen for mathematics noteworthy differences and where there was no difference for reading there was no difference for for reading also had mathematics. This form of replication is reassuring. noteworthy differences for mathematics. Replication of results with most and least effective 200 schools and in 30 most and least effective isolated schools Before commenting on the findings it would be wise to replicate the analysis The exercise was using the 200 most effective schools and 200 least effective schools and again repeated for the 200 by taking only the 534 isolated schools, examine the differences between the most effective and the 30 most effective and 30 least effective schools. These differences have been 200 least effective presented in Table 8.2 below. schools and also for the 30 most effective and 30 Table 8.2: Differences between top and bottom groups for the top and least effective isolated bottom 200 schools and the top and bottom 30 schools in isolated areas. schools. 200 most and least effective schools 30 most and least effective schools in isolated Variable Reading Math Reading Math Difference Difference Difference Difference Pupil characteristics Pupil age (in months) -0.24 -0.37 0.21 0.06 Pupil sex (ratio female) -0.13 -0.24 0.13 0.39 Family time (mins. Per day) 0.14 0.21 0.51 0.62 Private corner for study at home (ratio) 0.54 0.58 0.11 0.51 Meals per day 0.95 0.88 0.68 0.63 Travel time to school (mins.) -0.28 -0.34 0.06 -0.23 Days absent last month -0.65 -0.52 -0.19 -0.16 Grade repetition (ratio repeating) -0.06 -0.19 0.47 0.24 Borrow books 0.52 0.53 0.67 0.7 Materials 0.5 0.35 0.14 0.27 Textbooks 0.35 0.57 -0.42 -0.16 Homework and extra tuition Ensure homework done 0.42 0.52 0.16 -0.01 Help with homework 0.35 0.38 -0.41 -0.11 Home pays attention to work at school 0.41 0.46 -0.08 -0.13 Get Math homework 1.01 1.11 0.84 0.88 Math homework corrected 0.53 0.62 0.45 0.61 Get Reading homework 1.22 0.92 1.03 0.89 Reading homework corrected 0.41 0.46 0.34 0.28 Hours extra tuition (Vietnamese) 0.21 0.14 0.65 0.76 Hours extra tuition (Math) 0.06 0.02 0.19 0.64 Hours extra tuition V and M 0.36 0.22 0.92 1.2 Textbooks and subjects studied Have Vietnamese textbooks 0.31 0.5 -0.31 -0.02 Have Math textbooks 0.32 0.5 -0.07 0.28 Have Other textbooks 0.31 0.51 -0.53 -0.38 Use own Vietnamese textbooks 0.46 0.44 0.44 0.4 Use own Math textbooks 0.47 0.46 0.66 0.33 Opportunity to learn 0.2 0.14 -0.43 -0.23 School head's characteristics Head sex (ratio female) 0.44 0.45 0.46 0.15 Ethnic group of head (ratio non-Kinh) -0.18 -0.06 -0.15 0.15 Head Academic Education -0.06 0.04 -0.07 0.07 Head TTR 0.18 0.34 0.12 0.23 Head teaching (hours per week) 0.4 0.46 0.58 0.49 Head total possessions 0.25 0.3 -0.06 0.46 Characteristics of school Age of school 0.37 0.25 0.66 0.47 229 Vietnam Reading and Mathematics Assessment Study Distance from civilization -0.26 -0.15 -0.12 0.75 Permanent teachers 0.1 0.06 0.1 0 Temporary teachers -0.33 -0.42 -0.37 -0.39 Staff education 0.07 0.16 -0.24 0 Staff TTR 0.73 0.81 0.65 0.63 Total enrolment (school size) -0.18 -0.22 -0.44 -0.47 Grade 5 enrolment (grade size) -0.07 -0.09 -0.32 -0.45 Satellite campuses -0.14 -0.25 0.25 0.15 Class size -0.1 -0.1 -0.75 -0.69 Grade 5 class size 0 -0.04 -0.45 -0.67 Square meters per pupil 0.15 0.08 0.02 0.17 Condition of school building 0.07 0.12 -0.33 -0.16 School resources 0.44 0.47 0.39 0.27 Pupils can take books home 0.01 0.17 0.18 0.07 Importance of Head's activities Contacts with local community -0.12 -0.1 0.02 -0.14 Monitoring pupil's progress -0.16 -0.18 -0.21 -0.27 Administrative tasks 0.27 0.06 0.57 0.07 Observe/discuss lessons with staff -0.04 -0.07 -0.07 0 Professional development of teachers 0.16 0.01 -0.17 -0.23 Self development of Head 0.07 -0.05 0 -0.26 Organize extra curricular activities 0.1 -0.06 0 -0.25 Improve school conditions 0 -0.17 0.1 -0.24 Organize excellent competitions 0.2 0.12 0.34 0.26 Comments on the results The results for the differences in variables for the 100 and 200 most effective and least effective schools were basically the same. The differences between the 200 most effective and least effective schools have been used for comments. The results for the isolated schools only were somewhat different. Comments on the results have been given below. Pupils in the most Pupil characteristics: There was no difference between the two effective schools were: groups in the distribution of the sexes, in the amount of time that younger, less absent, pupils devoted to helping their families, and in grade repetition. The eating more meals per pupils in the least effective schools tended to be older than those in day, had private study the most effective schools and also to have been absent more days corner at home, had but these were not big differences. Pupils in the most effective more textbooks, and schools tended more to have a private corner at home where they could borrow books could study, to have more meals per day, and to have been absent more from school less. Their perception of being able to borrow books from school, library. and having more mathematics textbooks was greater than the pupils in the least effective schools. It should be noted that in Vietnam it is the responsibility of the parents to buy the materials and textbooks for their children. The provincial authorities should provide the textbooks only for pupils in very disadvantaged circumstances. For the isolated schools' differences, pupils in the most effective isolated 230 Vietnam Reading and Mathematics Assessment Study schools spent more time helping the family, had more meals per day, and could borrow books more than pupils in the least effective isolated schools. Thus, every effort should be made to ensure that all pupils have a private corner for study at home, eat regularly, are absent from school less, and have the required materials and textbooks. Parental help and extra tuition: There were major differences between the They also got more two groups of schools in the homework variables, especially for homework and had it mathematics. Those pupils in the most effective schools tended to get more corrected, and received homework set by their teachers, have parents who ensured that it was done more support for school and teachers who corrected more of the homework. The parents of these work from their families. children also 'paid more attention' to school work in general. The most In isolated schools, effective schools had pupils who tended to attend more extra tuition classes pupils in more effective in Vietnamese and mathematics. schools had more extra tuition. In isolated schools, pupils in the more effective schools were given more homework, had it corrected more, and received more extra tuition than the pupils in the least effective schools. Homework is an area of education where homes and schools can work together. In this Grade 5 survey it has been seen that it is important for the teachers to give homework, for the children to do it, for the parents to check that it has been done, and for the teachers to correct all of the homework. The importance of giving and correcting all homework should be stressed more in the pre- and in-service training of teachers. The management courses for head should also stress this aspect of schooling. If the school heads in all schools emphasise to the parents that checking that their children have done their homework is important, this may help. Alternatively, having parents sign the homework can sometimes be a useful way of ensuring this. It would be worthwhile considering devising various strategies to have this checking occur, evaluate the effect of the different strategies and then implement the most successful one. The following suggestions have been already made, but it is worth re-iterating that the importance of giving and correcting homework should be stressed in all teacher training courses and school head management courses. The inspectorate should ensure that homework is given regularly and corrected. Finally, a special study should be made of extra tuition in schools in isolated areas. Textbooks and subjects studied: In the most effective schools for reading Pupils in more effective comprehension in Vietnamese, the average pupil either had his or her own schools tended to study textbook or shared a textbook with one other person whereas in the less slightly more subjects effective schools there were some pupils who had to share with more than and had to share their one other pupil. This was true for both subject areas. There are nine subjects textbooks less. 231 Vietnam Reading and Mathematics Assessment Study in the curriculum but in the past not all schools have taught all subjects because the introduction of some were optional at the time of first introduction. In the 100 most and least effective school comparison, it can be seen that the average child studied only 8.74 subjects in the most effective schools and 8.65 in the least effective schools in Vietnamese. However the difference was small and there were no differences between the more effective and less effective schools in mathematics. In the isolated schools it was interesting to note that the pupils in the most effective isolated schools had fewer other textbooks but shared them less than the pupils in the least effective isolated schools. Again, it has been mentioned earlier, but it would be highly desirable for the curriculum centre to review the actual textbooks required. More effective schools School head's characteristics: the most effective schools tended to have more had heads who were school heads who were female, heads with more teacher training, heads who female, with more tended to teach more each week, and heads who had slightly more teacher training, and possessions at home. There was no difference in the ethnic group to which who taught more hours the heads belonged, nor in the amount of academic education they had per week. received. In general, though, the differences between the school head's characteristics were not great. In effective isolated In isolated schools it was the amount of teaching per week that the heads schools, school heads carried out that emerged as important. Pupils in the most effective isolated taught more hours per schools tended to have school heads who were teaching more hours than week. those in the least effective isolated schools. Presumably the least effective schools were in poorer locations so that it was not conducive for female heads to be there. Pupils in the most effective schools tended to have teachers teach more in a week. It can be argued that heads need to teach each week so that they are au fait with what is going on in the school and the types of pupils that the school has. This can only be gained through teaching. Policy suggestion 8.1: The education advisors and the management course personnel should encourage heads to teach each week. There were either only Characteristics of the school: It was re-assuring that there were no, or very small differences or no small, differences in the number of permanent teachers, grade 5 enrolment, differences between the class size, and square metres per pupil. This shows that the various levels of two group of schools for educational offices had done a good job, Although there were no differences the number of between the most and least effective schools in the head's perceived condition permanent teachers, of the school buildings and the rule about pupils being able to take book Grade 5 enrolment, home to read, it can be argued that the levels of both of these variables was class size, square meters too low. Even in the most effective schools the level of the conditions of the per pupils, rule for school buildings was between 2.40 and 2.50 signifying classrooms needing 232 Vietnam Reading and Mathematics Assessment Study either major or minor repairs. Only about 50 percent of pupils were in schools borrowing books, and where the heads said that they could take books home. This means that there perceived conditions of were fifty percent of schools where either there were no libraries or pupils school buildings. (but were unable to borrow the books to take home. very low levels for the last two variables.) The variables where there was some replication in the differences In more effective between the most and least effective schools were: the age of the school schools, there were (pupils in more effective schools tended to be in older schools, the fewer temporary distance from 'developed' social and commercial facilities in isolated teachers, the whole areas (but with a difference in sign for reading and mathematics in teaching staff had more isolated areas), the percentage of temporary teachers (the least effective teacher training, there schools having more); the amount of teacher training the whole staff had were lower enrolment, had (most effective schools having staff with more years of teacher fewer satellite training than the total staff in the least effective schools); total enrolment campuses, and more and Grade 5 enrolment in isolated schools (with the least effective school resources. schools having higher enrolments), having satellite campuses in isolated areas (least effective schools having more), and the number of items of school resources where the most effective schools had two to three more items of school resources than the least effective schools. These aspects of schools are important and require a long term strategy to correct the shortcomings in the least effective schools. Each province and indeed district should have a plan to reduce the number of temporary teachers and replace them with permanent teachers. Care should be taken to replace the temporary teachers with teachers who have had the full teacher training. At the same time, the planners in the provincial offices should examine the allocation of teachers to schools to ensure a more equitable distribution of permanent and fully teacher trained teachers to all schools. The lack of school resources is a problem. A study should be undertaken to identify those key resources needed in any school. Head's perceived importance of head teachers' activities: it was of interest that there were no differences between the heads of the most and least effective schools. School activities: The group of persons on the questionnaire committee for More effective schools the study identified six different activities that occur in Vietnamese primary had a school magazine schools that might be though to effect achievement in reading and produced by pupils and mathematics. There was quite a bit of disagreement about the variables. story-telling Information was collected on all of them. In other studies it has been found competitions, both that where the pupils are actively involved in the writing and producing of a reading related school magazine the reading achievement of schools where this occurs is activities. higher. In Vietnam it can be seen that there were differences between the most effective and least effective schools in the extent to which the school produced a school magazine, and the extent to which it held story telling 233 Vietnam Reading and Mathematics Assessment Study competitions. In other words, where the head and staff fostered those activities concerned with thinking and reading, the pupils in those schools achieved better. There were no consistent differences in values for camping expeditions, school day celebrations and having an open meeting of the school council. It is worth re-iterating that all teacher training courses and school heads' management courses should encourage the kinds of school-wide activities that foster intellectual development and especially in reading comprehension and mathematics. Parents of pupils in Community contribution to school: communities differ in the amount of more effective schools help they offer to schools. There were some small differences between gave more contributions. the most and least effective schools on one or two aspects of community/parental contribution. There was a higher percentage of pupils from the most effective schools whose parents helped in the payment of non-salaried staff, and to a lesser extent by building extra classrooms, and donating books. There is evidence, not only from Vietnam but from other countries, that the more a community feels and is involved in school matters the better the achievement in the school. Attention should be given to various activities that schools can undertake in order to have parents (and others in the community) play an active role in helping the school. This should be not only in the form of contributions to the school but also in the form of help of all kinds including acting as teaching assistants when this is required. Devising the activities involving parents in the lives of schools in all areas in the country would appear to be an important issue. The more effective Teacher characteristics: The most effective schools had more female schools had more female teachers, more teachers with a higher level of teacher training, more teachers teachers, with more being designated as 'excellent', having more education and having more teacher training, and possessions at home (for mathematics). It was interesting to note that the more of them being teachers in the least effective isolated schools devoted more hours to extra designated as excellent. work for money. One possible mechanism for improving the allocation of teachers to schools will be to create provincial data bases for planners to use for the re-allocation and new allocation of teachers to schools. Effective schools had Teaching resources: There were differences in the amount of classroom more teaching resources. resources between the most and least effective schools. This concerned classroom furniture and the books in the classroom bookcase with the most effective schools having more of them. 234 Vietnam Reading and Mathematics Assessment Study Teachers' teaching: There was no difference between the two groups of Teachers in effective schools for the number of hours per week teaching. However, it should be schools did not teach noted that the number of hours teaching was well below the international more hours but did meet average - even in countries where it is expected that teachers do extra work a higher percentage of in order to earn enough money. There were differences for the percentage of pupils' parents, were pupils' parents that were met by the teacher (the percentage was higher in the observed more most effective schools), in the number of times the classes were observed by frequently by colleagues, colleagues (in the most effective schools this occurred more frequently) and and received more amount of feedback from the principal to the teachers about their teaching (in feedback from school the most effective schools this occurred more frequently). In isolated schools and especially for mathematics the pupils in the most effective schools had heads about their teachers who met parents more, met a higher percentage of pupils' parents, teaching. tested more frequently, and were observed more frequently by their peers. It would seem that the teachers meeting a high percentage of pupils' parents and peer observation were associated with more effective schools and these two aspects of teacher behaviour should be encouraged. Peer observation of teachers teaching and the meeting of all parents should be stressed in teacher training courses and in management courses. To ensure that teachers recognise how these two activities can be preformed efficiently, The MOET could profitably prepare two booklets that offer specific guidance and illustrative examples of the conduct of peer observation and of conferring with parents. The booklets could be distributed in pre-service and in-service courses for teachers and school heads. Teachers' perceptions of inspectors: There were some small differences In more effective between the two groups of schools where the teachers in the most effective schools, the teachers schools had a more positive view of the visits of the inspectors. In had slightly more mathematics the teachers in the effective schools perceived inspectors to be positive view on suggesting teaching methods and encouraging contacts with other teachers inspectors. more than the teachers in the least effective schools perceived this to be happening. In isolated schools it was the providing information for self- improvement, encouraging contacts with other teachers, suggesting teaching methods and providing in-service training that were important. Heads' perceptions of pupil and teacher problems: a series of questions was Heads in more effective asked of the head about the frequency with which the Head had to deal with schools had to deal less behavioural problems. 'Never' was coded as 1 and 'sometimes or often' as 2. often with pupil and The first problem in the table above is to do with pupils arriving late at school. The figure of 1.8 in the most effective schools column means that 80 teacher behavioural percent of pupils were where the head said that the problem occurred problems. sometimes or often. The major differences between the most effective and least effective schools were for pupil absenteeism, pupils skipping classes, pupils dropping out of school, classroom disturbance by pupils, and vandalism and stealing. Health problems had very high frequencies in both groups. The major behaviour problems of teachers where there was a difference between the two types of schools were being late to school, being absent, skipping classes, alcohol abuse or possession. Again there was a high frequency for teacher health problems although there was not a difference between the most and least effective schools. 235 Vietnam Reading and Mathematics Assessment Study Teachers in more Teacher reading and math competence: The teachers in the most effective effective school had schools in reading and in mathematics had much more competence in both more subject matter subject areas. The difference of one standard deviation between the two knowledge. This was a groups of teachers (in the most and in the least effective schools) is very high. very large difference. Policy suggestion 8.2: The MOET should mount a special study in order to examine in detail the kinds of behavioural problems expressed by heads about both pupils and teachers. What were the most important variables to emerge form the analyses? But before commenting on the most important differences, it would be re- assuring if the same variables were to emerge from an analysis of the differences between the most and least effective 200 schools? Where there was a difference of more than 0.5 SD on a variable between the most and the least effective schools both in Reading and Math, the variable has been listed in the table below. Where this difference existed for both reading and math, the cells have been shaded. Table 8.3: Variables for different analyses having more than 0.5 SD difference 100 schools 200 schools 30 isolated schools Variable Read Math Read Math Read Math Home Family time (mins. Per day) Private corner for study at home (ratio) Meals per day Days absent last month Hours extra tuition (Vietnamese) Hours extra tuition V and M Use own Vietnamese textbooks Use own Math textbooks School Staff TTR Temporary teachers Class size Story telling competition Total number of pupil behavioural problems Total number of teacher behavioural problems Borrow books Teacher Teacher sex (ratio female) Teacher TTR Classroom furniture (Max 10) Meeting parents Percent of parents met Get Math homework Math homework corrected Get Reading homework Reading homework corrected Perceive inspectors as encouraging contacts with teachers Teacher's reading score Teacher's math score 236 Vietnam Reading and Mathematics Assessment Study It was interesting to note that the home variables emerging as important were different for the 100 and 200 school comparisons on the one hand and, on the other hand for the 30 isolated school comparisons except for the variable 'meals per day.' Between 100 the most and the least effective school, there were differences in 'using own Vietnamese and Math textbooks.' 'Private corner for study at home' and 'days absent last month' were the variables with big differences between 200 the most and least effective school. 'Time helping family' and 'extra tuition' came up for the isolated schools but 'days absent last month' did not. Among the school variables, 'average staff teacher training' and 'pupil The most important can borrow books' came up for any comparisons. The 'number of variables (more than 0.5 temporary teachers' only arose in the 100 schools comparison, class size standard deviation only occurred in the isolated schools and story telling only in the 100 difference) were singled out for further analyses school comparison. The 'total number of teachers' behavioural problems' (see Chapter 9). were for both 100 and 200 schools, whereas the 'total number of pupil behavioural problems' was only for 200 schools comparison. Amongst the teacher variables the 'teacher subject matter knowledge' emerged as a large difference between the least and most effective schools in all comparisons, but 'meeting parents' and the 'inspectors encouraging teachers to meet other teachers' only came through in the isolated schools comparisons. The 'percentage female', the 'teachers' level of teacher training' and 'classroom furniture' came through only for 100 and 200 school comparisons. "Get homework' both in Vietnamese and Math emerged in all comparisons, 'math homework corrected' for both the 100 and 200 school comparisons, and 'reading homework corrected' only for 100 school comparison. Summarizing the above, in terms of school and teacher variables, it can be said that similar findings emerged from the comparison of 100 and 200 schools' comparisons, but when only isolated schools were considered there were different variables that emerged. It is the variables that came through for both the 100 school and 200 school comparison that will be carried through to the next chapter where multivariate analyses will be reported. The differences of the most effective and least effective isolated schools should form the basis of a separate report that involves some observation work in some if those schools. 237 Vietnam Reading and Mathematics Assessment Study Policy suggestion 8.3: The effective schools in isolated areas are of great interest. Some of the variables that are different between more and least effective schools have been identified. However, there must be other factors in the most effective schools that have not been and are important. It is important to know what these are in order for all isolated schools to be improved. The MOET may well wish to consider undertaking a small observational study of selected most effective schools in isolated areas. Conclusion This chapter has been concerned with the identification of variables associated with the differences between effective and ineffective schools. An 'effective' school was one whose achievement was above that which could be expected from the type of pupil intake into the school. An 'ineffective' school was one whose achievement was much worse than could be expected given the pupil intake into the school. The quality of the pupil intake into the school was measured by a composite home background consisting of total possessions of the pupil's home, whether the pupil was Kinh or non-Kinh, the extent to which the pupil spoke Vietnamese outside the school, number of books at home, and the level of the parents' education. The 100 most effective and 100 least effective schools were identified and the differences on many independent variables were calculated in terms of the overall standard deviation. This exercise was repeated for the 200 most effective and least effective schools. This was repeated again for the 30 most effective and 30 least effective schools in isolated areas. In all there were only 534 isolated schools in the sample. The variables that were associated with differences between the most effective and least effective schools of greater than half an overall standard deviation for the 100 and 200 schools' comparisons have been presented below. Home Private corner for study at home (ratio) Meals per day Days absent last month Use own Vietnamese textbooks Use own Math textbooks School Staff TTR Temporary teachers 238 Vietnam Reading and Mathematics Assessment Study Story telling competition Total number of pupil problems Total teacher problems Borrow books Teacher Teacher sex (ratio female) Teacher TTR Classroom furniture (Max 10) Percent of parents met Get Math homework Math homework corrected Get Reading homework Reading homework corrected Teacher's reading score Teacher's math score These, then, were the variables that had the largest differences after the home background had been removed. Pupils who had a private corner at home in which to study, had three meals per day, were not absent very much, could borrow books from either the school or the classroom to take home and read, who received homework often, who were in schools with teachers who had more teacher training, where the head perceived that there were fewer teacher behavioural problems, where there was more classroom furniture, and where the teachers had a good knowledge of their subject matter all stood a better chance of receiving a better education than pupils who were not in such homes and schools. However, before rushing to any conclusions, further analyses are required. These will be reported in the next chapter. 239 Vietnam Reading and Mathematics Assessment Study Appendix 8.1: Means and sampling errors of school resources by region and school location Variables Singleton Multiple Original Ratio Recoded Composite Pupil characteristics 1 Pupil age in months 2 Pupil sex: 1=boy and 2=girl 3 Family time: number of minutes spent each day helping family 4 Private corner for study at home: 1= do not have; 2 = have 5 Meals per day 6 Travel time to school in minutes 7 Days absent last month 8 Grade repetition: number of times having repeated a grade 9 Borrow books: can take home books from school or class library: 0=cannot borrow and 1= can borrow 10 Materials: number of items out of 10 for doing school work 11 Textbooks: number of textbooks out of 29 Homework and extra tuition 12 Homework done: 0=don't get homework or never, 1=sometimes, 3=most of the time 13 Help with homework: same as for homework done 14 Home pays attention: extent to which someone at home 'pays attention' to homework 15 G et Math homework: 1=no homework 2=once or twice per month 3= about once a week 4= three or four times per week 16 Math homework corrected: 1=no homework 2-never corrected 3= sometimes corrected 4= most homework corrected 5=all homework corrected 17 Get Reading homework: 1=no homework 2=once or twice per month 3= about once a week 4= three or four times per week 18 Reading homework corrected: 1=no homework 2-never corrected 3= sometimes corrected 4= most homework corrected 5=all homework corrected 19 Hours per week extra tuition in Vietnamese 20 Hours per week extra tuition in Mathematics 21 Total hours extra tuition in Vietnamese and math Textbooks and subjects studied 22 Having Vietnamese textbooks (Max:8) 23 Having Math textbooks (Max:5) 24 Having Other textbooks (Max: 16) 25 Using own Vietnamese textbooks: 0=have none 1=share with others 2=share with one other 3=have own 26 Using own Math textbook: 0=have none 1=share with others 2=share with one other 3=have own 27 Opportunity to learn: Number of subjects studied (out of 9) School head's characteristics 28 Head sex: sex of school head: 1=male 2=female 29 Ethnic group of head: 1=other 2=Kinh 30 Head academic education: Number of years of academic education 31 Head TTR: Number of years of teacher training 32 Head teach: number of hours per week taught by head 33 Head total possession: number of possession in the home of head (out of 20) 240 Vietnam Reading and Mathematics Assessment Study Characteristics of school 34 Age of school: number of years old the school is 35 Distance from civilisation: the average number of kilometres from the school to a health clinic, the nearest road for cars, the public library, book shop, lower secondary school, and market 36 Permanent teachers: number of permanent teachers in school 37 Temporary teachers: number of temporary teachers per school 38 Staff education: average number of years academic education of staff 39 Staff teacher training: average years of teacher training of staff 40 Total enrolment (sometimes called 'school size'): total number of boys and girls in school 41 Grade enrolment: total number of boys and girls in Grade 5 42 Average class size (sometimes called 'class size'): enrolment divided by number of classes 43 Average Grade 5 class size: enrolment of Grade 5 divided by number of classes in Grade 5 44 Satellite campuses: number of satellite campuses per school 45 School area: square metre per pupil of instructional space 46 School head's perception of condition of school building: 1=school needs complete rebuilding, 2=some classrooms need major repair, 3=most or all classrooms need minor repairs, 4=some classrooms need minor repairs and 5=in good condition 47 School resources: number of school resources (out of 27) 48 School borrow: school head says that pupils may take books home from library Importance of school head's activities [1=not important, 2=important, and 3=very important] 49 Contacts with local community 50 Monitoring pupils' progress 51 Administrative tasks 52 Observing and discussing lessons with staff 53 Professional development of teachers 54 Self development of school head 55 Organising extra curricular activities 56 Improving school conditions and teacher living conditions 57 Organising 'excellence' competitions School activities to engender learning [1=never, 2=once per year, and 3=more than once per year] 58 School magazine - production of 59 Organising story-telling competitions 60 General knowledge quizzes 61 Organising camping expeditions 62 School day celebrations 63 Open meeting of school council Community contribution to school [1=no and 2=yes] 64 Building of school facilities such as classrooms, teacher houses 65 Donation of books to school 66 Payment of salaries on non-teaching staff 67 Help building 68 Maintenance of school facilities 69 Construction or repair of furniture 70 Donate books 71 Stationary 72 Salary of additional teachers 73 Additional salary 74 Salary of non-teaching staff 75 Additional salary for NT staff 76 Extra curricula activities 77 Assistance of teachers 78 Provisions of school meal 79 Land for housing 80 Land for farming Pupil and teacher behaviours seen to be problem by school head 81 Average occurrence of fourteen pupil problems 82 Average occurrence of nine teacher problems 241 Vietnam Reading and Mathematics Assessment Study Teacher characteristics 83 Teacher sex: male=0 female=1 84 Teacher ethnic: non-Kinh=1 Kinh=2 85 Teacher education: number of years education 86 Teacher TTR: number of years of teacher training 87 Teacher experience: number of years teaching primary school 88 Teacher excellent: named as excellent teacher this year 89 Teacher in-service effective: the extent to which the teachers deemed in-service training attended was effective 90 Teacher possessions: sum of possessions out of 21 91 Hours extra work for money Teaching resources 92 Books in classroom bookcase 93 Classroom furniture: sum of chalk board, chalk, wall chart, cupboard, bookshelf, bookcase, teacher chair, teacher table, fannd light 94 Classroom supplies: sum of dictionary, geometric instruments for use on board, teachers guide for Vietnamese, teachers guide for Math, reference material, and pictures as teaching illustrations 95 Classroom resources: the sum of classroom furniture and supplies Teacher's teaching 96 Teacher teach: number of hours per week teaching 97 Teacher work: hours teaching per week plus hours preparing lessons and marking homework 98 Meeting parents: frequency meeting parents: 0=never 1= once a year 2= once a term and 3=once or more a month 99 Percent of parents met 100 Teacher test reading: extent to which teachers give written test in reading 101 Teacher test math: extent to which teachers give written test in math 102 Advice from head: 0=never 1=once a year 2=once a term 3= once or more per month 103 Number of times observed by other teachers last month Teachers' perception of inspectors [1=no and 2=yes] 104 Number of inspections since 1998 105 Advises me 106 Criticises me 107 Suggests new idea 108 Clarifies curriculum objectives 109 Explains curriculum content 110 Recommends new teaching materials 111 Provides information for self-improvement 112 Contributes little to my classroom teaching (?? Maybe omit because of translation) 113 Makes suggestions on improving teaching methods 114 Encourages professional contacts with teachers in other schools 115 Provides in-service training to teachers 116 Finds faults and reports them to my employer Others and Achievement scores 117 School location 118 Pupil reading score 119 Pupil math score 120 Teacher reading score 121 Teacher math score 122 Pupil predicted reading score: each pupil's reading score is predicted from her/his home background 123 Pupil predicted math score: each pupil's math score is predicted from her/his home background 242 Vietnam Reading and Mathematics Assessment Study Chapter 9 INITIAL EXPLORATIONS OF RELATIONSHIPS AMONG VARIABLES Introduction T his study has been a cross-sectional sample survey of the state Care must be taken in of education in Grade 5 in April, 2001. The data were collected inferring causation from on one day. The study is therefore rather like having a static cross sectional data. picture taken of the situation of education on that day. It was not a longitudinal study or an experimental study with a control and treatment group. It is therefore not possible to prove cause and effect. The evidence in the study is has been based on various measures of association between variables and the outcome scores of mathematics and reading. Any inference about possible cause and effect must be a function of the memory, testimony and introspection of the person or persons, in this case the researchers, working on the study. It can happen that the inference may vary from one interpreter to another. This is ground for caution in making any inference, but it is not ground for making no inference. For example, the present writer believes from what he remembers (memory) of his personal experience as a teacher (testimony) and from thinking about it (introspection) that doing homework and having it corrected by the teacher and given back to the pupil is likely to help improve math achievement. The same can be said about several of the variables that have been reported in previous chapters. Similarly, it must be recalled that if there is no variance or very little Only variables that had variance in one of the variables there will be no correlation or a very variance could be small one. Take, for example, the case of the general education of examined. teachers. In a case where nearly all teachers have had 12 years of education there will be no correlation or one that is not far from zero. This does not mean that the general education of teachers is unimportant. It just means that because there was no variance there was no correlation. 243 Vietnam Reading and Mathematics Assessment Study Some simple and partial correlations In Chapter 2, the percentages of pupils reaching different skill levels in isolated, rural, and urban areas was shown in Table 2.8. If isolated is coded as 1, rural as 2, and urban as 3, the relationship between school location and the pupils' skill levels in reading can be calculated as a correlation. It is 0.21 showing that, in general, urban pupils score higher than rural pupils who score higher than pupils from isolated areas. It was important to In Chapter 3, it was 'getting math homework, doing it and having it corrected' examine both simple and (called 'math homework') that had a correlation of 0.15. This meant that the partial correlations, pupils who received and did homework and had it corrected achieved better with home background partialled out. in mathematics than those who had it given less. But, was this because those who got homework were in better schools because their parents were in a situation to have their children in better urban schools? When home background was partialled out from the correlation of 0.15, the partial correlation between 'get homework' and 'mathematics score' was 0.13. This means that irrespective of home background it was advantageous for a pupil to 'get homework'. Correlations of selected In Chapter 4 in Table 4.23, it was seen that the teachers' scores on the variables with mathematics test were highly correlated with the pupils' mathematics test achievement from scores (0.29). But, was this because the pupils came from better homes? The previous chapters were examined. partial correlation of 0.25 showed that some of the original correlation was due to 'better parents' but that with a partial correlation of 0.25 there was still a strong relationship between teacher math scores and pupil math scores. Irrespective of home background it was advantageous for a pupil to have a teacher with a high mathematics score. This relationship was also true at the provincial level. Those provinces where the teacher mean math scores were high also had a high pupil mean math score (see Figure 2.15). In Chapter 5, it was seen, in Table 5.18, that the correlation between school resources and pupils' reading scores was 0.26 and the partial correlation was 0.14. This meant, again, that some of the initial correlation was due to home background but that the correlation was still definite even when home background had been partialled out and therefore it was still advantageous to be in a school with a higher level of school resources. The relationship at the province level has been depicted in Figure 9.1 below. 244 Vietnam Reading and Mathematics Assessment Study Figure 9.1: Relationship between school resources and math achievement at the provincial level between schools It can be seen that for each province the line rises from left to right First, a correlation i.e. from low resourced schools in provinces to higher resourced matrix for a few variables was examined. schools in each province. Thus, it is slightly better for a pupil to be in a better resourced school if he or she wishes to achieve well in mathematics. But, is it the case that pupils who score well are from better homes, in urban areas, in schools with teachers with good scores and that are well resourced, and the 'get' homework? The correlation matrix at the pupil level of these five variables with pupil mathematics achievement has been given in Table 9.1 below. 245 Vietnam Reading and Mathematics Assessment Study Table 9.1: Correlation matrix of six selected variables Home Background School location 'Get homework' Teacher math score School resources School resources Home Background 1.0 .41 .08 .22 .43 .27 School location 1.0 .04 .15 .49 .19 'Get homework' 1.0 .07 .06 .15 Teacher math score 1.0 .26 .29 School resources 1.0 .25 Math achievement 1.0 All correlations were positive. It can be seen that pupils from 'better' homes tended to be in schools in urban areas (.41), and urban areas had schools that were better resourced than schools in rural and isolated areas (.49). Teachers with higher math scores tended to be associated with schools that had more resources (.26) and in schools with pupils from 'better' home backgrounds (.22). The correlation of 'get homework' with other variables were positive but low except for the relationship with pupil achievement. But, since all of the variables are inter-correlated is it possible to identify the relative strengths of association of each variable when account has been taken of the other variables? In Table 9.2, the results of a small regression analysis has been presented. Together the five variables accounted for 15.7 percent of the variation in pupil mathematics scores. Table 9.2: Regression analysis using five predictors(R˛= .157) Beta coefficients Home background .16 School location .05 Get homework .11 Teacher math score .22 School resources .09 The largest regression coefficient was .22 for Teacher math score and the second highest was .16 for home background. 'Get homework' was .11 and school resources and school location became low. The reason why school location became low was that it was highly correlated with home background and both were highly correlated with the criterion, namely pupil mathematics achievement. From this small analysis it can be said that if a pupil is to get good mathematics results it is advantageous if he or she comes from a good family, 246 Vietnam Reading and Mathematics Assessment Study attends a school where teachers have high mathematics scores and where the teachers give more homework. Since the MOET cannot change home background and school location, the A larger matrix of 26 two variables that can be changed are teacher math score and the giving and variables wax examined. correcting of homework. The former can be changed through more intensive It was clear that the teacher training and the latter through a mixture of pre- and in-service home backgrounds of training and the inspectorate. pupils were highly correlated with school But all in all there were 24 variables and two outcome variables - a total of and classroom 26. The inter-correlations have been presented, in Table 9.3, for those pupil resources, teacher variables that had simple correlations higher than 0.07 and for teacher and quality, the possession school variables that had correlations higher than 0.12. of textbooks. These variables were also A quick perusal of the table shows that Home Background has high highly correlated with correlations with pupil having a private corner for study at home (.31), pupil achievement. school and classroom resources (.49), have materials (.35), teacher possessions at home (.34), school location (.41), teaching staff teacher training (.33), percent of pupils in full-day shift, and percent female staff. Stated in other words this means that pupils from better home backgrounds were more likely to have a private corner at home, more likely to be in a school with good classroom and school resources, more likely to have more of their own textbooks, more likely to have teachers who are slightly richer, more likely to be in an urban area, more likely to be in a full-day school with more females on the teaching staff. In short children from richer families are much better off for teachers and schools than pupils from poorer families. But it can also be seen that pupil age and times having grade repeated were correlated .45. This is to be expected. But it could also be seen that pupils who were older were also in schools with fewer classroom and school resources (-.30). The variable 'teacher possessions in the home' was highly correlated with home background and school/classroom resources meaning that the urban areas had the richer teachers who were in schools with more resources. Schools with more resources were not only in urban areas (.41) and with richer teachers (.50), but also had a teaching staff that had had more teacher training (.44), had a higher percentage of pupils following the full-day school (.47), and had more parents giving contributions to the school (.40). In some cases it was seen that there were two variables that were very similar in nature that were highly correlated. For example, it could be seen that the years of teacher training of Grade 5 teachers and of the whole teaching staff was correlated .37. In this case it is probably better to drop one of them when looking for the relative effect of variables. 247 Vietnam Reading and Mathematics Assessment Study variables. three twenty for matrix Correlation 9.3 ableT 248 Vietnam Reading and Mathematics Assessment Study On the basis of the correlation matrix it was decided to form several constructs using principal components analysis. These constructs with their variables and variable loadings have been given in Appendix 9.1. A new correlation matrix was formed. This has been presented in Table 9.4. Table 9.4: Correlation matrix for final regression run Parental Probs Ethnic education School Home slow Get math Get Teacher TeacherTeachingPercent dealt Community home and quality circumstlearners homework reading reading math staff parents with by PRD500 PMA500construct support ances homework score score met head Ethnic home 1.00 0.32 0.33 -0.16 -0.28 0.04 0.05 0.11 0.18 0.21 -0.06 -0.03 0.15 0.12 0.41 Parental education and support 1.00 0.42 -0.27 -0.29 0.10 0.13 0.14 0.17 0.38 0.20 -0.09 0.34 0.33 0.41 School quality 1.00 -0.28 -0.30 0.05 0.09 0.22 0.28 0.57 0.25 -0.14 0.31 0.31 0.58 Home circumstances 1.00 0.19 -0.06 -0.07 -0.10 -0.12 -0.30 -0.11 0.08 -0.20 -0.20 -0.23 slow learners 1.00 -0.04 -0.04 -0.10 -0.14 -0.28 -0.08 0.09 -0.21 -0.19 -0.27 Get math homework 1.00 0.55 0.06 0.07 0.06 0.02 -0.01 0.15 0.15 0.05 Get reading homework 1.00 0.08 0.08 0.08 0.06 -0.04 0.16 0.17 0.04 Teacher reading score 1.00 0.57 0.21 0.08 -0.10 0.23 0.24 0.13 Teacher math score 1.00 0.24 0.09 -0.12 0.24 0.29 0.20 Teaching staff 1.00 0.28 -0.22 0.27 0.30 0.47 Percent parents met 1.00 -0.18 0.15 0.19 0.08 Probs dealt with by head 1.00 -0.11 -0.13 -0.05 PRD500 1.00 0.74 0.22 PMA500 1.00 0.19 Community construct 1.00 Several multiple regression analyses were conducted using pupil math score as the dependent variable. The results of the final analysis have been presented in Table 9.5. For reading the results have been presented in Table 9.6. Table 9.5: Regression run with math as dependent variable (R square=0.22) Unstandardized Standardized Coefficients Coefficients t Sig. B Std. Error Beta (Constant) 380.271 3.698 102.831 .000 A short list of variables slow learners -4.606 .541 -.046 -8.521 .000 Home circumstances -6.247 .530 -.062 -11.778 .000 was arrived at and a Ethnic home -3.775 .569 -.038 -6.637 .000 regression analysis was parent education and support 19.787 .587 .197 33.735 .000 Teaching staff 8.152 .651 .081 12.528 .000 undertaken. School quality 10.786 .685 .109 15.740 .000 Get math homework 10.039 .495 .100 20.296 .000 Teacher math score .195 .005 .195 37.719 .000 %parents met .363 .029 .066 12.338 .000 Problems dealt with -24.162 3.078 -.040 -7.849 .000 Community construct -5.042 .653 -.050 -7.725 .000 249 Vietnam Reading and Mathematics Assessment Study Table 9.6: Regression run with reading as dependent variable (R square=0.20) Unstandardized Standardized t Sig. Coefficients Coefficients B Std. Error Beta (Constant) 423.632 3.732 113.528 .000 slow learners -8.097 .549 -.080 -14.762 .000 Home circumstances -6.474 .538 -.064 -12.024 .000 Ethnic home -1.134 .576 -.011 -1.967 .049 Parent education and support 19.704 .597 .196 33.018 .000 Teaching staff 4.502 .661 .045 6.810 .000 school quality 11.468 .694 .116 16.516 .000 Get reading homework 9.540 .505 .095 18.878 .000 Teacher reading score .139 .005 .139 26.889 .000 Percent pupils met .161 .030 .029 5.390 .000 Problems dealt with -20.110 3.124 -.033 -6.438 .000 Community construct -.430 .663 -.004 -.649 .516 For mathematics it can be seen that the major predictors were 'parent education and support', 'teacher math score', 'school quality' and 'get math homework'. For reading the results were similar. The amount of variance accounted for by any one variable is the standardised regression coefficient * the simple correlation with achievement *100. The amounts of variance accounted for by these variables for mathematics was therefore: Variable Beta coefficient Correlation Variance explained Parental education/support .20 .33 7 Teacher math score .20 .29 6 School quality .11 .31 3 Get math homework .10 .15 2 The most important Even two percent of the variance is important so that it can be said that these variables to emerge variables are important. were 'teacher achievement', 'parent However, so far the data used have all been at the pupil level. That is, the education/care', 'school teacher and school data were disaggregated to the pupil level. This has quality', and 'get math provided an indication of relative importance of variables. The weakness homework'. of this approach has been this disaggregation. Cheung et al1 . showed how the major predictors could be identified no matter which form of It was then decided to analysis that was used but that a multilevel approach would provide undertake a multilevel more precise estimates. In the following section, both HLM (Raudenbusch analysis. and Bryk, HLM5/3L) and MLn (Rasbash et al.2) have been used to begin to explore the data. 1 Cheung K C., Keeves J P, Sellin N, and Tsoi S C. 1990. The analysis of multilevel data in educational research: Studies of problems and their solutions. International Journal of Educational Research. Vol. 14, No. 3 2 Rasbash et al., 2000 A user's guide to MLWin Version 2.1. University of London 250 Vietnam Reading and Mathematics Assessment Study Partitioning the Variance In this approach the total pupil variance of scores was partitioned into the amount due to a pupil being in a particular school or class. The first two pie charts (Figures 9.1 and 9.2) depict the situation for pupil reading scores for a three level analysis: school, class, and pupil. The first chart was calculated using HLM and the second MLn. Figures 9.3 and 9.4 are the same but for mathematics. Figures 9.5 and 9.6 are for a two level model using the school and the pupil levels for reading and finally Figures 9.7 and 9.8 are a two level model using mathematics. It will be recalled that intact classes were not tested in the design for this study. Half of the pupils were attached to a particular teacher because only two teachers per school were taken. Thus the number of pupils attached to a particular teacher ranged from five to nine. Although the first four charts have included this mini-class it must be recalled that the estimates for the class mean and the deviations from that mean may not be stable. Nevertheless it was agreed to try and make a crude estimate of the amount of variance that might be accounted for at the class level. Figure 9.1: Three level partitioning for reading Figure 9.2: Three level partitioning for reading (MLn) (HLM) Figure 9.3: Three level partitioning for math Figure 9.4: Three level partitioning for math (MLn) (HLM) 251 Vietnam Reading and Mathematics Assessment Study Figure 9.5: Two level partitioning of reading Figure 9.6: Two level partitioning of reading (MLn) (HLM) Figure 9.7: Two level partitioning of math Figure 9.8: Two level partitioning of math (MLn) (HLM) First, the variance was What does this partitioning of variance tell us? From Figures 9.1 and 9.2 it is examined to see how possible to state that the maximum amount of variance that could be much of it could be explained at the school level is just over 50 percent for reading. At the class attributed to the level it would be somewhere between 11 and 15 percent and at the pupil level differences between it would be about 34 percent. That is to say that from further analyses we may schools, between classes be able to explain part of the variance at each level but probably not all. within schools, and between pupils. By far the most variance was For mathematics, the maximum amounts are about 60 percent for the school between schools. level, between 11 and 16 at the class level, and 26 percent at the pupil level. Since the class level data in the data set are known to be somewhat unstable (because the classes were not really full classes), it was decided to restrict the HLM explanatory analyses to the school and pupil levels only. It was found (Figures 9.5 to 9.8) that the maximum amount of variance that could be explained was about 58 to 60 percent at the school level for reading and about 64 percent for mathematics. Thus, at the pupil level it was about 40 percent for reading and 35 percent for mathematics. 252 Vietnam Reading and Mathematics Assessment Study It is useful to have this information in terms of the maximum amount of variance to be explained at each level because this is what we are aiming for. It also tells us about the system of education. To re-state what has been said in earlier chapters the amount of variance between schools is very large compared with many countries. The partitioning of variance has confirmed what had already been seen in previous chapters. The variation among schools was very large; among classes within schools it was not very large and among pupils it was average. The major questions then become which were the variables in schools and classes (including teachers) that were having a large effect on pupils' achievement. Some exploratory analyses A series of small exploratory analyses were run using both MLWin and HLM. The main results of the The tables and figures for the first analyses have been presented in Appendix exploratory analyses 9.2 and a major HLM analysis has been presented in Table 9.3. Given the confirmed what have complexity of the analyses, no detailed explanation has been given. It is left been found in preceding to those understanding such analyses to read the appendices. chapters. There were extraordinarily high The main results of the analyses were: differences between schools, a major inequity in the system. there were large between school differences for both subjects The major factors the importance of home background, especially parental education and associated with the support, but also travel time, days absent, meals per day, and grade achievement between repetition in influencing pupil achievement. Being non-kinh and not pupils and between speaking Vietnamese outside of school was a disadvantage. schools were home the importance of being given homework and having it corrected background, teacher the greater 'effect' of school location on pupil reading score differences achievement, teaching than on pupil math score differences staff qualifications, the large 'effect' of teacher subject-matter knowledge on pupil, class and school quality, and school achievement homework being given the negative effect of school behavioural problems on pupil achievement and corrected. the positive effect on achievement of being in a school with more female teachers and having a class female teacher the positive effect of being in a school where the teachers had more teacher training, where the head had attended a management course, and where the class teacher had been designated as an excellent teacher the positive effect of being in a classroom and school with more classroom and school resources 253 Vietnam Reading and Mathematics Assessment Study When the relative effect of the variables was examined, the same results were obtained whichever form of analysis was employed. The largest effect was the teacher achievement score (the teacher subject-matter knowledge) followed by parental education and support, and then homework being given and corrected. The largest negative effects were for pupil behavioural problems and grade repetition. These effects were also cumulative. The best schools were in urban areas, had well-trained, female teachers with high achievement scores who gave much homework and corrected it and who were in classrooms and schools with many resources. The worst schools were the opposite. Until the schools are equalized on many of the variables, the large inequities in achievement will remain. But the inequities must be remedied by bring the lower levels up to the higher levels and not vice versa. The greatest disappointment from these multivariate analyses is that only 25 percent of total variance was accounted for. This means that there is a lot of variance, especially between schools, that has not been explained by the variables that were collected in this study. Before the repeat of this study, much conceptualization will be required about which other variables are 'responsible' for the differences between schools and how they might be measured. On the other hand, it should be stressed that the results of the multivariate analyses produced the same results - from a policy perspective - that were produced from the univariates analyses. This is reassuring. Conclusion The aim of this chapter was to establish the relative importance of variables as input to the MOET. Those of the variables that are considered to be malleable will be of great importance to the MOET since it provides evidence on where the most impact will result. From the previous chapters it was evident that there were pockets of schools where the achievement was high and other pockets where the achievement was low. The pupils in the schools in the isolated areas had low achievement and in their schools there was a paucity of equipment and supplies as well as of well trained teachers who gave and corrected homework. The pupils in urban areas such as Hanoi/Halong, Ho Chi Minh City and Danang and Hue achieved well and they were in schools where there was more equipment and supplies, and better qualified teachers who gave and corrected home work more frequently. 254 Vietnam Reading and Mathematics Assessment Study The analyses conducted in this chapter confirmed the large between school inequities for so many variables. However, it was variables such as the teachers' subject matter knowledge, their giving and correcting homework and the quality of the whole of the teaching staff within a school that were associated with large differences in achievement between pupils and schools. Equipment and supplies were also important. These are all variables about which something can be done by the relevant authorities for education. Many of the analyses confirmed the positive affect of many of the Ministry's initiatives taken during the last decade of the 1900s. The effect of excellent teachers, of the teacher training, of the school head management courses bear witness to this although it was seen that the management courses should include more training for the heads on how to deal with children having parents who had low educational levels and did not give much support to their children. But some lacunae remain. The achievement in the isolated schools is a matter for concern and it is in these areas (as well as in some urban pockets) that there is a lack of parental support for their children's education. This was also presumably coupled with the non-enrolment of children in school or early drop-out in some cases. This provides a strong case for the school intervention programs mentioned earlier in this report. To change parental attitudes and behaviours in the home is one of the most difficult things to do but, given its success elsewhere, it should be attempted. The main messages for policy suggestions were that teacher subject matter knowledge must be attended to urgently and in ways already suggested earlier in this report. Secondly, school intervention programs are likely to be useful in changing parental behaviours and attitudes at home and that these in turn will have a positive effect on pupil achievement. Thirdly, the giving and correcting of homework must be seen as a priority for all teachers and not just those in good schools. The ways in which this can be done can be written into short booklets but also included in teacher training programs (both pre- and in-service) as well as in the management courses for head teachers. Inspectors can also serve as reinforcing agents. In those schools experiencing pupil behavioural problems, special research studies should be carried out in order to discover why these problems are occurring and then devising ways of dealing with the problem. Finally, every effort should be made to eliminate grade repeating. This is much easier said than done because it requires a determined effort buy all concerned to provide a high quality education for all pupils, and as already seen, this is a mixture of many of the variables that have been examined in this report. 255 Vietnam Reading and Mathematics Assessment Study Appendix 9.1: Constructs used in Chapter 9. It will be seen that for each construct there is a title for the construct and then a list of variables. Next to each variable there is a coefficient in the column headed 'Component'. The coefficient is the weight that the variable has in the construct. Thus, in the first construct speaking Vietnamese has a weight of +.952 and Belonging to an ethnic group has a weight if -.952. But the variable 'belonging to an ethnic group' is coded as 1 for an ethnic minority and 2 for the majority group. Thus the two variables go together in equal weights to form the construct. In the last construct mentioned, that of 'community in which the school is located' it can be seen that being in an urban community was weighted .744, having electricity .655, having piped water .582 and, in a sense, the opposite of the above 'distance from public amenities', a rural variable with a weight of -.752. The first three variables were really urban variables. 1. Ethnic background (xchmethn) Component Speak Vietnamese outside school .952 Belonging to ethnic group -.952 2. Parental education and support (xcedcare) Component Books at home .595 Parent education .653 Have textbooks .561 Have private study corner .596 Possessions at home .822 3. Home circumstances (xchomcir) Component No. of meals per day -.579 Travel time to school .653 Days absent last month .604 4. Slow learner (xcsloln) Component Age in months .851 Grade repeated .851 256 Vietnam Reading and Mathematics Assessment Study 5. Teaching staff of school (xcstaff) Component Amount of staff teacher training .854 % female teaching staff .854 6. School quality (xcschres) Component Total school resources .806 Total classroom resources .761 % pupils in full-day schooling .695 Amount of parental contributions .629 The other variables to be used were the same as before and were: 7. Get math homework and have it corrected (xpmathhw) Component Get math homework 0.804 Corrected math homework 0.804 8. Get reading homework and have it corrected (xpreadhw) Component Get reading homework 0.842 Corrected reading homework 0.842 9. tma500 (Teacher math score) 10. trd500 (teacher reading score) 11. tpctmeet (percent of parents met) 12. xspbhpro (sum of problems dealt with by head) 13. pma500 (Pupil math score) 14. prd500 (Pupil reading score) 15. xcomm (community in which school is located: school location, electricity and piped water, and distance from public amenities) 15. Community in which school was located (xcomm) Component School location 0.744 Piped water 0.582 Electricity 0.655 Distance from public amenities -0.752 In some analyses in Chapter 9, the construct 'home background' has been used. The loadings for this construct have been given at the end of Chapter 3. 257 Vietnam Reading and Mathematics Assessment Study Appendix 9.2: The presentation of selected multilevel analyses Model 1 This analysis attempted to estimate the variance in pupil math scores. The results have been presented in Table 9.2.1. Starting with the column HB (for Home background) each further column can be regarded as a model. The estimation of variance explained by added covariates was based on the assumption that previously added covariates were still explaining the same amount of variance as in previous models. Therefore, it is possible that the estimates of some added explanatory variables were slightly underestimated. Table 9.2.1: Explanation of variance in mathematics score 3 level model (MLn) Levels Covariates HB HW TT TM TM, PR TM, PR, HB TM, PR, HW TM, PR, TT school 55 59 53 50 20 20 20 19 teacher 14 15 15 17 9 7 8 8 pupil 26 27 27 26 18 18 18 18 TM 7 7 7 7 7 PR 46 46 46 46 HB 5 2 HW 0 0 TT 6 2 total 100 101 101 100 100 100 99 100 HB=home background HW =math homework TT=staff teacher training TM=Teacher math score PR=Pupil reading It can be seen that HB accounted for 5 percent of the variance. Homework did not emerge as an important variable. When entered a single variables, teacher training accounted for 6 percent of the variance, and Teacher math score accounted for 7 percent of the variance. When the pupil reading score was entered as a predictor of pupil math score (even though it was measured concurrently with pupil math) then it accounted for 46 percent of the variance. The teacher math score remained constant at 7 percent. When teacher training was added to the equation with the pupil reading score and teacher math score then the variance accounted for by teacher training fell to 2 percent. 258 Vietnam Reading and Mathematics Assessment Study Model 2 A similar model was run for reading and the results were similar (see Table 9.2.2) Table 9.2.2: Explanation of variance in pupil math score 3 level model (MLn) Levels Covariates HB HW TT TR TR, HB TR, HW TR, TT school 46 51 47 47 42 47 43 teacher 12 13 13 15 12 13 13 pupil 34 34 35 34 34 34 35 teacher read (tr) 4 4 4 4 hb 8 8 hw 2 2 tt 5 5 total variance explained 100 100 100 100 100 100 100 The amount of variance accounted for was for the column where the teacher reading score and home background were entered. These were four and eight percent respectively. Why math homework was so important in the regression analysis and not in this MLn analysis is not known. Two further analyses were run to estimate the amount of variance accounted for by school location. These showed that the amount of variance accounted for by school location for pupil math was only of the order of 4 or 5 percent but for reading it remained high at about 16 percent. Why school location should affect pupil reading achievement so much more than pupil math achievement is not known. 259 Vietnam Reading and Mathematics Assessment Study Appendix 9.3: An HLM analysis of the Vietnam Grade 5 survey data2 The multilevel analyses reported in this appendix have been carried out in order to examine factors influencing achievement in mathematics and reading among Grade 5 primary school pupils in Vietnam. The computer package used for the multilevel analyses in this study is HLM5 for Windows developed by Raudenbush, Bryk and Congdon (2000). The HLM program was initially developed by Bryk et al. (1986) to find a solution for the methodological weakness of educational research studies during the early 1980s, which was the failure of many analytical studies to attend to the hierarchical, multilevel character of much of educational field research data (Bryk and Raudenbush, 1992). This failure came from the fact that "the traditional linear models used by most researchers require the assumption that subjects respond independently to educational programs" (Raudenbush and Bryk; 1994, p. 2590). In practice, most educational research studies select pupils as a sample who are nested within classrooms, and the classrooms are in turn nested within schools, schools within geographical location, state, or country. In this situation, the pupils selected in the study are not independent, but rather nested within organizational units and ignoring this fact results in the problems of "aggregation bias and misestimated precision" (Raudenbush and Bryk; 1994, p. 2590). The multilevel technique employed in this study has been used in other studies to tease out factors influencing student achievement in several developing countries. For example, Willms and Somers (2001) used hierarchical models to examine socioeconomic factors and other factors influencing mathematics and language achievement among Grades 3 and 4 pupils from 13 Latin American countries. In addition, a recent study by Howie (2002) successively employed this multilevel technique to tease out language and other factors influencing student achievement in mathematics in South Africa using data that were collected as part of the TIMSS-R study. The structure of this appendix is as follows. A short section has been provided in which the criteria employed to select the variables included in the multilevel analyses reported in this appendix have been presented. In the second section, the hypothesised multilevel models have been reported. The multilevel analyses have then been described and, finally, sections containing the results of the analyses have been presented and interpreted. 2This appendix was written by Njora Hungi of Flinders University, South Australia 260 Vietnam Reading and Mathematics Assessment Study Selection of variables It should be noted that some of the composite variables mentioned earlier and in the main body of the text have not been used in the multilevel analyses reported here. Generally, a variable was dropped if it was considered to provide information that could be obtained from the other variables included in these multilevel analyses. Variables were also dropped if it was considered that it would be less informative to use them as they were or it was considered that it would be easier to interpret the results if the individual variables constituting the variables were used singly in these analyses. For example, the information provided by the 'home background' construct (XHOMEBA3 - speaking Vietnamese, ethnic background, books at home, parental education, and wealth) could also be obtained from the two variables 'parental education and support' (XCEDCARE - books at home, private corner, parental education, and wealth) and 'ethnic background' (XCHMETHN - speaking Vietnamese, and ethnic background). In this example, it was considered that it would be more informative and easier to interpret the results if the two variables (XCEDCARE and XCHMETHN) were used separately rather than as a composite in XHOMEBA3. A complete list of the variables selected for examination in the multilevel analyses reported in this chapter has been presented in the next section together with descriptions of the models that were tested in this study. Hypothesised models The data for this study were collected at two different levels, namely: pupil, and school. Two Grade 5 teachers were selected at random from each selected school and then linked to their own class pupils. Since 20 pupils had been drawn at random within each selected school, this linkage of teachers to pupils resulted in some 5 to 9 pupils being linked to a class teacher. In this sense, a 'group of pupils' was not necessarily representative of the real class from which the pupils came. Despite the obvious limitations of these 'class groups', a decision was taken to conduct the HLM analyses using both two and three level models. For the two-level models (Figure A9.1), the hierarchical structure is such that pupils within school, that is, the Level-1 units are pupils and Level-2 units are schools while for the three-level model (Figure A9.2) the Levels 1, 2 and 3 units are pupils, class groups (to be called class) and schools respectively. For the purposes of simplicity, these diagrams (Figures A9.1 and A9.2) have not included illustrations of any hypothesised cross-level interaction effects. However, it should be noted that in the actual analyses several cross-level interaction effects were examined. It should further be noted that for each of the general types of model, two separate models were specified, one for 261 Vietnam Reading and Mathematics Assessment Study mathematics and the other for reading. The outcome variables of interest in these models are pupils' scores in mathematics (PMA500) and reading (PRD500) tests, both with a mean of 500 and a standard deviation of 100. Thus, for the multilevel analyses reported in this appendix, there were eight, 15 and 19 variables that were hypothesised to influence directly pupil achievement in mathematics (or in reading) at the pupil, class and school levels respectively (see Figures A9.1 and A9.2). It should be noted that the variables hypothesised to influence pupil achievement in mathematics and reading at the pupil and school levels in the two-level models were the same variables hypothesised to influence pupil achievement at the corresponding levels in the three-level models. It should further be noted that some of the variables examined for inclusion in the two-level models at the school-level and for inclusion in the three-level models at the class and school levels were constructed by aggregating the pupil-level data. For example, pupil-level data on the variable Days Absent were aggregated at the class-level in order to construct the variable Average Days Absent at the class-level while pupil- level data on this variable were aggregated to the school-level in order to construct the variable Average Days Absent at the school-level. 262 Vietnam Reading and Mathematics Assessment Study reading and mathematics for model hierarchical two-level Hypothesised A9.1: Figure 263 Vietnam Reading and Mathematics Assessment Study reading and mathematics for model hierarchical three-level Hypothesised A9.2: Figure 264 Vietnam Reading and Mathematics Assessment Study The names and codes of the predictor variables tested for inclusion at each level of the two-level (Figure A9.1) and the three-level (Figure A9.2) models have been provided in Table A9.1. The predictor variables listed together in Table A9.1 were considered to be alternative versions of the same underlying measure and therefore, they were not added into the model together to avoid problems associated with suppressor variables (Keeves, 1997). For example, at the pupil-level, the variable 'Private Corner' (PPRIVCOR) and the variable 'Parental Education & Care' (XCEDCARE) were considered as alternative versions of the same measure (Parental Education & Support) and consequently have been listed together in Table A9.1. For the same reason, the variables constructed by aggregated pupil-level data on these two variables at the class-level (PPRIVC_1 and XCEDCA_1) and at the school- level (PPRIVC_2 and XCEDCA_2) have been listed together in Table 9.1. Likewise, the variables XLOCA_2 and XCOMM_2 have been listed together indicating that they were considered to be alternative versions of the same measure (School Location). The correlation matrices of the variables at pupil, class and school levels can be found in Tables AA9.1, AA9.2, and AA9.3 respectively at the end of this appendix. Construction of the SSM files The original SPSS data files involved in this part of the study had 36,476 pupils in 7,221 classes nested in 3,635 schools. However, in the construction of the sufficient statistic matrix (SSM) files there were a few schools that were dropped because of insufficient data. Indeed, for the three-level analyses, the Ns in the SSM file were 35,420 for pupils, 7,090 for classes and 3,622 for schools while for the two-level analyses the Ns were 36,383 for pupils and 3,623 for schools. Weighting (with PWEIGHT3) was undertaken in the construction of these SSM files. The descriptive statistics of the variables included in these SSM files can be found in Tables AA9.5 and A9.6 at the end of this appendix. Analyses For both the two- and the three-level analyses, the first step undertaken was to run a null model in order to obtain the amounts of variance available to be explained at each level of the hierarchy (Bryk and Raudenbush, 1992). The null model is the simplest model because it contains only the dependent variable (for this study, pupil mathematics or reading score) and no predictor variables are specified at any level. The second step undertaken was to build up the pupil-level model or the 'unconditional' model at Level-1. This involved adding pupil-level predictors to the model, but without entering predictors at any of the other levels of the hierarchy. The purpose of this step was to examine which pupil-level variables had significant (.05 level) effects on the outcome variable. A 'step- 265 Vietnam Reading and Mathematics Assessment Study up' approach was followed to examine which of the pupil-level variables had a significant influence on achievement in mathematics and reading in each of the hypothesised models. Bryk and Raudenbush (1992) have recommended the step-up approach for inclusion of variables into the model to the alternative approach 'working-backward' where all the possible predictors are dumped into the model and then the non-significant variables are progressively eliminated from the model. The third step, which for the two-level analyses was the final step, involved adding the Level-2 predictors into the model using the step-up strategy mentioned above. In this step, the predictor variables examined for inclusion in the three-level models were the class-level variables while the predictors examined for inclusion in the two-level models were the school-level variables (see Table A9.1). The Level-2 exploratory analysis sub-routine available in HLM5 was employed for examining the potentially significant Level-2 predictors (as shown in the output) in successive HLM runs. Table A9.1: Variables tested on each level of the hierarchy Level Variable of Interest Variab le(s) Tested in HLMa Pupil Age in Month XPAGEMON Meals per Day PMEAL Travel Time PTRAVEL Days Absent PABSENT Parental Education and Support PPRIVCOR, XCEDCARE b Grade Repetition XPGREP Homework XPMATHHW, XPREADHW c Ethnic Background XCHMETHN Class Average Pupil Age XPAGEM_1 Average Meals PMEAL_1 Average Travel Time PTRAVEL_1 Average Days Absent PABSEN_1 Average Parent Support and SES PPRIVC_1, XCEDCA_1 c Average Grade Repetition XPGREP_1 Average Homework XPMATH_1, XPREAD_1 b Average Ethnic Background XCHMET_1 Classroom Resources/Materials XTRESC_1, XHAVMA_1 b Teacher Score TRD500_1, TMA500_1 c Teacher Gender TSEX_1_1 Teacher Training XTTTR_1 Excellent Teacher XTEX_1_1 Teacher Possessions XTHPOS_1 Percent Parents Met TPCTME_1 Average Pupil Age XPAGEM_2 Average Meals PMEAL_2 Average Travel Time PTRAVEL_2 Average Days Absent PABSEN_2 Average Parent Support and SES XCEDCA_2, PPRIVC_2 b Average Grade Repeat XPGREP_2 Average Homework XPMATH_2, XPREAD_2 c 266 Vietnam Reading and Mathematics Assessment Study Average Ethnic Background XCHMET_2 Pupil Behavior XSPBHP_2 Teacher Behavior XSTBHP_2 School Resources XSREST_2 Average Teacher Score TMA500_2, TRD500_2 c Average Teacher Training XSTT_1_2 Percent Full Day XSSHFT_2 Management Course SMANAG_2 School Location XLOCA_2, XCOMM__2 b Percent Female Teachers XSFEMT_2 Parental Contribution XSPARC_2 Reading Programs XSPROG_2 d Notes: a. Meanings of these codes can be found in Appendix 9.4. b. These variables are listed together because they are considered as alternative versions of the same measure and therefore were not included in the model simultaneously. c. Variable listed first is for testing in the mathematics models while the second variable is for testing in the reading models. d. This a Reading Program variable and was therefore tested for inclusion in the reading models only. The final step for the three-level analyses involved building up the model to the school-level through adding the significant school-level predictor variables into the model using the Level-3 exploratory analysis sub-routine and the step-up strategy. In all of the analyses described above, the general 'centering' approach followed was to enter the variables as grand-mean-centred (see Kreft, 1995 and Kreft et al., 1995). However, the dummy variables (Teacher Gender at the class-level and Management Course at the school-level) were entered uncentred in order to facilitate the interpretation of interaction effects. In addition, the estimated coefficients of a variable was 'fixed' at a particular level of the hierarchy if the reliability estimate of the predictor dropped below 0.05 value at that level (see Bryk and Raudenbush, 1992, p.110; Pituch, 1999, p.201; Raudenbush and Bryk, 2002, p.125) or if the estimation procedure would not converge when the variable was added to the model (Hox, 1995). Specifying a variable as 'fixed' constrains its slope or intercept to be the same across all the Level-2 and/or Level-3 units. The alternative is to specify a variable as 'random', which allows the slopes and intercepts to vary among the Level-2 and/or Level-3 units (Raudenbush et al., 2000). Results and discussions In the sub-sections that follow, the results of the analyses described above are presented and discussed. However, for purposes of simplicity, only the results of variance partitioning obtained from the null model, and fixed effects and variance explained results obtained from the final models have been reported 267 Vietnam Reading and Mathematics Assessment Study here. For the fixed effects, a discussion of the final models at the pupil-level has been presented followed by discussions of the models at the class (for the three-level models only) and school levels, respectively. A discussion of cross-level interaction effects in the final models has been provided in a separate sub-section. Variance partitioning Tables A9.2 and A9.3 display estimates of variances involved in the two-level and three-level models for the two outcome variables respectively. A discussion of the procedure employed to calculate the percentages of variance available (given in Tables A9.2 and A9.3) is to be found in Raudenbush and Bryk (2002; p.68-95). The first pie chart included in each of these two tables summarises the situation for pupil mathematics scores while the second pie chart summarises the situation for pupil reading scores. Needless to say, the pie charts included in Table A9.2 are for the two-level analyses while those included in Table A9.3 are for the three-level analyses. Hence, the results in Tables A9.2 and A9.3 show that based on a two-level analysis, 32.2 percent and 67.8 percent of the variation of pupil mathematics scores were at the pupil and school levels respectively, and based on a three- analysis the percentages were 25.1, 11.3 and 63.6 for pupil, class and school levels respectively. The corresponding percentages for pupil reading score based on a two-level analysis were 40.1 and 59.9 for pupil and school levels respectively, and the corresponding percentages based on a three-level analysis were 33.0, 11.2 and 55.8 for pupil, class and school levels respectively. These percentages of variation of pupil scores at the various levels of the hierarchy were the maximum amounts of variance available at those levels that could be explained in subsequent analyses. Table A9.2: Variance partitioning based on the two-level models Term Variance (%) Variance Component Available Chat Mathematics Pupil 0 2 3233.94 (32.2) School 0 6795.77 (67.8) Total 10029.71 Reading Pupil 0 2 3999.46 (40.1) School 0 5966.23 (59.9) Total 9965.68 268 Vietnam Reading and Mathematics Assessment Study Table A9.3: Variance partitioning based on the three-level models Term Variance (%) Variance Component Available Chat Mathematics Pupil 0 2 2483.17 (25.1) Class 0 1112.43 (11.3) School 0 6290.52 (63.6) Total 9886.13 Reading Pupil 0 2 3216.29 (33.0) Class 0 1091.07 (11.2) School 0 5442.07 (55.8) Total 9749.43 Generally, for both outcome measures, the variation of pupil scores at the school-level based on the two-level analysis followed closely the variation at the school-level based on the three-level analysis. Thus, a three-level analysis of these data seemed to disentangle the amounts of variances available at the class-level from the amounts of variances available at the pupil-level. Arguably, this separation of the class-level variance from the pupil-level variance provides a better image of the situation at these two levels (that is, class and pupil) compared to the image provided by the two-level analyses. However, as far as the amount of variance available to be explained at the school-level is concerned, it did not seem to matter markedly whether a two- level or three-level analysis was employed. In summary, the results from the two-level analyses displayed in Table A9.2 indicated that in Vietnam, the variation between pupils within schools in terms of their achievement in mathematics and reading at Grade 5 was roughly around a half of what the variation in performance between schools was. However, the results from the three-level analyses further revealed that about a half (for mathematics) or a third (for reading) of this variation (as seen in the two-level analyses) between pupils within schools could be attributed to the differences between the classes that the pupils belong to. Importantly, regardless of the type of analysis that was employed, these results indicated that the variance between schools in Vietnam is bigger compared to what is generally reported at similar grade levels in other developing countries. For example, Willms and Somers (2001), utilising data from Grade 3 and 5 pupils from 13 Latin American countries, found that the variance between schools in mathematics achievement range from 19.5 to 269 Vietnam Reading and Mathematics Assessment Study mathematics for model hierarchical two-level Final A9.3: Figure 270 Vietnam Reading and Mathematics Assessment Study reading for model hierarchical two-level Final A9.4: Figure 271 Vietnam Reading and Mathematics Assessment Study mathematics for model hierarchical three-level Final A9.5: Figure 272 Vietnam Reading and Mathematics Assessment Study reading for model hierarchical three-level Final A9.6: Figure 273 Vietnam Reading and Mathematics Assessment Study r s 3.32 0.04 0.04 0.32 0.88 0.56cs 1.91 0.68 2.39 2.42 1.32 0.57 2.92 0.01 1.77 0.53 0.10 0.97 2.80 1.48 1.89 4.16 6.39 3.87 7.53 0.54 S.Erro Three-level Coeff. 482.97 -0.37 -0.10 -2.19 1.03-1 5.58 -4.70 3.50 6.72 2.95 2.19 0.10 4.09 1.44 9.03 1.51 -10.00 12.51 -0.24 -2.02 -7.25 13.10 15.32 -31.24 10.62 18.59 Reading s models 1.18 0.03 0.03 0.23 0.71 0.44 0.67 2.27 2.24 1.24 0.51 1.87 4.15 6.40 1.80 0.01 3.82 7.23 0.54 1.87 S.Error wo-levelT three-level Coeff. 490.87 -0.31 -0.13 -1.72 -9.44 3.36 4.01 6.04 -9.37 2.78 2.46 13.58 13.96 -31.44 -4.36 0.15 1.231 21.83 1.54 15.08 and c s two- 3.27 0.03 1.04 0.03 0.28 0.75 0.49 0.59 2.23 2.50 1.22 0.49 2.67 0.01 1.73 1.69 0.55 0.24 0.88 1.37 2.09 4.53 7.07 4.29 7.68 0.58 S.Error the from Three-level els. end Coeff. 483.81 -0.14 2.55 -0.12 -2.47 -8.73 6.58 1.85 6.14 -7.36 2.75 1.65 15.45 0.13 3.79 5.16 1.43 0.51 -2.51 10.57 14.65 15.62 -36.75 15.16 20.10 1.39 mod the at fix reading suf reading s s Mathematics the _2 1.22 0.03 0.80 0.03 0.22 0.61 0.41 1.16 1.58 0.53 2.18 2.24 1.18 0.47 2.10 4.43 7.03 1.88 0.01 4.43 4.28 7.60 0.57 1.97 in and have S.Error wo-levelT includedis variables 1 Coeff. 490.56 -0.13 2.09 -0.1 -1.96 -7.18 6.01 3.28 -4.01 2.30 6.23 -7.59 2.59 2.33 14.29 13.15 -33.54 -3.82 0.21 15.83 1.501 22.68 1.23 10.98 variable mathematics school-level second of for the centred a codes while school-level class-level while Included the the grand-mean models at at end estimates are vary vary the s ariableV to to at fix of VEL THHW/XPREADHW VE_1 TH_2/XPREAD_2a displayed suf effect AGEMON ARC_2 variables leftis leftis are Code XP PMEAL PTRA ABSENTP XPGREP XPMA PMEAL_2 XPGREP_2 XCHMETHN XCEDCARE XPGREP_2 SMANAG_2b XCOMM__2 XSFEMT_2 TMA500_1/TRD500_1a TSEX_1_1b XTEX_1_1 XTRESC_1 XTHPOS_1 PTRA ABSEN_1P XPGREP_1 XCEDCA_1 XPMA PMEAL_2 XSPBHP_2 ABSEN_2P mathematics _1 TMA500_2/TRD500_2a XSTT_1_2 XSSHFT_2 XSFEMT_2 XSP XCEDCA_2 the other .05 in all ficient ficient p have fixed coef coef at Final includedis this this centred, of of ficients variables Repetition Repetition eacherT not coef Course Support& Support& first A9.4: parameter parameter class-level Meals Grade Support Grade Location Female listed variable of significant and ableT verageA verageA verageA Management School Percent imeT Education Education Absent Repetition Score eachersT Absent ariableV- Dummy- Residual- Residual- Only- Codes- Day Name Day with with with with with with Resources/Materials Month per imeT Education Score Gender eacherT Possessions ravelT Days Grade Parental Homework Meals Days eacherT rainingT Full Female Contribution Parental in Absent Repetition Background Behavior ariableV Intercept Age Meals ravelT Days Grade Homework interaction interaction Ethnic Parental interaction interaction interaction interaction eacherT eacherT Excellent Classroom eacherT verageA verageA verageA verageA verageA verageA Pupil verageA verageA eacherT Percent Percent Parental verageA a b s c Level Pupil Class School Notes: 274 Vietnam Reading and Mathematics Assessment Study 41.2 per cent. Thus, the results presented in Table A9.2 appear to warrant a careful look at the Vietnam schooling system to establish exactly what is happening. Final model estimates The final two-level hierarchical models for mathematics and reading have been presented in Figures A9.3 and A9.4 respectively while the same, but for the final three-level models, have been presented in Figures A9.5 and A9.6. Only the factors that had a significant (.05) direct (or interaction) effect on pupil achievement have been displayed in Figures A9.3 to A9.6. An 'effect' was considered to be significant at p=.05 level if its coefficient taken in absolute terms was more than twice its standard error (given in parenthesis in these diagrams). Fixed effects estimates from the final two-level and three-level models HLM analyses for mathematics and reading have been presented together in Table A9.3. That is, Table A9.3 displays a summary of the information presented in Figures A9.3 to A9.6. These results at the various levels of hierarchy are discussed next. Pupil-level model From Table A9.4, it can be observed that the results of the two-level analyses strongly agreed with the results of the three-level analyses regarding which pupil-level variables have significant influences on achievement in mathematics and reading. For both the two- and the three-level models, the results in Table A9.4 show that all the eight pupil-level variables examined in this study had a significant influence on achievement in mathematics. These eight pupil-level variables were: Age in Months, Meals per Day, Travel Time, Days Absent, Grade Repetition, Homework, Ethnic Background, and Parental Education & Support. All but one (Meals per Day) of these eight pupil-level variables also had significant influences on achievement in reading in both the two- and three-level models. It is worth noting that in the unconditional models at Level-1, Meals per Day also had a significant influence on achievement in reading but this variable was eventually deleted because it lost its significance when variables at the subsequent levels of the hierarchical structure were added. By way of caution, sizes of metric coefficients of the variables provided in Table A9.4 do not indicate the relative magnitude of effects and can not therefore be used to rank the variables in terms of their relative degree of influence on the outcome. However, the signs of the coefficients indicate the directions of effects and can be interpreted meaningfully if the coding of the variables and their standard deviations are considered Accordingly, for both the two- and the three-level models, the negative coefficients for the variables Age in Months, Travel Time, Days Absent and Grade Repetition indicated the following effects on achievement in 275 Vietnam Reading and Mathematics Assessment Study mathematics and reading among Vietnamese Grade 5 pupils when other things were equal. 1. Age in Months: Pupils of younger age were likely to achieve better than their older counterparts. 2. Travel Time: Pupils who took a shorter time to travel to school (or live near the school) were likely to achieve better than those pupils who took a longer time to travel to school (or live far from school). 3. Days Absent: Pupils who were never (or rarely) absent from school were likely to achieve better than those pupils who were frequently absent from school. 4. Grade Repetition: Pupils who had never repeated a grade were likely to achieve better than pupils who had repeated a grade one or more times. On the other hand, the positive coefficients for the variables Meals per Day, Homework, Ethnic Background and Parental Education & Support indicated the following relationships between these factors and achievement of Vietnamese Grade 5 pupils in mathematics and reading when other things were equal. 5. Meals per Day: Pupils who ate more meals per day were likely to achieve better in mathematics (but not necessarily in reading) than pupils who ate fewer meals per day. 6. Homework: Pupils who were given homework more frequently and had it corrected were likely to achieve better than pupils who were given homework and had it corrected less frequently. 7. Ethnic Background: Pupils who always spoke Vietnamese outside school or belonged to the Vietnamese ethnic majority group (Kinh) were likely to achieve better than pupils who never speak Vietnamese outside school or belong to the Vietnamese ethnic minorities groups. 8. Parental Education & Support: Pupils from homes with more educated parents, more books and other possessions, and who had a private study corner were likely to achieve better than pupils from homes with less educated parents, fewer books and other possessions and no a private study corner. From Table A9.4 it can also be seen that there were interaction effects involving pupil-level variables (Homework and Parental Education & Support) and some school-level variables. These interaction effects have been discussed later in a separate sub-section. Class-level model For the three-level models, the results in Table A9.4 showed that out of the 21 class-level variables examined in this study (list in Table A9.1), seven had a significant influence on achievement in mathematics. These seven class- level variables were: Teacher Math Score, Teacher Gender, Excellent 276 Vietnam Reading and Mathematics Assessment Study Teacher, Classroom Resources/Materials, Teacher Possessions, Homework, Average Days Absent and Average Parental Education & Support. Five out of these seven class-level variables that had a significant influence on achievement in mathematics also had a significant influence on achievement in reading. They were Teacher Reading Score, Teacher Gender, Classroom Resources/Materials, Average Days Absent, and Average Parental Education & Support. A further two class-level variables (Average Travel Time and Average Grade Repetition) had a significant influence on achievement in reading but not in mathematics. Among the five class-level variables that had a significant influence on achievement in mathematics as well as in reading, four (Teacher Score, Teacher Gender, Classroom Resources/Materials and Average Parental Education & Support) had positive coefficients and one (Average Days Absent) had a negative coefficient. Thus, other things being equal, the directions of effects of these five variables on achievement in mathematics and reading among Vietnamese Grade 5 pupils were as follows. 1. Teacher Score: Pupils taught by teachers who were more knowledgeable or skillful on the subject matter (i.e. with higher subject-matter scores) were likely to achieve better than pupils taught by teachers with lower subject-matter scores. 2. Teacher Gender: Pupils taught by female teachers were likely to achieve better than pupils taught by male teachers. 3. Classroom Resources/Materials: Pupils in classes with a lot of teaching and learning resources were likely to achieve better when compared to pupils in classes with few or hardly any teaching and learning resources. 4. Average Parental Education & Support: Pupils in classes in which a majority of the pupils had more educated parents and were given educational support were likely to achieve better than pupils in classes in which a majority of the pupils were from less educated parents and were hardly ever provided with educational support. 5. Average Days Absent: Pupils in classes in which a majority of the pupils were never (or rarely) absent from school were likely to achieve better than pupils in classes in which a majority of the pupils were more often absent from school. In addition, for mathematics, the positive coefficients for the class-level variables Excellent Teacher and Teacher Possession indicated the following effects on achievement among Vietnamese Grade 5 pupils when other things were equal. 6. Excellent Teacher: Pupils taught by teachers who were recognised as excellent in teaching were likely to achieve better in mathematics than pupils taught by non-excellent teachers. 7. Teacher Possessions: Pupils taught by teachers who had many possessions (rich) were likely to achieve better in mathematics than pupils taught by teachers who had few or no possessions (poor). 277 Vietnam Reading and Mathematics Assessment Study For reading, the negative coefficients for the class-level variables Average Travel Time and Average Grade Repetition indicated the following effects on achievement among Vietnamese Grade 5 pupils, other things being equal. 8. Travel Time: Pupils in classes in which a majority of the pupils took a shorter time to travel to school (or lived near the school) were likely to achieve better in reading when compared with pupils in classes in which a majority of the pupils took a longer time to travel to school (or lived far from school). 9. Grade Repetition: Pupils in classes in which a majority of the pupils had never repeated a grade were likely to achieve better in reading than pupils who were in classes in which a majority of the pupils had repeated a grade one or more times. School-level model For the two-level models, it can be seen from the results in Table A9.4 that 10 out of the 19 school-level variables examined in this study (listed in Table A9.1) had a significant influence on achievement in mathematics. These 10 variables were: Average Homework, Average Meals, Pupil Behavior, Average Days Absent, Average Teacher Score, Teacher Training, Percent Full Day, Percent Female Teachers, Parental Contribution, and Average Parental Education & Support. All but one (Teacher Training) of these 10 school-level variables that had a significant effect on achievement in mathematics also had a significant effect on achievement in reading in the two-level models. Similar to what was observed for the pupil-level model, the results of the two-level analyses strongly agreed with the results of the three-level analyses in terms of which of the school-level variables had significant influences on achievement in mathematics and reading. Indeed, six school-level variables that had a significant effect on achievement in mathematics and reading in the two-level model also had a significant effect on achievement in mathematics and reading in the three-level models. These six variables were: Average Homework, Average Meals, Pupil Behavior, Percent Full Day, Percent Female Teachers and Parental Contribution. Furthermore, for both mathematics and reading, all except one (Teacher Training, for mathematics) of the school-level variables that had significant effects in the two-level (but not in the three-level) analyses had significant effects at the class-level, and therefore, their effects were observed in both types of models. Hence, for the two- and three level models, the results in Table A9.4 indicated the following school-level effects on achievement in mathematics and reading among Vietnamese Grade 5 pupils, all other things being equal. 3From the SSM file, see (Table AA9.5) 278 Vietnam Reading and Mathematics Assessment Study 1. Average Homework: Pupils in schools where homework was given and corrected more frequently were likely to achieve better than pupils in schools where homework was given and corrected less frequently. 2. Average Meals: Pupils in schools where a majority of the pupils ate more meals per day were likely to achieve better than pupils in schools in which a majority of the pupils ate fewer meals per day. 3. Pupil Behavior: Pupils in schools in which the occurrences of pupil behavior problems were rare were likely to achieve better than pupils in schools in which the occurrences of such problems were very frequent. 4. Percent Full Day: Pupils in schools where pupils were attending full- day school were likely to achieve better than pupils in schools not having full-day schooling. 5. Percent Female Teachers: Pupils in schools where more of the members of staff were female were likely to achieve better than pupils in schools where fewer staff members were female. 6. Parental Contribution: Pupils in schools where parents were involved in school development oriented activities were likely to achieve better than pupils in schools where parents were not involved in such activities. In addition, for the two-level models only, the results in Table A9.4 indicated the following school-level effects on achievement in mathematics and reading, other things being equal. 7. Average Days Absent: Pupils in schools where a majority of the pupils were never (or rarely) absent from school were likely to achieve better than pupils in schools where a majority of the pupils were more often absent from school. 8. Average Teacher Score: Pupils in schools with teachers who had higher subject-matter scores were likely to achieve better than pupils in schools with teachers with lower subject-matter scores. 9. Average Parental Education & Support: Pupils in schools where a majority of the pupils were from more educated homes and were given educational support were likely to achieve better than pupils in schools where a majority of the pupils were from less educated homes and were hardly ever provided with educational support. 10. Teacher Training: Pupils in schools with teachers who had had many years of professional training were likely to achieve better in mathematics (but not necessarily in reading) than pupils in schools with teachers who had fewer years of professional training. This relationship is of interest because it could be observed even after the influence of the variable Average Teacher Score [(8) above], had been controlled for in the model. The following examples illustrate the impact of the model coefficients displayed in Table A9.4 on pupil's achievement. For the two-level model the achievement in mathematics of Pupil 1 whose age is two standard deviations below the mean age of the pupils in Grade 5 279 Vietnam Reading and Mathematics Assessment Study in Vietnam is likely to be 490.56 + {-0.13 x (-12.29 x 2)} = 493.76, that is [intercept + (coefficient twice standard deviation3)]. For the same model, the achievement of Pupil 2, whose age is two standard deviations above the mean age is likely to be 490.56 + {-0.13 x (12.29 x 2)} = 487.36. These calculations assume that all the other things are equal, that is, the two pupils being compared differed only in age and that all their other characteristics (individual and group) were identical. This means that the difference in mathematics achievement between these two pupils due to the influence of age alone could be 493.76 - 487.36 = 6.40 score points on the original mathematics scale. This difference might appear to be small when compared to the estimated grand mean (490.56), but it should be borne in mind that it is greater than two standard deviations of the mean score, and therefore, it is significant at .05 level, and should be of interest to the parents, teachers and policymakers. For the socioeconomic variables and again considering the two-level mathematics model, achievement of Pupil 3 whose score on Parental Education & Support scale is two standard deviations below the mean on this scale is likely to be 490.56 + {6.23 x (-1.05 x 2)} = 477.73. Likewise, based on the same model, the achievement of Pupil 4 whose score on this scale is two standard deviations above the mean is likely to be 490.56 + {6.23 x (1.05 x 2)} = 503.39. Again, these calculations assume that all the other things are equal. Thus, the difference in mathematics achievement between Pupil 3 and Pupil 4 due to the influence of parental education and support alone could be 503.39 - 477.73 = 25.66 score points on the original mathematics scale. Suppose now this effect of Parental Education & Support and the effect of Meals per Day are considered simultaneously in this model. Furthermore, suppose Pupil 3 (from an impoverished home) gets below average meals per day, which when placed on the Meals per Day scale is say two standard deviations below the mean while Pupil 4's (from a rich home) score on this scale is say two standard deviations above the mean. The two-level model predicts that Pupil 3's mathematics score adjusted for influences due Parental Education & Support and Meals per Day is likely to be 477.73 + {2.09 x (- 0.46 x 2)} = 475.81 while the score of Pupil 4 adjusted for these two factors is likely to be 503.39 + {2.09 x (0.46 x 2)} = 505.31. This means that the difference in mathematics achievement between Pupil 3 and Pupil 4 due to the influence of socioeconomic factors at the pupil-level alone could be 505.31 - 475.81 = 29.50 score points on the original mathematics scale. If the effects of socioeconomic status at the school-level are now added, this two-level mathematics model predicts that if Pupil 3 attends a school where Average Parental Education & Support and Average Meals per Day are each two standard deviations below average will be likely to achieve another 25.28 score points lower. On the other hand, if Pupil 4 attends a school where 280 Vietnam Reading and Mathematics Assessment Study Average Parental Education & Support and Average Meals per Day are each two standard deviations above average, this pupil will be likely to achieve an additional 25.28 score points higher. Thus, the difference between mathematics achievement between these two pupils due to socioeconomic factors alone at the pupil and the school-level is likely to be 80.06 score points (29.50 for the pupil-level factors plus 50.56 for school-level factors). It should be borne in mind that this difference (80.06) could increase if the interaction effects between the socioeconomic status variables and other variables were included in the calculations. Interaction effects The results in Table A9.4 also indicated interaction effects in the two-level and three-level models for mathematics as well as reading. There were interaction effects that involve: (a) Pupil-level Homework with the following school-level variables: (i) Average Meals (in the two-level mathematics model only), and (ii) Average Grade Repetition (in the two-level mathematics model and three-level reading model). (b) Pupil-level Parental Education & Support with the following school-level variables: (iii) Average Grade Repetition (in the two- and three-level models for both mathematics and reading), (iv) Management Course (in the two- and three-level models for both mathematics and reading), (v) School Location (in the two- and three-level models for both mathematics and reading), and (vi) Percent Female Teachers (in the three-level models for both mathematics and reading). All the above interactions were cross-level interactions since they involve interaction between variables at a higher level with variables at a lower level (Raudenbush and Willms, 1991). Hox (1995, p.26) has pointed out that "the effect of the interaction and the direct effects of the explanatory variable that make up the interaction must be interpreted as a system". In the sub-sections that follow, discussions of the cross-level interaction effects listed above have been presented. Graphical representations of the interaction effects have been used to enhance these discussions. The coordinates of the graphs used in these discussions were calculated from the 281 Vietnam Reading and Mathematics Assessment Study final estimation of the fixed effects obtained from the final models (results in Table A9.4). Lietz (1996) has described the procedure employed to calculate the coordinates of these graphs. A book by Aiken and West (1996) was consulted for the interpretation of these interaction effects. Homework with Average Meals on Mathematics achievement Figure A9.7 shows the graphical representations of the interaction effects between pupil-level Homework with school-level Average Meals and mathematics achievement. The coordinates of this graph were calculated from the final estimation of the fixed effects obtained from the two-level model for mathematics (see results in Table A9.4). The graphical representation in Figure A9.7 shows that pupils who received more mathematics homework were generally likely to achieve better in mathematics than their counterparts who received less mathematics homework. Moreover, an increase in the amount of homework given was likely to result in a larger increase in mathematics achievement of pupils in schools where the average meals per day was high than of pupils in schools where the average meals per day was low. That is, mathematics homework had more impact in schools with high average number of meals per day. Thus, it appears that if pupils were to get maximum benefit from more homework given, then it would be necessary to provide them with more meals per day, perhaps to get the necessary energy to do the homework. Figure A9.7: Impact of the interaction effect of pupil Homework with the Average Meals per Day in school on Mathematics achievement 282 Vietnam Reading and Mathematics Assessment Study Homework withAverage Grade Repetition on Mathematics (or Reading) achievement In Figure A9.8 the graphical representations of the interaction effects between pupil-level Homework with school-level Average Grade Repetition and mathematics achievement have been shown. The coordinates of this graph were calculated from the final estimation of the fixed effects obtained from the two-level model for mathematics (see results in Table A9.4). A graph similar to the one in Figure A9.8 was obtained for reading based on the three-level results. Thus, because the graphical plots obtained for the two subjects were basically identical, it is considered appropriate to present only one of the plots. Nevertheless, it should be emphasized that whatever conclusion reached in this sub-section for mathematics achievement applies for reading achievement as well. Figure A9.8: Impact of the interaction effect of pupil Homework with the Average Grade Repetition in school on Mathematics achievement In Figure A9.8 it was seen that pupils who received more mathematics homework were generally likely to achieve better in mathematics than their counterparts who were given less mathematics homework regardless of whether the pupils were in schools with many repeaters or in schools with hardly any repeaters. Nevertheless, in situations where pupils were given little or no homework, pupils in schools with many repeaters were likely to achieve better when compared with pupils in schools with few or no repeaters. However, in situations where pupils were given plenty of homework, pupils in schools with few or no repeaters were likely to achieve better in mathematics than pupils in schools with many repeaters. In other words, Homework had more impact in schools with fewer grade repeaters than in schools with more grade repeaters. These are indeed very interesting results because they imply that by adopting 283 Vietnam Reading and Mathematics Assessment Study a policy of giving more homework to the pupils, schools that had a policy against grade repetition could easily outperform schools with a policy in favour of grade repetition. As was mentioned above, the same arguments presented above apply for reading as well since the graphical plots obtained for the interactions effects between pupil-level Homework with school-level Average Grade Repetition and mathematics or reading achievement were basically identical. Parental Education & Support with Average Grade Repetition on Mathematics (or Reading) achievement The graphical plots obtained for the interaction effects between pupil-level Parental Education & Support with school-level Average Grade Repetition and mathematics or reading achievement were basically identical. Thus, in order to avoid repetition, only the graph for mathematics has been presented here but it should be borne in mind that whatever conclusion is reached in this sub-section for mathematics achievement applies for reading achievement as well. Figure A9.8: Impact of the interaction effect of pupil Homework with the Average Grade Repetition in school on Mathematics achievement In Figure A9.8 it was seen that pupils who received more mathematics homework were generally likely to achieve better in mathematics than their counterparts who were given less mathematics homework regardless of whether the pupils were in schools with many repeaters or in schools with hardly any repeaters. Nevertheless, in situations where pupils were given little or no homework, pupils in schools with many repeaters were likely to achieve better when compared with pupils in schools with few or no repeaters. However, in situations where pupils were given plenty of 284 Vietnam Reading and Mathematics Assessment Study homework, pupils in schools with few or no repeaters were likely to achieve better in mathematics than pupils in schools with many repeaters. In other words, Homework had more impact in schools with fewer grade repeaters than in schools with more grade repeaters. These are indeed very interesting results because they imply that by adopting a policy of giving more homework to the pupils, schools that had a policy against grade repetition could easily outperform schools with a policy in favour of grade repetition. As was mentioned above, the same arguments presented above apply for reading as well since the graphical plots obtained for the interactions effects between pupil-level Homework with school-level Average Grade Repetition and mathematics or reading achievement were basically identical. Parental Education & Support with Average Grade Repetition on Mathematics (or Reading) achievement The graphical plots obtained for the interaction effects between pupil-level Parental Education & Support with school-level Average Grade Repetition and mathematics or reading achievement were basically identical. Thus, in order to avoid repetition, only the graph for mathematics has been presented here but it should be borne in mind that whatever conclusion is reached in this sub-section for mathematics achievement applies for reading achievement as well. The coordinates of the graph presented in Figure A9.9 were obtained from the two-level model for mathematics (see results in Table A9.4). Figure A9.9: Impact of the interaction effect of pupil Parental Education & Support with the Average Grade Repetition in school on Mathematics achievement 285 Vietnam Reading and Mathematics Assessment Study From Figure A9.9 it can be seen that regardless of the number of grade repeaters in the school, generally pupils who were from homes with more educated parents, more books and other possessions, and had a private study corner were likely to achieve better in mathematics (or reading) than pupils from homes where the opposite was the case. Nevertheless, pupils from more supportive homes were likely to achieve better in mathematics (or reading) if they were in schools with fewer grade repeaters than if they were in schools with many grade repeaters. On the other hand, pupils from less supportive homes were more likely to perform better in mathematics (or reading) if in schools with many grade repeaters than if in schools with a few or no grade repeaters. That is to say that parental support has more impact in schools with fewer grade repeaters. These results are of interesting because they indicate that pupils from less supportive homes may need to be in schools with greater grade repeating policies to lift their levels of achievement in both mathematics and reading. Parental Education & Support with Management Course on Mathematics (or Reading) achievement The graphical representations of the interaction effects between pupil-level Parental Education & Support with school-level Management Course and mathematics achievement have been presented in Figure A9.10. The coordinates of this graph were obtained from the three-level model for mathematics (see results in Table A9.4) but any conclusions drawn from this plot should apply for reading as well, because an almost identical plot was obtained using the results from the three-level reading model. From Figure A9.10 it can be seen that pupils from more supportive homes were likely to achieve better in mathematics (or reading) if they were in schools where the head teacher had completed a management course than if they were in schools where the teacher had not completed such a course. However, pupils from less supportive homes were more likely to perform better in mathematics (or reading) if in schools where the head teacher had not completed a management course than if in schools where the teacher had completed this course. That is, the parental support had more impact in schools where the head teacher has completed a management course. Thus, it seems that the skills learned from the management course appeared to make the head teacher better able to work with supportive parents but not with non-supportive parents. On the other hand, the head teacher who had not completed a management course seems to help more the pupils from non-supportive parents but appeared not to have the required skills for advising the supportive parents on how to boost their children's academic achievement. Since the effect of the management course was in general beneficial then school heads should be trained in it. However, more needs to be done with the parents to make them more supportive. 286 Vietnam Reading and Mathematics Assessment Study Figure A9.10: Impact of the interaction effect of pupil Parental Education & Support with the completion of Management Course by the head teacher on Mathematics achievement Parental Education & Support with School Location and Mathematics (or Reading) achievement In Figure A9.11 the graphical representations of the interaction effects between pupil-level Parental Education & Support with School Location and mathematics achievement have been presented. The coordinates of this graph were obtained from the two-level model for mathematics (see results in Table A9.4) and any conclusions drawn from this plot apply for reading achievement as well, because a similar plot was obtained using the results from the two-level reading model. It can be seen from Figure A9.11 that pupils from more supportive homes were likely to achieve better in mathematics (or reading) if they were in urban schools or schools located in urban areas than if they were in rural schools or isolated schools. However, pupils from less supportive homes were likely to achieve better in mathematics (or reading) if they were in rural schools or isolated schools than if they were in urban schools. That is parental support was more effective in urban schools than in rural or isolated schools. Parental Education & Support with Percent Female Teachers on Mathematics (or Reading) achievement In Figure A9.12 the graphical representations of the interaction effects between pupil-level Parental Education & Support with school-level Percent of Female Teachers and mathematics achievement have been presented. The coordinates of this graph were obtained from the three-level model for Vietnam Reading and Mathematics Assessment Study Figure A9.11: Impact of the interaction effect of pupil Parental Education & Support with School Location on Mathematics achievement Figure A9.12: Impact of the interaction effect of pupil Parental Education & Support with Percent Female Teachers in school on Mathematics achievement 288 Vietnam Reading and Mathematics Assessment Study mathematics (see results in Table A9.4) but any conclusions drawn from this plot also apply for reading achievement, because an almost identical plot was obtained using the results from the three-level reading model. Thus, in Figure A9.12, the steeper slope for the 'high proportion of female teachers in school' line indicates that parental support had more impact in schools with higher proportions of female teachers than in schools with lower proportions of female teachers. As a result, it emerged that pupils from more supportive homes were likely to achieve better in mathematics (or reading) if they were in schools with a higher proportion of female teachers than if they were in schools with a lower proportion of female teachers. However, pupils from less supportive homes were more likely to achieve better in mathematics (or reading) if in schools with a lower proportion of female teachers than if in schools with a higher proportion of female teachers. The relative strength of the predictors Thus far the results have been given of the effects of the independent variables on the two criteria of pupil reading and mathematics achievement. But, nothing has been said about the relative effect. In other words, which variables were more important than others. In order to compare the relative magnitude of effects of variables, all the variables are standardised before putting them into multilevel analysis. This provides standardised regression coefficients, which when taken in absolute terms, indicate the relative magnitude of effects and can therefore be used to rank the variables in terms of their relative degree of influence on the outcome. However, Hox (1995) reports that the following formula can be employed to derive the standardised regression coefficients from the metric (unstandardised) coefficients. (Unstand. Coeff) (Std. Dev Explanatory Var.) Std. Coeff. = ------------------------------------------------------ [9.1] (Std. Dev. Dependent Var.) (Hox, 1995; p.25) This formula was applied to the metric coefficients in Table A9.4 to get the standardised results in Table A9.5. For example, for the two-level mathematics model, the results in Table A9.5 indicate Average Teacher Score the greatest magnitude of effect on achievement in mathematics (0.20), followed by Average Parental Education & Support (0.08) and Average Homework (0.08). Indeed, for both mathematics and reading and for both two- and three-level analyses, the results in Table A9.5 indicate that Average Teacher Score has the greatest magnitude of effect followed by Average Parental Education & Support and Average Homework. 289 Vietnam Reading and Mathematics Assessment Study It can be seen that grade repetition and pupil behavioural problems are negatively associated with achievement. Estimation of variance explained The results of the final estimation of variance components for the final two- level and three-model and the results of the analyses of the variance components obtained from the null models (provided in Table A9.2) are presented in Table A9.6 in rows 'a' and 'b' respectively. From the information in Table A9.6 rows 'a' and 'b', the information presented in rows 'c' to 'f' were calculated. A discussion of the calculations involved here is to be found in Bryk and Raudenbush (1992; pp.60-76 and p.177). For example, for mathematics, in the three-level analyses, the predictors included in the final model explain 13.5 per cent of 25.1 per cent variance available at the pupil-level and that is equal to 3.4 per cent (that is, 13.5 × 25.1) of the total variance explained at the pupil-level. Similarly, for the same model, predictors included in the final model explain 1.5 per cent (that is, 13.4 per cent of 11.3 per cent) at the class-level, and 20.3 per cent (that is, 31.9 per cent of 63.6 per cent) at the school-level. Thus, the total variance explained by the predictors included in the final three-model for mathematics is 3.4 + 1.5 + 20.3 = 25.2 per cent, which leaves 74.8 per cent of the total variance unexplained. At the school-level, the information provided in Table A9.6 shows that the results of the two-level analyses generally agreed with the results of the three-level analyses regarding the total amounts of variances explained (row 'e'), and consequently, the amounts left unexplained (row 'f'). However, for both mathematics and reading, the total amounts of variances explained in the two-level analyses at the pupil-level were slightly smaller (by around two to three per cent) compared with the combined amounts of variances explained at pupil and the class levels in the three-level analyses. As a result, the total amounts of variances explained in all the levels combined (given in bold in row 'e') in the two-level analyses were slightly smaller (by 1.5 and 3.0 per cent for mathematics and reading respectively) than in the three-level analyses. Thus, the three-level analyses offer only a small advantage over the two-level analyses as far as the amounts of variance explained in the final models are concerned. In addition, the information provided in Table A9.6 shows that within the same type of analysis, the total amounts of variances explained in all the levels combined for mathematics followed closely the amounts explained for reading. Nevertheless, the total amounts explained for mathematics were marginally larger (by around one to two per cent) when compared to the amounts explained in the corresponding reading models. Regardless of the type of analysis employed, the total amounts of variances left unexplained (in the shaded cell of row 'f') for both mathematics and reading are very large ( 75 per cent). These large amount of variance left 290 Vietnam Reading and Mathematics Assessment Study unexplained strongly indicate that there are other important factors influencing the achievement in mathematics and reading of the Vietnamese Grade 5 pupils that have not been included in the models developed in this study. This should raise some concern while interpreting the fixed effects results presented in this appendix. However, it should be borne in mind that the numbers of pupils, classes and schools involved in this study were very large, and therefore, it is unlikely that the inclusion of other significant factors would results in a loss of significance of most of the factors included in the final models. Nevertheless, there is a clear need for a further study to develop models that are the most appropriate for explaining pupil achievement and which maximize the total variance explained. For such models to be achieved, it would require a thorough survey of the situation on the ground to establish what other important factors have been left out in this study. The information obtained from such a survey would then be used in the development of a data collection instrument aimed at capturing the pieces of information needed to construct variables that would maximize the total variances explained. Since the largest proportion in unexplained variance lies at the school-level, and with considerable variability between schools, the survey undertaken should initially be directed towards the identification of characteristics associated with higher performing schools contrasted with lower performing schools. Table A9.6: Variance explained using the two-level and the three-level analyses Two-level Model Three-level Model Pupil School Total Pupil Class School Total (N=36,383) (N=3,623) (N=35,420) (N=7,090) (N=3,622) Mathematics a Null Model 3233.94 6795.77 10029.71 2483.17 1112.43 6290.52 9886.13 b Final Model 2978.61 4677.00 2147.57 963.16 4285.67 c Var. Available 32.2% 67.8% 25.1% 11.3% 63.6% d Var. Exp. 7.9% 31.2% 13.5% 13.4% 31.9% e Total Var. Exp. 2.5% 21.1% 23.7% 3.4% 1.5% 20.3% 25.2% f Var. left UnExp. 29.7% 46.6% 76.3% 21.7% 9.7% 43.4% 74.8% Reading a Null Model 3999.46 5966.23 9965.68 3216.29 1091.07 5442.07 9749.43 b Final Model 3745.39 4088.33 2830.19 919.36 3618.78 c Var. Available 40.1% 59.9% 33.0% 11.2% 55.8% d Var. Exp. 6.4% 31.5% 12.0% 15.7% 33.5% e Total Var. Exp. 2.5% 18.8% 21.4% 4.0% 1.8% 18.7% 24.4% f Var. left UnExp. 37.6% 41.0% 78.6% 29.0% 9.4% 37.1% 75.6% 291 Vietnam Reading and Mathematics Assessment Study Conclusion In this appendix the multilevel analyses undertaken in order to tease out the pupil, class and school level factors influencing achievement in mathematics and reading among Grade 5 Vietnamese pupils have been reported. For each of the two outcome measures (mathematics and reading), two different types of hierarchical models were hypothesised and examined in this appendix: two-level and three-level models. Generally, the results of the two-level analyses agreed with the results of the three-level analyses regarding the amounts of variance available to be explained especially at the school-level. The results from the two types of models were also very similar in terms of the amounts of total variance that was explained in the final models. The results were also very similar for the individual-level and group-level factors that had significant effects on achievement in mathematics and reading. Despite the small amounts of variance explained in the final models, the results described were extremely interesting especially because they showed how teacher characteristics such as teacher subject matter knowledge and teacher training influenced pupil achievement. The results were also interesting because they showed how individual-level and group-level (class or school) factors influenced pupil achievement when considered simultaneously. For example, the results from three-level models showed that the variable Average Days Absent at the class-level had a negative effect on pupil achievement in both mathematics and reading. Most important, the negative effect of the variable Average Days Absent was significant (.05) despite the fact that the effect of the variable Days Absent had been included in the model at the pupil-level. This implied that a high rate of absenteeism at the class-level affected regular attendees within the class as well. Such a conclusion could not have been arrived at based on the conventional multiple regression approaches, which do not allow for modelling of the effects of the factors at the pupil and group levels, together with cross-level interactions carried out simultaneously. References Aiken, L. S. and West, S. G. (1996). Multiple Regression: Testing and Interpreting Interactions. Newbury Park: Sage Publications. Bryk, A. S. and Raudenbush, S. W. (1992). Hierarchical Linear Models: Application and Data Analysis Methods. Newbury Park: Sage Publication. Bryk, A. S., Raudenbush, S. W. and Congdon, R. T. (1986). Hierarchical Linear Modeling with the Hierarchical Linear Modeling/2L and Hierarchical Linear Modeling/3L Programs, [Computer Software]. Chicago: Scientific Software International. Howie, S. (2002). English Language Proficiency and Contextual Factors Influencing Mathematics Achievement of Secondary School Pupils in South Africa. Enschede: Print Partners Ipskamp. Hox, J. J. (1995). Applied Multilevel Analysis. Amsterdam: TT-Publikaties. 292 Vietnam Reading and Mathematics Assessment Study Keeves, J. P. (1997). Suppressor Variables. In Keeves J. P. (Ed.). Educational Research, Methodology, and Measurement: an International Handbook (pp.695-696). Oxford: Pergamon Press. Kreft, I. G. G. (1995). The Effects of Centering in Multilevel Analysis: Is the Public School the loser or the winner? A New Analysis of an Old Question. Paper Presented at AERA 1995, Section D. San Francisco, April 18-22 1995. Kreft, I. G. G., Leeuw, J. and Aiken, L. S. (1995). The Effects of Different Forms of Centering in Hierarchical Linear Models. Multivariate Behavioral Research, 30 (1), 1-21. Lietz, P. (1996). Changes in Reading Comprehension across Culture and over Time. German: Waxman, Minister. Pituch, K. A. (1999). Describing School Effects with Residuals Terms. Evaluation Review, 23 (2), 190-211. Raudenbush, S. W. and Bryk, A. S. (2002). Hierarchical Linear Models: Applications and Data Analysis Methods (2nd ed.). California: Sage Publications. Raudenbush, S. W. and Willms, J. D. (1991). The Organisation of Schooling and Its Methodological Implications. In S. W. Raudenbush and J. D. Willms (Eds.), Schools, Classrooms, and Pupils: International Studies of School from a Multilevel Perspective (pp. 1-12). San Diego: Academic Press. Raudenbush, S. W., Bryk, A. S., Cheong, Y. F. and Congdon, R. T. (2000). HLM 5: Hierarchical Linear and Nonlinear Modeling, [Computer Software]. Lincolnwood, IL: Scientific Software International. Raudenbush, S.W. and Bryk, A.S. (1994). Hierarchical Linear Models. In Husén, T and Postlethwaite. T. N. (Eds.). International Encyclopedia of Education: Research and Studies (Second Edition), Oxford: Pergamon Press. pp. 2590-2596. Willms, D. J. and Somers, M. A. (2001). Family, Classroom, and School Effects on Childrens Educational Outcomes in Latin America. School Effectiveness and School Improvement, 12 (4), 409-445. Table AA 9.1: Correlation matrix at the pupil-level Age Meal Pupil Pupil Ethnic Home in per Travel Days Private Grade Homework Homework Reading Math Back- Possessions Month Day Time Absent Corner Repetition (maths.) (reading) Score Score ground & Care Age in Month 1.00 Meal per Day -0.12 1.00 Travel Time 0.17 -0.05 1.00 Days Absent 0.11 -0.05 0.08 1.00 Private Corner -0.15 0.12 -0.04 -0.07 1.00 Grade Repetition 0.45 -0.05 0.10 0.08 -0.09 1.00 Homework (mathematics) -0.04 0.05 -0.01 -0.06 0.05 -0.03 1.00 Homework (reading) -0.04 0.06 -0.02 -0.06 0.08 -0.03 0.55 1.00 Reading Score -0.21 0.15 -0.12 -0.12 0.16 -0.16 0.15 0.16 1.00 Math Score -0.18 0.15 -0.11 -0.12 0.15 -0.14 0.16 0.18 0.74 1.00 Ethnic Background -0.34 0.05 -0.22 -0.11 0.19 -0.23 0.04 0.05 0.21 0.16 1.00 Home Possessions & Care -0.34 0.21 -0.19 -0.14 0.62 -0.21 0.10 0.12 0.34 0.32 0.41 1.00 293 Table AA 9.2: Correlation matrix at the class-level Table AA 9.3: Correlation matrix at the school-level Table AA9.4: Meaning of the variable codes used in HLM analyses Level Variable Label Variable Name Pupil XPAGEMON Pupil age in month PMEAL Number of meals per day PTRAVEL Pupil travel time to school PABSENT Days absent in the last month (March 2001) PPRIVCOR Private study corner at home XPGREP Grade repetition XPMATHHW Mathematics homework get and correction XPREADHW Reading homework get and correction XCHMETHN Ethnic background XCEDCARE Parental education and care Class XPAGEM_1 Pupil age in month aggregated at the class-level PMEAL_1 Number of meals per day aggregated at the class-level PTRAVEL_1 Pupil travel time to school aggregated at the class-level PABSEN_1 Days absent in the last month aggregated at the class-level PPRIVC_1 Private study corner at home aggregated at the class-level XPGREP_1 Grade repetition aggregated at the class-level XPMATH_1 Mathematics homework and correction aggregated at the class-level XPREAD_1 Mathematics homework and correction aggregated at the class-level XCHMET_1 Ethnic background aggregated at the class-level XCEDCA_1 Parent support for education and socioeconomic status aggregated at the class-level XHAVMA_1 Pupil have materials in class XTRESC_1 Classroom resources (max=16) TRD500_1 Teacher reading score(Mean = 500, SD = 100) TMA500_1 Teacher mathematics score (Mean = 500, SD = 100) TSEX_1_1 Teacher sex XTTTR__1 Years of teacher professional training XTEX_1_1 Excellent teacher XTHPOS_1 Teacher total possessions (max=21) TPCTME_1 Percent parents met School XPAGEM_2 Pupil age in month aggregated at the school-level PMEAL_2 Number of meals per day aggregated at the school-level PTRAVEL_2 Pupil travel time to school aggregated at the school-level PABSEN_2 Days absent in the last month aggregated at the school-level PPRIVC_2 Private study corner at home aggregated at the school-level XPGREP_2 Grade repetition aggregated at the school-level XPMATH_2 Mathematics homework and correction aggregated at the school-level XPREAD_2 Mathematics homework and correction aggregated at the school-level XCHMET_2 Ethnic background aggregated at the school-level XCEDCA_2 Parent support for education and socioeconomic status aggregated at the school-level XSPBHP_2 School head's perception on average occurrence of pupil behavior problems XSTBHP_2 School head's perception on average occurrence of teacher behavior problems XSREST_2 School resources (max=27) TRD500_2 Teacher reading score aggregated at the school-level (Mean = 500, SD = 100) TMA500_2 Teacher mathematics score aggregated at the school-level (Mean = 500, SD = 100) TSEX_2_2 Teacher sex XSTT_1_2 Average years of staff teacher professional training XSSHFT_2 Per cent full day pupils XTHPOS_2 Teacher total possessions (max=21) SMANAG_2 Management course XLOCA_2 School location (urban/rural) XSFEMT_2 Per cent female teachers XSPARC_2 Parental contribution (max=14) XSPROG_2 Reading programs XCOMM__2 Community (developed/undeveloped) Table AA 9.5: Descriptive Statistics of the variables included in the two-level SSM file Table AA9.6: Descriptive Statistics of the variables included in the three-level SSM file LEVEL-1 DESCRIPTIVE STATISTICS VARIABLE NAME N MEAN SD MINIMUM MAXIMUM XPAGEMON 34212 135.55 12.29 100.00 238.00 PMEAL 35733 2.71 0.46 1.00 3.00 PTRAVEL 35381 17.83 13.60 0.00 150.00 PABSENT 35632 0.67 1.53 0.00 30.00 PPRIVCOR 35539 1.79 0.41 1.00 2.00 XPGREP 35562 0.24 0.53 0.00 3.00 XPMATHHW 35690 -0.01 1.02 -6.94 0.92 XPREADHW 35699 0.00 1.00 -4.55 1.02 PRD500 35760 492.91 101.68 62.85 892.75 PMA500 35755 493.51 101.19 21.98 919.34 XCHMETHN 35763 -0.29 1.27 -3.63 0.37 XCEDCARE 34652 -0.16 1.04 -4.57 3.69 LEVEL-2 DESCRIPTIVE STATISTICS VARIABLE NAME N MEAN SD MINIMUM MAXIMUM XPMATH_1 7090 -0.00 0.71 -6.94 0.92 XHAVMA_1 7090 8.17 1.08 2.00 10.00 XPREAD_1 7090 0.00 0.75 -4.55 1.02 TRD500_1 7090 497.57 99.86 81.05 761.57 TMA500_1 7090 496.84 100.19 67.42 707.82 TSEX_1_1 7090 1.74 0.44 1.00 2.00 XTTTR__1 7090 2.05 0.57 0.00 4.00 XTEX_1_1 7090 0.70 0.46 0.00 1.00 XTRESC_1 7090 10.65 2.06 0.00 16.00 XTHPOS_1 7090 11.12 3.53 0.00 21.00 TPCTME_1 7090 79.72 17.60 0.00 100.00 XHOMEB_1 7090 -0.16 1.01 -3.93 2.71 PMEAL_1 7090 2.72 0.34 1.50 3.00 XCHMET_1 7090 -0.16 1.07 -3.63 0.37 XPAGEM_1 7090 135.05 8.25 117.50 195.00 PTRAVE_1 7090 17.35 8.74 1.00 90.00 PABSEN_1 7090 0.64 0.86 0.00 8.00 PPRIVC_1 7090 1.80 0.27 1.00 2.00 XPGREP_1 7090 0.22 0.31 0.00 3.00 XCEDCA_1 7090 -0.08 0.84 -4.18 3.50 LEVEL-3 DESCRIPTIVE STATISTICS VARIABLE NAME N MEAN SD MINIMUM MAXIMUM XPMATH_2 3622 -0.00 0.60 -6.94 0.92 XPREAD_2 3622 0.00 0.64 -4.55 1.02 PMEAL_2 3622 2.72 0.31 1.77 3.00 PTRAVE_2 3622 17.47 7.56 1.13 75.00 XSPBHP_2 3622 0.28 0.17 0.00 1.00 XSTBHP_2 3622 0.17 0.16 0.00 1.00 PABSEN_2 3622 0.64 0.71 0.00 7.00 XPGREP_2 3622 0.23 0.26 0.00 3.00 TRD500_2 3622 497.01 92.94 98.93 761.57 TMA500_2 3622 496.11 94.74 111.52 707.82 SMANAG_2 3622 1.79 0.40 1.00 2.00 XSLOCA_2 3622 2.06 0.59 1.00 3.00 XSTT_1_2 3622 1.97 0.33 0.65 3.35 XSSHFT_2 3622 0.17 0.32 0.00 1.00 XSREST_2 3622 10.29 4.80 0.00 25.00 XSFEMT_2 3622 0.78 0.19 0.00 1.00 XSPARC_2 3622 3.69 2.23 0.00 14.00 XSPROG_2 3622 -0.16 0.89 -2.47 2.38 XCOMM__2 3622 -0.12 1.10 -7.11 1.25 PPRIVC_2 3622 1.80 0.24 1.00 2.00 XCHMET_2 3622 -0.18 1.05 -3.63 0.37 XPAGEM_2 3622 135.19 7.37 122.88 198.83 XCEDCA_2 3622 -0.10 0.78 -3.75 2.85 Vietnam Reading and Mathematics Assessment Study Chapter 10 SUMMARY AND CONCLUSIONS Introduction T his study has been a cross-sectional sample survey of Grade 5 This was the first large- pupils' achievement together with general information on the scale national survey of pupils, their teachers, and their schools. Reasonably accurate pupil achievement.. estimates were required for each of 61 provinces as well as for the country as a whole. This involved collecting data from 3660 schools. Only one day was allowed for data collection. The testing time took the whole day. This was the first time in Vietnam that a national survey of this kind in education had been carried out. The study has yielded an enormous amount of information as can be seen from the foregoing chapters. As a first survey ir could not provide information on every aspect of the educational system. Indeed, one proposal for a follow-up activity is that more information, using in some instances finer measures and involving observation and interviews, should be collected. The data analyses and writing of this report were conducted in the period March to July, 2002. There are many more analyses that could be undertaken. However, it is believed that the major analyses have been already carried out and the major results are known. This is not to say that further analyses of the existing data set should not be undertaken. As many as possible should be undertaken and reported over the months ad years to come. From the analyses undertaken so far and presented in this publication, a series of recommendations have been made to the MOET for its consideration. In this final chapter an attempt has been made to draw together the various suggestions made throughout the book. It has been seen that the conditions of schooling are far from ideal. The various suggestions have been clustered into seven major categories later in this chapter. 299 Vietnam Reading and Mathematics Assessment Study Review of major findings First, the major findings The same categories have been used as later for the recommendations. It was have been reviewed, thought to be desirable to review the major findings before coming to the and... actual recommendations. The categories concern different aspects of schooling. They do not reflect any priority of action. Comments on the priority are made after the recommendations. The Home ...categorized into seven The attitudes of the parents towards school and schooling are important. It is themes: well known that when home and school work in co-operation this can lead to ˇ Home... better achievement by the children. One of the first findings was that there were fewer girls than boys in Grade 5 and in some provinces there were only 40 percent girls enrolled. This home problem was more acute in the isolated school areas. It was felt that school intervention programs may well help in this matter. This is where the school needs to attract the mothers of such pupils (in and out of school) to come to school ostensibly to work on some project or other but the hidden object is to have the teachers in a subtle way get the message to the mothers that daughters should go to school and not drop out or be absent. Mothers should also be 'taught' to show an interest in what their children are doing at school by asking them to relate in detail what they learned each day and also to ask their children to read to them or to tell them the story of what they have just read. Since meals per day turned out to be quite a strong predictor of pupil achievement then this too is a matter to be stressed with mothers. Materials and textbooks in classrooms and at school ˇ Materials and Several pupils were in classrooms where they did not have the wherewithal textbooks in classrooms to write or read. In some cases there was no blackboard and in other cases no and schools... teacher guide. Although 99 percent of pupils had the Vietnamese and Math textbooks there were many supplementary textbooks that pupils did not have. It was primarily homes that were poor where the parents could not afford these supplementary materials. The personnel of the curriculum group should review the necessity of all of these materials being available to pupils. The group could prepare a consolidated list of books and materials that all pupils should have. Material resources ˇ Other material Questions were asked about 27 different school resources. However only resources... 12 of them were on the benchmark list of the Ministry. Although the percentage of pupils in schools meeting the benchmarks was high, there 300 Vietnam Reading and Mathematics Assessment Study were still some pupils in schools that did not meet the benchmarks. The same was true of classroom benchmarks. Where pupils were in classrooms without blackboards, it would seem that there would be not much chance of learning occurring. It was suggested, inter alia, that the benchmark list be revisited and revised and that one final list be produced instead of the two lists at present and then each list sub-divided into two different sets of benchmarks, one for rural and urban schools and one for isolated schools. Although the Ministry has a set of national standards, it is quite clear that the reality of the implementation is not fully related to the national standards. It is important that there be an approach to implementation that emphasises increments from the base to increase the standards gradually. It was also clear that there should be a system of collecting data in a systematic fashion about the implementation in schools of the basic material resources. These data could be kept on a data base that is used for ensuring the gradual implementation of resources until all schools meet the standards. What was surprising was the dearth of reading books, in the form of a classroom library or school library, in the schools. Indeed there seemed to be a lack of a reading culture in the schools. It was suggested that the MOET should consider a book flood program. Book flood programs have worked well elsewhere and also increased the achievement levels of reading comprehension of pupils. Above all, there was a good deal of variance among schools for school and classroom resources including books in classroom library. Again it was the isolated schools that had least, the rural next least and the urban schools that had most. Human resources in schools Teachers are the main human resources in schools. It has been shown that for ˇ Human resources... a pupil to have female teachers with more years of teacher training is an advantage. It was also shown that where the whole staff was more female dominated and where the whole staff had a good level of teacher training, then pupils in such schools had higher achievement levels. Finally it was shown that there were several teachers with low achievement levels in both math and reading and that where the teachers in provinces had low levels of achievement then so did the pupils. Indeed teacher subject matter knowledge was the largest determinant of pupil achievement. As such it is the most important factor to be tackled by the Ministry. All kinds of action can be taken to ensure that the subject matter knowledge of teachers on both reading and math is improved. 301 Vietnam Reading and Mathematics Assessment Study It was further shown that the quality of teachers based on their mathematical knowledge or reading knowledge as well as on their years of teacher training and sex were very unevenly distributed across schools. It was the isolated schools that suffered the most. This imbalance must be corrected and in the first instance there maybe should be more human resources allocated to those schools that have been most deprived until now. Again, a system of collecting and storing data about teachers with a view to using such a data base for allocation of teachers to schools should be implemented. If the overall level of teacher subject matter knowledge is raised then in theory the imbalance across schools should be much diminished. This is a matter for close monitoring. It was shown that the average contact time in Vietnam was below the average for many OECD countries. The MOET may wish to consider revising the number of desirable contact hours. Process variables ˇ Process of schooling... The most important variable to emerge from the list of process variables was that teachers who gave homework and corrected it had pupils with higher achievement than teachers who neither gave no home work or gave it and then did not mark it. This has direct implications for teacher training. It was also a fact that many teachers believed that the best activity for pupils to learn was to learn by reciting tables and in general by learning by rote. The test itself did not really assess what might be called the higher order skills of the understanding of principles and the application of knowledge and comprehension for solving a new problem. It was felt that more research and development work was needed on optimal approaches to the learning of higher order skills in order to ensure 'better' learning in future. Achievement ˇ Achievement... In general achievement was higher in urban areas than in rural areas and higher in rural areas than in isolated areas. It was the very low levels of achievement in some isolated areas that were noteworthy. It was felt that some further guidance from the curriculum centre about skill based assessment to help teachers assess where their pupils are would be useful. It was further felt that some intervention to help teachers in targeted areas improve their own standards of subject matter knowledge would be desirable. 302 Vietnam Reading and Mathematics Assessment Study Further data collections In some cases the data from this study did not allow particular questions to ˇ Further data be investigated and further small studies were suggested. Above all, it was collections. suggested that repeat surveys of the kind just conducted should be carried out about every four or five years. In order that comparisons can be made with the present study the first step is to develop a good archive of the present data set and accompanying instruments and notes. This should be begun immediately. Plans should also be made for the next study. It would be desirable that those responsible for the study learn some basic statistics and learn to read English, the language in which 90 percent of research studies and research methodology is written up. There were several aspects of primary education that were not examined in sufficient detail in the current study. These included more information on the conditions of schooling on satellite campuses, multi-grade schooling, extra tuition, teaching methods in both subject areas, and part versus full-day schooling. These aspects could be covered in extra small research studies. The demands for action Many of the suggestions listed later in the chapter will demand a high level of co-operation between the communities, the provincial authorities and the central government. Thus far, as far as the author is aware, the general principle has been that the central authorities determine policy but it is at the provincial and district levels that action occurs. But it is obvious from some of the results that the central authorities need to monitor what is going on at the provincial level in order to check on equity issues. Where there are disparities among the provinces, then it is up to the central government to point out to particular provinces that they are lagging behind in matters of equity and together the central authorities and provincial authorities must take action to redress the matter. All of the persons involved must have the determination to make things happen in the first place. The policy suggestions have been written in a provocative manner with the aim of generating wide spread discussion and debate. It is to be hoped that the officials within the MOET will join this process of debate and when the various actions have been determined then it might be useful if the MOET were to form a small committee to ensure that the actions occur when they are meant to occur. This committee might report directly to one of the Vice-Ministers. 303 Vietnam Reading and Mathematics Assessment Study In the following section the policy suggestions have been grouped under the headings used at the beginning of this chapter. Next to each suggestion three pieces of information have been supplied: 1. Relevant department(s): The name off the unit, branch or department in the central ministry and/or provincial office that should be given responsibility for initiating action with respect to each policy suggestion. The recommendations in 2. Implementation time: A very approximate time estimate for each of the seven implementing each policy suggestion based on a three-point scale: categories have been 'short' - around three to nine months; 'medium' - around one to two presented together with the unit(s) responsible years; and 'long' - around three to five years. and an implication of 3. Estimated costs: An approximate cost estimate for implementing each the time and cost likely policy suggestion according to a three-point scale: 'low' - for initiatives to be involved in the that require no increased expenditure and could be accommodated implementation of the within existing budgets through redeployment of staff, more efficient recommendations. use of resources, and redefining existing procedures; 'moderate' for activities that need to be funded in addition to current Ministry operations; and, 'high' - for large-scale investments in capital works and human resources. It is suggested that the Ministry work on establishing priorities for the various actions that ensue from the suggestions. The Home The following suggestions concern the home of the pupils. The suggestions are particularly aimed at pupils in isolated area schools. Suggestion Unit(s) responsible Time Cost Policy suggestion 3.3: The MOET may wish to consider introducing home intervention programs through schools with associated radio programs in order to increase parental interests in their children's education. These programs should involve showing parents how to exhibit their interest in the children's education, have them send their MOET daughters to schools, keep them there, reduce the children's time spent helping the family, ensuring that homework NIES has priority, and trying to provide a private study corner. The radio program should emphasize similar behaviours. PE Dept The Mekong Delta might be a place to start. Provinces Short Low Policy suggestion 3.5: The provincial offices might like to consider ways of educating parents in the importance of regular meals and nutrition. In addition, specific provinces may wish to provide meals in targetted schools. Provinces Short Low Policy suggestion 5.2 b) The distance from public amenities for some very isolated communities is large and the possibility of increasing community cultural activities associated with school learning may be desirable. Provinces Medium Medium PE Dept Policy suggestion 5.11: The MOET might wish to consider establishing a study group to examine which action NIES might be taken for school-home relationship. Provinces Short Low 304 Vietnam Reading and Mathematics Assessment Study Materials and textbooks at school The following suggestions concern all of Vietnam but the isolated areas were in fact somewhat worse off. Suggestion Unit(s) responsible Time Cost Policy suggestion 3.2: The MOET may wish to consider increasing the material and human resources to isolated MOET schools. Provinces Medium High Policy suggestion 3.4: The MOET might wish to consider: a) getting up-to-date information on the availability of supplementary reading materials and library books; b) strengthening the current availability of the three different types of books in school libraries; c) the possibility of having book flood programs in selected schools; and d) PE Dept. Short to encouraging the PTAs to organize the donation of the books to the libraries. Provinces medium Medium Policy suggestion 4.5: The MOET might (a) wish to review and state which materials and textbooks as a core packet are really required by all pupil; b) conduct a provincial review of the lack of materials and books in all schools; and c) consolidate one national curriculum in all schools. PE Dept. Medium Low Policy suggestion 6.1: The Ministry might wish to review and, if necessary, to revise benchmark standards for the educational environment that are deemed to be reasonable for the proper functioning of primary schools. The notion of incremantally improving resources rather than having the 'model' idea would probably be fruitful. PE Dept Short Low Material resources excluding textbooks Suggestion Unit(s) responsible Time Cost PE Dept. Policy suggestion 5.2: a) Given the age of some school buildings, the MOET may wish to conduct a review of the P & F Dept quality of all primary school buildings. Provinces Short Low Policy suggestion 5.9: The MOET might wish to compare the findings from this study with their own assessments of the condition of school buildings. If there are grounds for concern, then appropriate plans should be made to have a detailed four year plan of rebuilding. Priority should be given to the isolated schools in the Northwest and PE DEpt. Northeast regions. P & F Dept Medium High Policy suggestion 5.10: The MOET may wish to decide on the minimum resources needed for a school to function PE Dept. well and then re-run the analyses in order to target Educational Priority Areas (EPAs). NIES Short Low Policy suggestion 7.1: The MOET may wish to consider equalizing classroom resources and school resources PE Dept. between provinces through ensuring that between school and classroom resource differences within a province are P & F Dept. addressed at the provincial level. Provinces Long High Human resources (including teacher training) Suggestion Unit(s) responsible Time Cost Policy suggestion 4.1: The MOET and provincial authorities might wish to review the system of how teachers are allocated to schools and then revise the procedures in order to ensure more equity of the quality of the teaching force in all schools. For a short period of time the MOET might even consider having better quality teachers in the isolated schools in order to compensate for the relatively poor teachers they have had over the last so many years. Teacher Dept. It is also suggested that a rotation scheme be tried out whereby good teachers from good schools much spend Personnel Dpt. some time in schools in isolated areas. However, it is recognised that it is very difficult to devise good incentives for PE Dept. teachers to work in isolated areas. Provinces Long Medium 305 Vietnam Reading and Mathematics Assessment Study Policy suggestion 4.2: The MOET might wish to consider devising various means of increasing the subject matter knowledge of existing and future teachers of reading and mathematics. This may involve updating the existing teacher training course content to take account of the new curriculum and ensuring a national level of subject matter Teacher Dept. knowledge for the teachers to become certified. For existing teachers it may involve special in-service courses or PE Dept. special programs targetted at specific groups of schools. NIES Long Medium Policy suggestion 4.5c: The MOET might wish to consider introducing a system of re-allocation of teachers to schools, especially in the poor and isolated areas in order to achieve an equity of the allocation of quality teachers to all schools. ? ? ? Policy suggestion 4.6: The MOET should clarify the roles of the inspectorate to all schools emphasizing the MOET supportive rather monitoring role, and then have a good control mechanism to ensure that that the system Inspectorate works well. PE Dept Short Low Policy suggestion 5.1: The MOET may wish to review the data on the distribution of school heads' characteristics across schools and, together with the relevant provincial directors of education, devise methods whereby there will be a more equitable allocation of heads to schools in terms of their education, teacher training and management PE Dept training. Every effort should also be made to have the heads teaching at least two or three periods a week. Provinces Long Low Policy suggestion 5.8: The MOET might wish to consider including content in the school management program and PE Dept. in-service teacher training courses in order to have school heads and teachers diminish these problems in school. Teacher Dept. Medium Low Policy suggestion 7.2: The provincial authorities in the regions of the Central Highlands and the Southeast and Mekong Delta should take action to decrease the variation in the school heads' teaching experience (Central Highlands) and the number of inspectors visits to schools (Southeast and Mekong Delta). Provinces Medium Low Process Suggestion Unit(s) Time Cost responsible PE Dept. Policy suggestion 2.4: The MOET might wish to consider specific training for district and provincial officers in NIES techniques for monitoring and intervention against competency levels and benchmarks. Provinces Medium Low Policy suggestion 2.5: The MOET might wish to consider accelerating the shift in the primary school reading curriculum to ensure that more pupils are adequately prepared for the wider range of genre of reading encountered in secondary school learning materials. This can be helped by the introduction of more books into the classroom PE Dept. and school libraries. NIES Medium Low Policy suggestion 2.6: The MOET might wish to consider urgent action for schools in low performing regions where isolation and poverty combine to align with low competency levels (in EPAs). The action will consist of an integrated MOET package involving action in the homes and in the classrooms and schools. NIES Medium Medium Policy suggestion 2.7: The MOET might wish to consider urgent intervention plans for teachers and pupils with lower performance levels, especially in EPAs where the combined effects of isolation and poverty are linked with low PE Dept. teacher skills. Teacher Dept. Short High Policy suggestion 3.6: The MOET might wish to conduct experiments on ways of having children speak more Vietnamese. These experiments might include a) enrolling these children in kindgarten and having special PE Dept. Vietnamese lessons; b) having intensive three months Vietnamese course for such children before they begin Cen. for Ethnic school; c) having a more flexible curriculum such that an increased amount of reading and speaking Vietnamese Education becomes cross-curricular: and d) introducing the concept of a flexible curriculum and cross-curricular activities in all Teacher Dept. teacher training courses and school management courses. Provinces Medium High Policy suggestion 3.7: The MOET and the provincial authorities should re-stress in their guidelines the importance of giving and correcting homework. In the monitoring (inspection) system special checks should be made in the PE Dept. schools to ensure that this is happening. Provinces Short Low Policy suggestion 4.4: a) the MOET might wish to reconsider the minimum and maximum contact hours per week PE Dept. for teachers, given the existing low number of hours compared with other countries; b) the MOET may wish to Teacher Dept. undertake a special study of teachers meeting parents and then improve the teacher training programs for Personnel Dept. improving school, family, and community interactions. NIES Long High Policy suggestion 5.7: In order to strengthen the culture of reading in schools, the Curriculum Centre may wish to decide on which reading associated activities (e.g. pupils producing a school magazine) in school are desirable and NIES then experiment on the best way to have them run. PE Dept. Short Low Policy suggestion 8.1: The education advisors and the management course personnel should encourage heads to PE Dept. teach each week. Provinces Short Low 306 Vietnam Reading and Mathematics Assessment Study Achievement Suggestion Unit(s) responsible Time Cost Policy suggestion 2.1: The MOET might wish to consider the introduction of curriculum and assessment frameworks and benchmarks written in criterion referenced (profile) format. This can be done as the new curriculum PE Dept. is introduced. It will encourage a competency-based developmental interpretation of pupil performances. NIES Medium Low PE Dept. Policy suggestion 2.2: The MOET might wish to use developmental assessment procedures to plan intervention NIES programs for specific groups of pupils and teachers. Provinces Medium Low Policy suggestion 2.3: The MOET might wish to consider the introduction of performance monitoring programs at province and district level in which school, district and province level performances are monitored against NIES competency levels and benchmarks. Provinces Medium Low Further data collections Suggestion Unit(s) responsible Time Cost MOET Policy suggestion 1.1: The MOET might wish to begin to implement the way in which the data from its annual PE Dept. census of schools is recorded. It may wish to modify some of the questions in the census form. Center for EMIS Short Low Policy suggestion 3.1: The MOET might wish to undertake follow-up surveys of Grade 5 at five year intervals in PE Dept. order to establish, inter alia, changes in important educational indicators over time. NIES Medium Medium Policy suggestion 3.8: The MOET might wish to consider mounting a separate detailed study of extra tuition in PE Dept. selected provinces in order to obtain a much more detailed account of what is happening, why, how, and if it can NIES used to improve pupil attainment. Teacher Dept. Short Low Policy suggestion 4.3 The MOET might consider a review in annual school census of amount of classroom furniture and teaching materials in order to identify shortfall. A start could be made in schools with satellite campuses. NIES Short Low Policy suggestion 4.7: It is suggested that a small R&D team conduct a detailed study of new teaching methods that will result in appropriate methods being recommended. The team could use the results just reported above as a PE Dept. starting point, but then review the international literatures on the matter, devise the new methods and experiment NIES with them until the best new and appropriate methods are known. Teacher Dept. Short Low Policy suggestion 5.4: This study has been concerned with cognitive learning only. Given that the social development is also an important feature of primary schooling, it might be desirable for the MOET to initiate a study on the social development of pupils in primary school (and in that context examine the relationship between the PE Dept. social integration of pupils and total enrolment of schools). Provinces Short Low PE Dept. Policy suggestion 5.11: The MOET might wish to consider establishing a study group to examine which action NIES might be taken for school-home relationship. Provinces Short Low PE Dept. Policy suggestion 8.2: The MOET should mount a special study in order to examine in detail the kinds of NIES behavioural problems expressed by heads about both pupils and teachers. Provinces Short Low Policy suggestion 8.3: The effective schools in isolated areas are of great interest. Some of the variables that are different between more and least effective schools have been identified. However, there must be other factors in the most effective schools that have not been and are important. It is important to know what these are in order for all PE Dept. isolated schools to be improved. The MOET may well wish to consider undertaking a small observational study of NIES selected most effective schools in isolated areas. Provinces Short Low 307 Vietnam Reading and Mathematics Assessment Study What to change when this study is repeated For the next major This was a first attempt at undertaking a national sample survey of primary study, it will be only schooling. Since one of the aims was to obtain 'good' estimates of pupil necessary to conduct a achievement and pupil, teacher, and school independent variables at the national survey and not province level a large sample was drawn. It is suggested that in future it will to draw a sample for be sufficient to draw a national sample for good estimates at the national level each province and not at the province level. This will considerably reduce the sample size separately. This will be and cost of conducting the data collection. To obtain a sampling error of a considerably decrease percentage of 5 percent at the 95 percent confidence limits and with the roh the cost of study. of 0.664 (say 0.70) for Grade 5 schools in the country it will be necessary to have a sample of 286 schools and with 20 randomly drawn Grade 5 pupils per school the number of pupils involved will be 5720. Nevertheless, although the sample will be much smaller, new kinds of problems will be faced. There are always new and different problems that arise even though it is thought that the project is an exact repeat of the first one. This requires flexibility on the part of the researchers. If comparisons are to be made with the results from this first project then the maxim 'If you want to measure change, then don't change the measure' is wise advice. There are two questions that should be added. The first one is to have information from the pupil on which satellite campus or main school that he or she is in. The second concerns information, again from the pupil, on whether or not he is in a multi-grade class and if so what kind of multi- grade class. Finally, other variables should be added to the questionnaires. It was only possible to explain some 20 percent of the variance and this is not very much. There are other variables that were not collected in this first study that were clearly important in explaining variance among pupils' scores in reading and math. But what are they? This is a matter for reflection by the researchers in the country as well as by those in charge of the system. Conclusion The implication of these The various suggestions for action that were made through out the report suggestions will not be were grouped under several main headings. A first attempt to state the level easy, but if the of operation in the administrative set-up in Vietnam where the action needs implementation is to take place has been made. This needs to be rethought and made more successful then the precise. Secondly, an attempt was made to state whether the time needed to levels of achievement implement the suggestion was short, medium or long and whether the cost should rise and there would be low, medium, or high. Again, this needs to be checked. Finally some comments were made about what might be done in a repeat study. will be more equity between schools. This has been a gargantuan study and all of those who did the hard work need to be thanked again and again. Those responsible for the future action have to first digest the results and then secondly ponder on what needs to be changed in which priority. We wish them luck. 308