The Educationl Watch was set up in 1998 by a group of institutions and individuals aiming to provide annually an independent assessment of the primary education situation in Bangladesh. The first Watch report on the overall status of primary education was published in 1999. The second Watch looks more closely at the quality aspects of primary education. Education Watch 2000 has two foci: (i) Assessment of achievement of students in terms of terminal competencies at the end of the primary cycle; and (ii) Assessment of the state of teacher education at the primary level.

Under the guidance of an Advisory Board and a Working Group, two separate research teams worked on the above. The first study (Focus I) was undertaken by a group of researchers from BRAC Research and Evaluation Division. The second one (Focus 2) was jointly done by researchers from Bangladesh Institute of Development Studies (BIDS) and Institute of Education and Research (IER) of Dhaka University. Under the umbrella of the first study, primary source materials (viz., textbooks and teachers' guides) were evaluated, two separate tools were developed, a national survey was conducted, and pedagogical practices in classrooms were observed. For the second study, documentation of current stock of trained teachers, curriculum, course content, pedagogy, duration and cost of training courses was done, and relationship between classroom practices and training was investigated.

As Education Watch 2000 consists of two independent stand-alone studies, these are published separately (Volumes II and Ill). The present publication (Volume I) draws on the two other volumes. It reports on quality of primary education as it is reflected in competencies acquired by students and the factors, including skills and performance of teachers, that affect quality.

Overall performance in cognitive competencies
Less than two percent of the students completing five years of primary education acquired all 27 competencies tested. Students did better in those competencies which depended on recall and did not involve much thinking and reasoning.

The following shows the overall subject-wise achievement of the terminal competencies (percent of students achieving all competencies).

By types of institutions, one percent of the students in government schools, 0.9 percent in private schools and 6 percent in non-formal institutions achieved the 27 competencies. The analysis was also done in terms of the average number of competencies achieved. On average, the children achieved 16.1 competencies (or 60% of all competencies) which varied between school type, residence and gender of students. The students of non-formal schools, urban students, and boys achieved higher averages than their respective counterparts in other categories.

At the national level, students showed 'poor' performance in three competencies, 'mediocre' in nine, 'satisfactory' in 12 and 'excellent' in only three. The competencies which the students found to be very difficult are 'writing English', 'problem solving in Mathematics', and 'life history of prophet Mohammed (SM) or the preachers of Own religion'. On the other hand, the competencies that the students found very easy are 'listening Bangla', 'duties as a member of society' and 'importance of good health'. The students of non-formal institutions showed 'excellent' performance in six competencies; it was three for government and only one for private school students. The students did better in items which can be described as acquisition of knowledge rather than gaining understanding and applying knowledge and skills in specific situations. School level analysis showed a wide intra-school variation in performance. Highest variation was observed in rural non-formal institutions and lowest in urban government schools.

With nearly 80 percent enrolment, 72.7 percent completing the primary cycle, and 1.6 percent attaining the competencies, less than one percent of the children leave their primary school age with the nationally determined competency. In 2001-02 budget terms, the country will be spending Tk. 14.3 billion (US$250 million) in primary and mass education sector. If the above statistics hold true, 99 percent of this resource may be going nowhere.


Reflection of terminal competencies in the source materials
Of the 53 terminal competencies, 19 were reflected 'excellently', 21 'fairly well', eight 'well', and five 'satisfactorily' in the textbooks of Class I to V. On the other hand, in the teachers' guides 36 competencies were reflected 'excellently', 10 'fairly well', four 'well', and three 'satisfactorily'. None of the competencies were found 'poorly' reflected in the source materials.

Learning achievement in Bangla and English languages
Three competencies each under Bangla and English languages- reading, writing, and listening - are included in the test instruments. These were tested with 10 question items for Bangla and 7 for English.

Bangla being the mother tongue, students did better in Bangla than in English in all three areas of assessment. All the competencies in Bangla were achieved by 36.5 percent of the students, whereas 9.4 percent achieved the same result in English. In both subject areas, students did best in listening, less well in reading and worst in writing. Urban students did significantly better than their rural counterparts in all the six competencies. The girls did significantly better in writing English and the boys in reading Bangla. No gender difference was observed in other four competencies. Statistically significant variation according to school type was found in four competencies. The students of non-formal schools did better in two competencies, viz., writing Bangla, and reading and writing English; the students of government schools did better in listening English but the students of private schools could not show such performance in any. Students of all school types did equally well in reading and listening Bangla. Seven percent of the students achieved none of the Bangla and 15.5 percent achieved none of the English competencies.

Learning achievement in Mathematics
Competencies in Mathematics include arithmetic skills and knowledge of geometric shapes. The five competencies assessed under this area included basic numbers, four basic rules, problem solving in everyday life, measurement units, and identification of geometric figures. A total of 15 question items were included in the instrument.

Performance in Mathematics was worse than that in Bangla, but slightly better than that in English language. Only 11.6 percent of the assessed students achieved all five maths competencies and 13.3 percent achieved none. Urban students did significantly better in four competencies than the students of rural areas. No area-wise difference was noticed in 'measurement units'. Boys did better than girls in all the math competencies. Performance of students varied by school type in three competencies including 'basic arithmetic', 'problem solving', and 'measurement units'. Nineteen percent of students in non-formal schools achieved all five Mathematics competencies compared to 10.6 percent in the government schools.
 

Learning achievement in Environmental Studies
Environmental Studies include Social Studies and General Science. Six competencies under social studies and nine under general science were tested. The General Science section includes health and hygiene, and science and technology. A total of 31 question items were put in the instrument under these areas.

Less than a fifth of the students achieved all competencies in Environmental Studies. The students of non-formal centres performed better in both the subjects compared to other students. In Social Studies, boys outperformed girls in all the six competencies. On the other hand, in Science, boys did better in five competencies and no gender difference was noticed in other four competencies. The students of urban areas did significantly better than those of rural areas in all the competencies of Environmental Studies. The achievement level varied significantly by type of school for most of the competencies. In Social Studies, students in non-formal institutions showed best performance in three competencies, the students of government schools did best in one and both of the sub-system did equally well (or equally worse!) in one competency. On the other hand, in Science, the students of non-formal schools were ahead in six competencies, and the students of government schools were so in one competency.
 

Learning achievement in Religious Studies
Only one competency was considered under this area. This was 'life history of prophet Mohammed (SM) or the preachers of own religion'. Only 26.7 percent of the students achieved this competency; 30.3 percent for boys and 23.3 percent for girls. On average, 42.5 percent of the students of rural areas and 22.7 percent of the students of urban areas achieved this competency. Analysis by school type showed that 28.3 percent in government, 18.4 percent in private and 29 percent in non-formal schools achieved this competency. Gender difference favouring boys prevailed in all three types of institutions. Students of urban areas did better than their rural counterparts in two sub-systems, viz., government and private urban government schools.
 

Classroom observations
The case study of pedagogical practices in classrooms reveal that poor physical facilities, inadequate teaching materials (including textbooks and copybooks), memory-based teaching style and lack of remedial measures in the classroom for slow learners are the reasons for poor performance in the primary schools. Such inadequacies are more prevalent in private schools and least in non-formal schools. An outcome-oriented effort by teachers focusing on acquisition of competencies by students is in general absent in classroom transactions and pupil assessment.
 

Performance in non-cognitive competencies
Assessment of non-cognitive competencies of students relate to attitudes and values was undertaken only in government schools. The methodology followed was to have teachers rate their own students on the basis of a structured questionnaire. The degree of subjectivity of this method is an weakness of this assessment. Sixteen percent of the students satisfactorily achieved all the non-cognitive competencies-15.8 percent girls and 16.1 percent for boys; and 16.3 percent for rural and 14.6 percent for urban students. On average, the students achieved 68 percent of the competencies; this was 67 percent for rural and 73 percent for urban students. The average achievement was 67 percent for girls and 69 percent for boys.
 

Background characteristics and learning achievement
How far the socio-economic and educational characteristics of the students affect their learning achievement was examined. A negative relationship between age and learning achievement was observed. On the other hand, performance of students increased with the increase in parental education, self-perceived food security status and access to mass media.

Access to private tutor contributed in learning achievement. Students whose guardians discussed academic matters with their teachers, whose parents provided tutoring at home and whose guardians participated in school meeting achieved more competencies compared to those who did not do such things. However, unexpectedly, students' participation in co-curricular activities had a negative relationship with their learning achievement.

Schools with 40 or fewer students per teacher showed better performance than those with higher student-teacher ratio. The performance of students increased with the increase in the educational qualification of the teachers, but it decreased with the increase in the length of their service. Schools with more teachers having professional training did better than those with a lesser proportion of the teachers having such training. Distance between school and the local education authority had negative relationship with the achievement of the students. The performance of students increased with the increase of visits made by such authority. Multivariate regression analyses confirmed relationship of various socia-economic and educational variables on competency-based learning achievement of students. Except for a very few cases, statistically significant correlation of these variables were noticed on the achievement of students in all types of institutions. This indicates that their learning achievement is influenced to a considerable extent by their family background, the support they receive from their families and the school environment.
 

Teacher education
Two types of teacher education are commonly found for primary school teachers in Bangladesh. The most common is the 10-month Certificate-in-Education (C-in-Ed) course offered by Primary Teacher Training Institutes (PTI). The other is the foundation training offered by different non-governmental organisations (NGOs) for non-formal school teachers, the duration of which is usually 15 days. An overwhelming majority of the teachers in the government and the non-formal schools received basic professional training; however, very few teachers of the English medium schools got such training. There are 54 PTIs and training capacity of these is increasing. Learning outcomes defined in the PTI curricula vary by subject. Outcomes are largely descriptive and theoretical; scope for activity-based learning being limited. Lecture method is predominant in the PTI classrooms and practical skills are not emphasized. Total cost for producing a C-in-Ed was estimated at approximately Tk. 72,000. Majority of the educators in PTIs had professional training like M. Ed or B. Ed.; but, none had any practical experience of primary school teaching or even received training on primary education. Physical facility in the PTIs is largely unsatisfactory. PTI instructors considered themselves overburdened and class sizes were large. On average, an instructor had to take 17 classes per week. Classroom observations linking teachers' training and pedagogy practices revealed that interactive classroom practices were more prevalent in government schools than in the private ones. However, the non-formal classes were more interactive than the classes of above two sub-systems. The degree of interactivity apparently made some difference in student achievement as revealed in the competency scores, but overall performance of students remained low in all categories of institutions. There is a need for a systematic review of the relevance of PTI training and develop workable models for teacher's professional preparation and their effective performance in the classroom.

The findings about overall low pupil achievement and consequent wastage of scarce resources; the differences in achievement between urban and rural children, between boys and girls, among different types of schools and among domains of learning; teachers' skills and performance; and school-related and social factors affecting children's learning do point to policy issues which must be addressed to improve the quality of teaching and learning in primary education.

A focus on quality: The deplorably low achievement in terms of competencies acquired by students in primary education clearly points to the need for the highest priority to quality improvement in primary education development efforts. The progress made in enrolment and closing of gender gap in enrolment are liable to be totally negated if rapid and substantial improvement cannot be made in quality reflected in learning outcomes of students. The priority to quality needs to be focused on specific ways of overcoming obstacles to quality improvement, many of which has been identified in this study. Current primary education development activities supported by external assistance (such as the multi-faceted Primary Education Development Project and various quality-related projects) appear not to have much of an impact on quality. The plethora of projects and to what extent they address the root causes need to be critically reviewed as a starting point for developing a comprehensive and effective quality improvement strategy. Important elements of such a strategy are indicated by the findings of the study as noted below.

Urban-rural difference: With the overwhelming majority of primary school age children living in rural areas, educational policies and programmes need to give special attention to specific conditions in rural areas causing lower achievement of rural children The focus of attention should include per capita resource allocation, distribution of school facilities, direct and indirect costs of schooling to poor households, enforcing performance standards of schools and teachers through stronger supervision and accountability to the community, and introducing an objective system of monitoring and assessing learning achievement of children.

Gender gap: The trend towards narrowing the gender gap in participation in education and learning achievement has to be maintained. It has to be kept in view that the overall achievement remain very low for all children and the effort to improve total achievement has to continue along with combating traditional attitudes about gender roles and expectations reflected in classroom practices, teachers' behaviour and attitudes and the content of learning material.

The formal and non-formal education divide: The policy question is how the advantages of the non-formal schools can be incorporated into formal schools, even when the differences in the circumstances of the two are recognised. There are no simple answers, but establishing a stronger performance accountability of teachers and schools specifically focusing on learning outcome, re-examining the effectiveness of teacher training and supervision, ensuring that learning materials and textbooks are available on time to all children, meaningful involvement of parents in school and better communication between parents and teachers of their children have to be important items on any quality improvement agenda. Another important move would be to promote local-level comprehensive planning (at the village, union and upazila level) for educational services of acceptable quality for all children, involving in the process all stakeholders including NGOs, government and government-assisted institutions, community leaders and the local government structure. The aim would be to identify and implement essential quality improvement measures in all primary education institutions in the area.

Achievement in different domains: It is essential that the curriculum, definition of competencies and learning objectives, preparation of learning materials and textbooks, teacher training (especially in respect of teaching language and Mathematics to young children) and supervision, and assessment of learning outcome accord a high priority to equipping children with the basic tools of literacy and numeracy. It is necessary to reconsider the list of competencies and the relative emphases on them in order to ensure that necessary time and effort are directed to acquiring the core literacy and numeracy competencies by students.

Teachers' skills and performance: Mere tinkering with the curriculum of the existing model of formal teacher training and its expansion will not help improve the quality of primary education. The premises and assumptions of the current programme regarding training objectives, training methods and conditions for use of training in classroom in other words, why training is not making the expected contribution to better learning outcome - have to be probed rigorously, followed by a fundamental rethinking about effective teacher training and creating the conditions for use of the training in classroom. The non-formal teacher training method is not necessarily the model to be replicated for all schools, if only because the better student performance of non-formal primary education still falls short of an acceptable national norm for achievement in primary education competencies. However, a serious rethinking of teacher training will have to take into account many relevant lessons from the approach followed by NGOs in training their teachers.

Family-related factors: The policy implication, from the point of view of national and state obligation to provide quality basic education to all children, is that the school programme needs to be designed and other ancillary measures taken to compensate for serious family-related deficiencies impeding children's learning. NGOs, targeting specifically disadvantaged groups, have applied this approach with positive results. Among the measures that can be contemplated are: providing learning materials to students and eliminating direct and indirect, official and informal, costs to poor parents; eliminating the need for paid private tutoring; flexibility in school programme and routine to suit seasonal agricultural workload for families and situations of working children and so on. A pertinent policy issue is whether the substantial government spending on 'food-for-education' and the stipendprogramme, which consume over a third of the recurring budget of primary education, would not be better spent directly on improving teaching-learning condition, such as providing learning materials to poor children and supporting volunteer tutors for children who need extra help in their lessons.

School-related factors: The important policy message here is the need for establishing a system of accountability of school and teachers' performance to parents and the community. The community, in turn, needs to be encouraged to be involved in creating the condition for the school and the teachers to function properly.

Decentralisation of educational management: Autonomous district education authorities should become the pivotal entity for overall planning and management of primary and secondary education. The district authorities should manage educational resources provided by the government and other resources derived from other sources and support community and school-based plans and programmes for quality primary education for all. The decentralization process and district-based management should be initiated on a trial basis in a few districts to help capacity-building and to learn how the decentralised system can be protected from corruption and politicization, two major problems in the education system, which have crippled various reform efforts.

Resources: More than doubling of the share of GNP to education in the next five years to 5 percent from the present level of over 2 percent will be an essential measure for quality improvement in primary education. Even this increase, if achieved, will bring Bangladesh to the average level among developing countries' spending for education in GNP share. However, the distributional aspect of this increased allocation will be important. This must be diverted to activities that enhance quality of education.

Short-term measures

• Undertake an independent review of the quality improvement aspects of the major primary education development projects including various separate quality improvement projects in order to develop a comprehensive and coordinated quality improvement strategy and programme, addressing root causes of poor quality.
   
• Begin increasing public budget for primary education (within the framework of a goal of doubling the share of education expenditure in GNP), targeting the new resources to quality improvement measures.
   
• Revisit the list of the 53 competencies, retain those which are competencies in real sense of the term, and redefine these, as necessary, in terms of measurable pedagogic outcomes.
   
• For the new list of competencies, design fresh orientation programmes for teachers and their supervisors in terms of both teaching-learning and assessment processes. Devise and introduce measurement indicators and standards for the competencies.
   
• Raise awareness among teachers and their supervisors about the existing state of the quality of primary education in terms of the attainment of the terminal competencies, and the need to improve it.
   
• Consider deferring the introduction of English until Class IV.
   
• Introduce new materials and books to engage students in creative and problem solving exercises in various subject areas. Also train teachers to handle this new demand on them.
   
• Re-examine existing and required capacities in the supervisory system from a perspective of improving learning outcome of students; and monitor the performance of teachers, their supervisors and schools from this perspective; consider instituting incentive and reward systems for better performing schools.
   
• Supply textbooks on time and Khata (copybook) and pencils free of cost to students, particularly those who are disadvantaged. If resource is a constraint (which we believe is a matter of priorities and choice rather than scarcity), divert resources from the Food-for-Education and stipend programmes.
   
• Support and encourage more interaction between the formal and non-formal systems and institutions including those for teacher training in order to build on and make best use of the strengths of each other.

Medium and long-term measures

• Hold implementation of the proposed eight-year primary cycle until the existing system is more responsive to the need of the students; priority should be to improve quality and performance standards in existing institutions rather than engage in a major institutional restructuring venture with consequent strain on resources and management capacity.
   
• Create autonomous district education authorities for overall planning, management and accountability of primary education; begin with a trial in a number of districts to ensure success and learn lessons.
   
• Improve the reach and outcome performance of secondary school system to create aspirations among primary students and their parents.
   
• Double the GNP share of public resource allocation for education to five percent in the next five to seven years. Distribute this increased resource to quality enhancing measures.