ETHIOPIAN SECOND NATIONAL

LEARNING ASSESSMENT OF
GRADE 4 STUDENTS
120

100

80

64.5
60
51.7
48.5

39.7 38.7
40

20

Mean
0 ±SD
READING MATH COMPOSITE ±1.96*SD
SCIENCE ENGLISH

National Organization for Examinations

Addis Ababa
December 2004

Funded under USAID Contract No. 663-C-00-02-00349-00

BASIC EDUCATION STRATEGIC OBJECTIVE

 Technical Working Group

Erkihun Desta Chairman NOE

Endale Shewangizaw Secretary >>
Getahun Worqineh Member >>
Hailu Chane >> >>
Kasu Zewdie >> >>
Kefeleghn Tsegie >> >>
Yohannes Afework >> >>
Zewdu Gebrekidan >> >>
Teshome Nekatibebe Technical Advisor AED/BESO II

A national assessment may be defined as an exercise

designed to describe the level of achievement, not of
individual students, but of a whole education system or a
clearly defined part of one (such as grade 4 pupils or 11 –
year- olds).(Kellaghan and Greany, 2004)

National Organization for Examinations i

 Table of Contents
List of Tables iv

Acronyms vi

Acknowledgments vii

CHAPTER ONE: INTRODUCTION 1

1.1 Ethiopia 1
1.2 Purpose of the Study 3
1.3 Key Research Questions 4
1.4 Specific Objectives of the Study 4
1.5 Significance of the Study 5
1.6 Limitations of the Study 6
1.7 Definition of Terms 7

CHAPTER 2: A REVIEW OF RELATED LITERATURE 9

2.1 Introduction 9
2.2 The Concept of National Learning Assessment 9
2.3 Purposes and Functions of National Learning Assessment 10
2.4 Methodological and Technical Considerations in National Learning Assessment 11
2.5 Experiences of Other Countries Regarding National Learning Assessment 12
2.6 Background and Community Factors that Influence Student Achievement 15
2.7 School Based Factors that Influence Student Achievement 16
2.8 Historical Overview of Learning Assessments in Ethiopia 18

CHAPTER THREE: FRAME OF ANALYSIS AND METHODOLOGY 20

3.1 Frame of Analysis 20
3.2 Design 21
3.3 Sampling Procedures 21
3.4 Instruments of Data Collection 25
3.5 Validity and Reliability of Instruments 26
3.6. Data Collection and Administration of Instruments 27
3.7 Methods of Data Analysis 28

National Organization for Examinations ii

CHAPTER 4: FINDINGS OF THE STUDY 29
4.1 Findings on Grade 4 Achievement Tests 29
4.2 Overall Achievement of the Students by Subject 31
4.3 Grade 4 Pupils’ Background and Academic Achievement 46
4.4 Grade 4 Pupils’ Attitudes in Socially Relevant Issues 51
4.5 Relationships between Grade 4 Pupils’ Achievement and Factors 66
4.6 Comparative Analysis between EBNLA and ESNLA 72
4.7 Analysis of Qualitative Data 77

CHAPTER 5: SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 89

5.1 Summary 89
5.2 Recommendations 95

REFERENCES 97

National Organization for Examinations iii

 List of Tables
Table 1. Description of variables .................................................................................................................................20
Table 2. Sample schools planned and visited by region............................................................................................23
Table 3. Sample students, teachers and directors by region ....................................................................................24
Table 4. Participants of focus group discussions by region......................................................................................24
Table 5. Number of participants involved in field work by route ...........................................................................28
Table 6. Summary of the descriptive statistics for all subjects at national level ...................................................29
Table 7. T-test for equality of means by gender at national level............................................................................30
Table 8. T-test for equality of means by location at national level..........................................................................30
Table 9. Summary of the descriptive statistics for reading comprehension by region .........................................31
Table 10. One- way analysis of variance on reading comprehension by region ....................................................31
Table 11. Homogenous subset groups for reading comprehension.........................................................................32
Table 12. T-test for equality of means for reading comprehension by gender within regions ............................32
Table 13. T - test for equality of means for reading comprehension by location within regions.........................33
Table 14. Summary of the descriptive statistics for English by region..................................................................33
Table 15. One way analysis of variance of English across regions .........................................................................34
Table 16. Homogeneous subset groups of English by region ...................................................................................34
Table 17. T-test for equality of means for English by gender within regions ........................................................35
Table 18. T-test for equality of means in English by location within regions ........................................................36
Table 19. Summary of the descriptive statistics for mathematics by region..........................................................36
Table 20. One way analysis of variance on mathematics by region ........................................................................37
Table 21. Homogenous subset groups of mathematics by region............................................................................37
Table 22. T-test of mean difference for mathematics by gender within regions ..................................................38
Table 23. T-test of mean difference for mathematics by location within regions.................................................39
Table 24. Summary of the descriptive statistics for environmental science by region .........................................39
Table 25. One way analysis of variance on environmental science by region........................................................40
Table 26. Homogenous subset groups of environmental science by region............................................................40
Table 27. T -test of mean difference for environmental science by gender within regions ..................................41
Table 28. T-test for mean difference in environmental science by location within regions .................................42
Table 29. Summary of the descriptive statistics for composite mean by region....................................................42
Table 30. One way analysis of variance on composite mean by region ..................................................................43
Table 31. Homogenous subset groups of mean composite score by region ............................................................43
Table 32. T -test of mean difference for composite mean by gender within regions.............................................44
Table 33. T-test of mean difference for composite mean by location within regions...........................................45
Table 34. Time needed to reach school and academic achievement........................................................................46
Table 35. ANOVA on time required to reach school ...............................................................................................46
Table 36. Watching non-educational television and student achievement............................................................47
Table 37. Listening to the radio (other than educational radio) per week .............................................................47
Table 38. Absenteeism and achievement ...................................................................................................................48
Table 39. Frequency at which attendance is taken....................................................................................................48
Table 40. Daily meals and achievement......................................................................................................................48
Table 41. Interest in learning English.........................................................................................................................49
Table 42. Interest in learning mathematics................................................................................................................49
Table 43. Interest in learning environmental science ...............................................................................................49
Table 44. English textbook distribution.....................................................................................................................50
Table 45. Mathematics textbook distribution ............................................................................................................50
Table 46. Environmental science textbook distribution ...........................................................................................50
Table 47. Attitude towards health care......................................................................................................................52
Table 48. Attitude towards environmental protection..............................................................................................52
Table 49. Attitude towards civics and ethics.............................................................................................................53
Table 50. Attitudes towards harmful cultural practices..........................................................................................53
Table 51. Attitude towards education........................................................................................................................53
Table 52. We can prevent the spread of diseases through personal and environmental hygiene .......................54
Table 53. I would like my parents to give birth over and over again so that I can have many brothers
and sisters. ............................................................................................................................................................54
Table 54. As far as we are aware of the modes of transmission of HIV/AIDS and take care, we will not
be infected by the virus. ......................................................................................................................................55
Table 55. Blood tests cannot help identify HIV/AIDS infection ............................................................................55
Table 56. Helping HIV/AIDS infected people is a social obligation ........................................................................55
Table 57. Learning with those with HIV/AIDS causes no problem ........................................................................56

National Organization for Examinations iv

Table 58. Eating meat regularly is a proper dietary practice. .................................................................................56
Table 59. Drinking river water can expose one to water borne diseases................................................................56
Table 60. Visiting historical places is wasting both money and time. .....................................................................57
Table 61. Gardening gives me pleasure. .....................................................................................................................57
Table 62. One should plant trees to prevent erosion.................................................................................................58
Table 63. Keeping the environment clean is everyone's duty. .................................................................................58
Table 64. It is possible to prevent the consequences of drought. .............................................................................58
Table 65. It is necessary to burn forests to drive away wild animals that destroy crops. ....................................59
Table 66. Women are equal to men in knowledge and thinking..............................................................................59
Table 67. It is appropriate to return property found lost either to its owner or to the school
administration. .....................................................................................................................................................60
Table 68. I believe that group work is important for academic success. ................................................................60
Table 69. Gathering information about our surrounding is necessary to have knowledge about it. ..................61
Table 70. Solving differences through dialogue is good culture. .............................................................................61
Table 71. Boys and girls should equally participate in household chores after school. ........................................62
Table 72. Early marriage is a practice that should be abolished.............................................................................62
Table 73. Female circumcision is a harmful traditional practice that should be abolished................................63
Table 74. Absenteeism and coming late to school are signs of laziness...................................................................63
Table 75. Student participation in teachers’ evaluation is appropriate. ...............................................................64
Table 76. Going to school and staying at home are the same to me. .......................................................................64
Table 77. Learning is a means to earn a higher status.............................................................................................65
Table 78. Because I am a student, I am more knowledgeable and mature than those not in school. .................65
Table 79 Spearman Rho Correlation with Grade 4 Composite Score....................................................................66
Table 80. School-level multiple regression results on the Grade 4 composite achievement score details
on reduced models ...............................................................................................................................................68
Table 81 School level multiple regression results on the composite score model summary.................................69
Table 82 Contributions of each block to R-square of completely reduced model .................................................70
Table 83. G4 EBNLA national baseline mean score (by the end of 1992 E.C.) ..................................................72
Table 84. G4 ESNLA National mean score (by the end of 1996 E.C.) ...................................................................72
Table 85. Learning Progress: Comparison between the two national learning assessments ...............................73
Table 86 Reading: average change in percent correct scores from Baseline to SNLA........................................74
Table 87. English: average change in percent correct scores from Baseline to ESNLA ....................................75
Table 88. Mathematics: average change in percent correct scores from Baseline to SNLA..............................75
Table 89. Environmental science: average change in percent correct scores from Baseline to SNLA.............76
Table 90. Average difference in percent correct scores across Grade 4 achievement tests..................................76
Table 91. Student learning as viewed by teachers, students and parents by region. ...........................................78
Table 92. School supplies and school facilities...........................................................................................................80
Table 93. Teacher behavior, availability and competence .......................................................................................81
Table 94. Student behaviors influencing academic achievement ............................................................................83
Table 95 The influence of the curriculum on students achievement.......................................................................85
Table 96. The influence of parental support on school achievement in Ethiopia..................................................87
Table 97. Summary of Grade 4 achievement scores by region ................................................................................90

National Organization for Examinations v

Acronyms
BESO I: Basic Education System Overhaul
BESO II: Basic Education Strategic Objective
ENBA: Ethiopian National Baseline Assessment
ERGESE: Evaluative Research of the General Education System in
Ethiopia
ESDP: Education Sector Development Program
ESR: Education Sector Review
ETP: Education and Training Policy
FDRE: Federal Democratic Republic of Ethiopia
ICDR: Institute of Curriculum Development and Research
EMIS Educational Management Information Service
IEA: International Association for the Evaluation of Educational
Achievement
IIEP: International Institute for Education Planning
MLA Monitoring Learning Achievement
MOE: Ministry of Education
NAC: National Advisory Council
NCES National Center for Education Statistics
NAEP: National Assessment of Educational Progress
NLA: National Learning Assessment
ESNLA: Ethiopia Second National Learning Assessment
NOE: National Organization for Examinations
PISA: Program for the International Student Assessment
EFA Education for All
REB: Regional Education Bureaus
SACMEQ: South African Consortium for Monitoring Educational
Quality
SNNPR: Southern Nations, Nationalities and People’s Region
SPC: School Pedagogical Center
TWG: Technical Working Group
USAID: United States Agency for International Development

National Organization for Examinations vi

Acknowledgments

A number of people and institutions have contributed to the successful completion of the
Second Ethiopian National Learning Assessment. AED/BESO II provided the required
financial and technical assistance for carrying out the study. Dr. Elizabeth Randolph and
Mr. Benjamin Piper have provided an immense technical support as external consultants to
the project. Dr. Randolph has displayed a great deal of dedication in the design of the
study, analysis of data, and training of personnel and provision of resource materials
required for the assessment. Mr. Piper from Harvard University has provided a technical
support in the analysis of statistical data. Ato Mulu Nega from the Institute of Educational
Research (Addis Ababa University) has contributed to the study on several occasions as a
local consultant. Ato Gizaw Tasissa of the Department of Curriculum and Instruction
(Addis Ababa University) has assisted in the analysis of qualitative data upon request.
Duel acknowledgment also goes to Ato Bedru Kedir and Ato Fiseha W/Michael who
contributed to writing the report and preparing guidelines for field work respectively.
Members of the National Advisory Council of the Ethiopian Second National Learning
Assessment devoted their time to advise and support the Technical Working Group. In
addition to his duties as a chairman of the National Advisory Council of the Second
Ethiopian National Learning Assessment, His Excellency Ato Dereje Terefe, Vice
Minister of Education, has constructively responded on different occasions to our requests
to provide comments on draft reports. There is no doubt that this study would not have
been completed without the full support of the data collectors from NOE and regions, in
particular Zone and Woreda Education department heads, school directors, teachers,
parents and students. Finally, USAID/Ethiopia deserves special thanks for making the
whole study possible in cooperation with the Ministry of Education of Ethiopia.

National Organization for Examinations vii

The Ethiopian Second National Learning Assessment Grade 4 Report

CHAPTER ONE: INTRODUCTION

1.1. Ethiopia
Ethiopia, covering an area of 1,133,380 sq. km (Keller, 2004), is located in the North Eastern
part of Africa. The country is bounded in the North by Eritrea and Djibouti, in the East and
South East by Somalia, in the South West by Kenya, and in the West and North by Sudan.
Located in the sub tropical zone, Ethiopia possesses varying topography mainly plateaus,
highland chains at the center, and low lands at the coastal areas. The climate varies from very
hot to very cold temperature and from arid to high rainfall areas following the topography.
The principal rainy season of the country is between June and September, whereas dry season
is from October to May with short lived and weak rain in February and March.
Wide variations in land structure, soil type and climate are the main determinants for the
diversified agricultural products, vegetation, wild life and mineral resources in Ethiopia. The
nature of the plateaus and other factors like deforestation, over grazing, poor land
management and the cultivation system in the high lands accelerate the degree of erosion of
the top soil causing low productivity.
1.1.1 Population and Culture
With an estimated population of 68.6 million (World Bank, 2003), Ethiopia is the third
populous country in Africa next to Nigeria and Egypt. Different nationalities and ethnic
groups having diversified cultures and more than 80 spoken languages inhabit the country. Of
these languages, about 22 are used as media of instruction in schools. Amharic is the Federal
Government’s working language. There are two dominant religions, Christianity and Islam,
which have a long history and many followers in the country.
Since August 1995, Ethiopia has become a Federal Democratic Republic constitutionally. It
comprises 9 regional states and three city administrations. 4 of the regions are not only large
in size but also have better infrastructure. Two of the regions are city administrations while
the rest are considered as emerging regions.
1.1.2 Economy
Traditional agricultural and pastoralist practices employ about 80% of the population. The
most important food crops grown primarily for local consumption are cereal grains. Coffee,
cotton, oil seeds, pulses, fruits, vegetables as well as leather and skin are some of the
commercial products used for internal processing as well.
In addition to the backwardness and the low productivity of the agricultural sector, the
mining and industrial sectors are also at a lower level of development. But some promising
potentials and developments are being observed around the production of hydro-electricity,
airlines and other communication sectors. The main financial sources of the government are
some local revenues, loans and donor funds. The Ethiopian currency is called Birr; about 8.60
Birr is equivalent to one US dollar.

National Organization for Examinations 1

The Ethiopian Second National Learning Assessment Grade 4 Report

1.1.3 Education System

Ethiopia is one of the least developed countries in the world. Social indicators (health and
education) have remained at a lower level compared to other Sub-Saharan countries. The
country is hard hit by recurrent drought. About 44% of the total population is living below
poverty line (MOFED, 2002). Infant mortality rate is relatively high. Life expectancy is about
42.1 years (World Bank, 2002) with a declining trend due to HIV/AIDS.
The government of Ethiopia (FDRE) has set up various policies and strategies that could
alleviate poverty and guide the overall development of the society. Special focus in the
development policies has been given to rural and agricultural development. In general, there
are 4 specific areas of emphasis in the development strategy of the government: agriculture
led industrialization, justice and civil service reform, government decentralization and
empowerment, and capacity building in public and private sectors.
The strategy and program framework of capacity building, in particular, focuses on education
in order to increase the production of trained his manpower and upGrade the competence of
the existing labor force in Ethiopia. Special attention in the development of the education and
training system has been given to ensure access to primary education for all citizens, and
simultaneously improve the quality and relevance of the sector for the expected socio-
economic development.
To this end, the government of Ethiopia issued a new Education and Training Policy (ETP) in
1994. Overall, ETP was a response to the observed crises in the Ethiopian education sector,
and it aimed at ensuring equity, accessibility, quality and relevance of the education. The
following strategies were formulated to realize the goals of ETP:

• Restructuring of the education system (new educational structure)

• Administrative change through decentralized management
• Curriculum change – development of new curriculum
• Use of local languages as media of instruction
• Changes in teacher training and professional development opportunities (new
teaching career structure)
• New approaches to classroom and national examinations
• Change in educational finance (introducing cost sharing in upper secondary and
tertiary education)
Particular priority areas of the ETP are curriculum change, decentralization and teacher
education.
The Ministry of Education has formulated a 20-year development strategy to implement the
ETP. The Education Sector Development Program I (ESDPI), implemented from 1997-2002,
was the first phase of the 20-year program followed by the ESDP II (2002-2005).
The overall objectives of the ESDP II regarding primary education focus on the inculcation of
good citizenship and actualization of the goal of universal primary education by the year
2015.
The program has also set directions regarding assessment and quality improvement. These
include:

National Organization for Examinations 2

The Ethiopian Second National Learning Assessment Grade 4 Report

• conducting assessment studies on selected languages of nations and

nationalities that are used as media of instruction,
• developing learning competency guidelines to assess the profile of
education standards at each level for each subject,
• producing manuals for promoting continuous assessment, and
• carrying out national educational assessment on Grades 4 and 8 to assess
the skill development level of students in key subject areas.
The implementation of the ETP has been underway for the last 12 years. One major measure
taken is the decentralization of educational management following the federal structure of
administration. According to this arrangement, all educational institutions, except the
universities, are under the jurisdiction of the regional governments.
The previous twelve years of general education structure (6+2+4) has been changed into ten
years. The new structure comprises eight years of primary education subdivided into two
cycles of basic education (1-4) and general education (5-8); a general secondary education (9-
10) with a terminal (school leaving) examination and a preparatory education (11-12). Those
who could not join the preparatory program go to technical and vocational training that takes
one to three years to complete. Those who successfully complete the preparatory program go
to higher education of undergraduate studies (3-6 years).

The implementation of ETP has brought about encouraging results in terms of access to
primary education and schooling in the country. Some of the milestones include the
following.
• The Gross Enrolment Rate for primary education (1-8) increased from 45.8% to
64.4% between 1998/99 and 2002/03. The increase has been 18.7% for boys, 18.5%
for girls and 18.6% for the total.
• The Net Enrolment Rate for primary education has increased from 39.6% in 1998/99
to 54% in 2002/03. The gender gap in net enrolment rate has decreased from 15.1%
in 1998/99 to 13.4% in 2002/03.
• The proportion of qualified teachers has reached 30.9% in 2002/03 for upper primary
(5-8).
• The pupil teacher ratio for primary education has risen from 51:1 to 64:1 between
1998/99 and 2002/03, which is a 25.5% increase. This indicates a rapid growth of
enrolment rate at this level.
• The education share of government budget expenditure rose from 15.6% in 1997/98
to 18.8% in 2002/03.

1.2 Purpose of the Study

Education is moving from being a privilege for the few to becoming the right for all.
However, this quantitative expansion has brought about serious challenges to its quality.
Quality does not mean only what goes into schools, but also what goes in the mental and
physical changes of children. It is important to develop the knowledge, skills, attitudes and
habits of pupils in addition to giving emphasis to input factors.
Some developing countries have tried to assess and measure student achievement and
improve their educational systems. Nevertheless, most countries still apply public

National Organization for Examinations 3

The Ethiopian Second National Learning Assessment Grade 4 Report

examinations for certification, selection and promotion. Improving student learning has
remained one of the most desired goals of educational processes.
In Ethiopia, quality assurance has been an important part of the reform process. To this end,
the Ethiopian National Baseline Learning Assessment for primary education was carried out
in 1999/2000. Currently, the Second National Learning Assessment is taking place in the
country.
The main purpose of conducting the Second National Learning Assessment is to provide
information about learning attainments by students and the factors that determine those
attainments in the Ethiopian primary education so that attention is paid to the improvement of
the system as a whole.

1.3 Key Research Questions

1. To what extent did Grade 4 students achieve the stated curriculum in key subjects
and to what degree does their performance vary across regions, gender, location and
medium of instruction?
2. What do the students’ background information and interest look like in relation to
their overall achievement?
3. What do Grade 4 students’ general attitudes, beliefs and preferences look like in
relation to pro-social behavior and socially relevant issues at national and regional
levels?
4. What are the factors that influence students’ achievement in the primary schools of
Ethiopia?
5. Is there any progress from the baseline regarding pupils’ learning achievement?
6. What are the qualitative assessments and judgments of different groups (directors,
teachers, students and the community) on the efficiency, problems and solutions
concerning student learning in schools?
7. What are the possible implications of the findings of the study for improving student
performance and school quality in Ethiopia?

1.4 Specific Objectives of the Study

The Ethiopian Second National Learning Assessment of students in Grade 4 has the
following specific objectives:

1. Analyze the national student learning achievement results in Grade 4 mathematics,

English, environmental science and basic reading comprehension and student
attitudes in socially relevant issues
2. Analyze Grade 4 students’ achievement in mathematics, English, environmental
science and basic reading comprehension results by gender, location, and region
3. Find out Grade 4 students’ general attitude towards socially relevant issues at
national and regional level
4. Determine the relationship between Grade 4 students’ background variables and their
overall achievement in the given subjects
5. Determine the level of Grade 4 students’ learning progress from baseline by
comparing scores obtained in the first and second national learning assessment

National Organization for Examinations 4

The Ethiopian Second National Learning Assessment Grade 4 Report

6. Provide baseline data for Grade 4 student attitudes on socially relevant issues
7. Explain the factors that influence Grade 4 students’ achievement
8. Assess the opinions and judgments of different groups: directors, teachers, students
and the community on the efficiency and problems of learning in schools
9. Summarize the implications of the findings of the Second National Learning
Assessment for the improvement of school quality and effectiveness in Ethiopia

1.5 Significance of the Study

Student learning assessment involves a systematic process of collecting relevant, valid and
timely information about the outcomes of schooling so that decisions are made about the
learning and development of students, curriculum, educational programs and educational
policy. Student learning assessment provides the necessary feedback and objective evidence
required to maximize the outcomes of educational efforts. Such assessments summarize what
learners know, understand, and can do in relation to some or all of the learning goals
determined in the curricula.
Over the last decade, substantial attempts have been made to expand primary education, and
improve access, equity and efficiency in Ethiopia. Now the emphasis has shifted towards
improving quality in all areas and in particular towards student learning achievement. This
national learning assessment, therefore, provides an indication or feedback of where students’
achievement in the country stands in relation to the stated profiles of the curriculum.
A student learning assessment can provide baseline information from which progress can be
measured during and at the end of a key stage in education. Since it focuses on actual
learning, it enables one to find out the extent to which an educational system is effective as a
whole. If it is properly integrated into the system of education, student learning assessment
can help actors and stakeholders to focus their collective attention, examine their
assumptions, and create a shared academic culture dedicated to assuring and improving the
quality of education. The Ethiopian Second National Learning Assessment is a nationwide
program and a continuation of the Ethiopian National Baseline Learning Assessment. The
first national assessment has provided a benchmark from which improvement can be
measured. In this respect, the Second National Learning Assessment will serve as a key tool
for monitoring changes or improvements since the time the Ethiopian National Baseline
Learning Assessment was conducted. Since it includes new subjects, it also creates baseline
data.
The Education and Training Policy of the Federal Government decentralizes education in the
sense that regions plan and administer primary education under the guidelines and standards
set by the Ministry of Education. Moreover, the policy states that primary school children
should learn in their mother tongues. Regions implement the Education and Training Policy
by taking into consideration their own specific conditions. This implies that some of the
features of these regions affect the practice of primary education in relation to curriculum
development, material preparation, teacher education, school management, teacher practices,
school-community relations, language of instruction etc., and the extent to which students
learn from their schooling. The Ethiopian Second National Learning Assessment contributes
to monitor how expected national standards have been implemented and if each of the regions
has developed realistic mechanisms to convert national guidelines into local tools for school
development.
Information on the relationship between student learning outcomes and school inputs
provides an immense potential to policy makers to identify, allocate and manage the
resources of education to improve quality. The Ethiopian Second National Learning

National Organization for Examinations 5

The Ethiopian Second National Learning Assessment Grade 4 Report

Assessment provides such information along side the achievement results so that the most
influential determinants of learning are properly known and managed.
Ethiopia expends a considerable amount of its public finance on education. In order for the
education sector to justify this expenditure and retain support, both the government and the
public require that the money expended produce the required skills. The Ethiopian Second
National Learning Assessment provides access to such relevant information and this,
hopefully, promotes accountability in the system.

1.6. Limitations of the Study

Tests were more of an objective type and this did not allow the measurement of high order
thinking and skills students acquired in schools. The curriculum intends several proficiencies
other than the ability to recognize and know the proper answers from the given alternatives.
Future studies may consider the measurement of various domains of learning as much as
possible.
Sample schools were not visited before the field work started. The consequence had been that
current changes in the status of the schools demanded the replacement of few of the sampled
schools. Data obtained from EMIS showed that some schools were full primary schools
having Grades 1-8, but it was found out during the field study that these schools had only
Grades1-4.
Gambella region was not included in the study due the problem that prevailed in the area. The
inclusion of such an emerging region would have provided further insight into how schools
function in the relatively less developed regions of the nation.

National Organization for Examinations 6

The Ethiopian Second National Learning Assessment Grade 4 Report

1.7. Definition of Terms

National learning assessment: Sometimes is known as system assessments, assessments of
learning outcomes or learning assessments. This activity may be defined as an exercise
designed to describe the level of achievements, not of individual students, but a whole of the
education system or a clearly defined part of it (Kellaghan and Greany, 2004). In the
Ethiopian Second National Learning Assessment, the main exercise was to measure the level
of student achievement of the whole primary education system by taking Grades 4 and 8 as
particular parts for investigation. The assumption was that Grade 4 was a terminal point for
investigating the outcomes of the first cycle primary education and Grade 8 for investigating
the outcomes of the whole primary sector of education.
Student home background: There is the recognition that student home background
variables are important determinants of student achievement. Chen (1996) reviews a lot of
literature to identify three sets of variables related to home background that influence student
achievement. These are home socio-economic status (SES), family configuration and parental
support. Home socio-economic status is measured using such indicators as parental
occupation, level of parental education, parental income and the prestige of parental
occupation. Family configuration measures include family size, sibling sex and spacing, and
birth position in a family. Measures of family support are both psychological and practical
and include such factors as academic guidance and support, stimulation to explore and
discuss ideas and events, language environment, academic aspirations and expectations, and
work habits of a family. In the Ethiopian Second National Learning Assessment, student
home background variables included the agreement between home and instructional
language, student chore time, distance to school, number of family members in school,
availability of television at home, availability of radio at home, and daily meals.
Student behavior: Activities, thoughts and feelings students demonstrate during their
learning are included in this conceptual construct in the Ethiopian Second National Learning
Assessment. Notable variables in this category included students' sense of ownership in
school property, time spent listening to the radio, time spent watching television, leisure
reading ( reading non-textbook materials), interest in English, interest in mathematics,
interest in chemistry, interest in biology, interest in physics and the use of school library.
Government and non-government primary schools in Ethiopia: Primary schools in
Ethiopia are divided into government and non-government schools in terms of ownership.
Government primary schools are administered and controlled by the government. Non-
Government schools are all schools controlled by institutions other than the government, and
these may include religious institutions, NGOs, the private sector, communities, etc. In the
Ethiopian Second National Learning Assessment government schools are those schools that
are administered and controlled by the government of Ethiopia.
Levels/cycles of primary schooling in Ethiopia: The Ministry of Education (2004)
classifies the levels or cycles of primary school in Ethiopia into two structures. These are the
First Cycle Primary School (Grades 1-4) structure and the Second Cycle Primary School
(Grades 5-8) structure. The Ethiopian Second National Learning Assessment used this
definition as a reference for its operations. But, schools can also be First Cycle to comprise
Grades 1-4, and Full Primary comprising Grades 1-8.
School structure: Chen (1996) includes three main variables in school structure. These are
school size, class size and cycles of schooling. School structure in the Ethiopian Second
National Learning Assessment was used to include the following variables: time to reach the
main road from the school, time to reach the Woreda education office from the school,

National Organization for Examinations 7

The Ethiopian Second National Learning Assessment Grade 4 Report

availability and condition of school pedagogical center, availability of football and volleyball
fields, availability and conditions of school library and school pedagogical center, income
generated from the sale of hay and vegetables, and school location.
Curriculum materials: These are materials that support the implementation of the curricula.
Curriculum materials are those resources that serve as daily guides for students and teachers
in directing activities related to instruction and learning. In the current study, curriculum
materials include the syllabus, teacher’s guide, student textbooks and reference books.
Teacher variables: Fuller (1986) classifies teacher variables into teacher quality and teacher
behavior in the classroom. Teacher quality variables include schooling, social background,
verbal proficiency, and motivation of teachers. Teacher behaviors in the classroom include
the efficient use of instructional time, the level of performance standards and expectations set
for students, the extent to which teachers evaluate students’ performance, and teachers' ability
to motivate students to learn. In the Ethiopian Second National Learning Assessment, teacher
variables include both quality and behavior indicators. Quality variables include teacher
qualifications, teachers’ total experience, experience at the school, distance to school, teacher
training on new syllabus, teacher training on new teaching methods, teacher training on new
assessment techniques, teacher training on student discipline and teacher training on
textbooks. From teacher behaviors, teacher perceptions or judgment of textbook difficulty
and student learning attitudes are included.
School management: is to mean the responsibility of the director to plan, organize, lead and
control schools to work well (Sergiovanni, 1995) as well as his own personal quality as a
leader (Fuller, 1986). In the Ethiopian Second National Learning Assessment, this component
includes director’s qualification, teaching experience, social obligations outside of directing,
director’s supervising of teachers, the distance the director has to walk home, director’s
meeting with teachers, and director’s ability in generating funds from different sources.
School quality: The term “school quality” has no agreed upon definition. However, in
relation to the Ethiopian Second National Learning Assessment, school quality refers to the
outcomes of education as measured by student’s level of academic performance or
achievement.
Instruction: This refers to teachers’ teaching practice and classroom organization (Chen,
1996, Fuller, 1996). Such variables as efficient use of instructional time, teachers’
expectations of pupils’ performances, time spent for preparing lessons, frequency of
homework, marking students’ homework, keeping order in the classroom, organizing
feedback in a systematic way, use of correctives in helping children to learn what they failed
to learn are studied in this variable. In the Ethiopian Second National Learning Assessment,
instruction included teacher’s instructional quality, the sum of homework assigned in subjects
by teachers, student understanding of subjects, parental academic support, and the percentage
of contents of curriculum taught so far.
Language of instruction: A medium in which instruction takes place in the classroom.

National Organization for Examinations 8

The Ethiopian Second National Learning Assessment Grade 4 Report

CHAPTER 2: A REVIEW OF RELATED LITERATURE

2.1 Introduction
The main purpose of education, especially of primary education, is to enhance economic and
social development of a country by creating learning opportunities at individual, community,
and national levels, and to expand literacy and give basis for further training and self-
education. To attain such major aims, various countries have been designing and employing
different strategies for expanding access and improving quality of schooling.
In developing nations, the actual reality of schooling is getting worse from time to time due
to the decreasing educational expenditure, on one hand, and the rapid increase of enrolment
rates on the other. This in turn contributes to the poor quality of schooling in general and to
the low level of students' achievement in particular. Consequently, there is a growing
awareness and shift of attention towards improving students’ achievement and the quality of
education. To this end, educational research has become the best tool to identify factors
(variables) that can affect students’ achievement and seek ways to design, implement and
then measure improvements.

2.2 The Concept of National Learning Assessment

Educational systems are known for public examinations, but national learning assessments
are relatively new (Kellaghan and Greany, 2004). The importance of monitoring learning
achievement grew rapidly after the 1990 World Declaration of Education for All (EFA) in
Jomtien. This declaration necessitated the introduction of a system or national assessment to
determine if children were acquiring useful knowledge, reasoning ability, skills and values
that schools promised to deliver.
The term assessment is used to refer to the process of gathering, interpreting, and applying
outcomes data on programs or entire curricula to improve program effectiveness, particularly
as measured by student learning outcomes. It is an ongoing process aimed at understanding
and improving student learning. It involves making expectations explicit; setting appropriate
criteria and high standards for learning; systematically gathering, analyzing, and interpreting
evidence to determine how well performance matches those expectations and standards, and
using the resulting information to document, explain, and improve performance (Angelo,
1999).
Learning outcomes are changes that occur within the student, and not what the instructor does
That is, learning outcomes are a level of knowledge, skills, abilities that a student has
attained. They are essential and enduring knowledge, abilities (skills), and attitudes (values,
dispositions) that constitute the integrated learning needed by a graduate of schools or
programs.
National learning assessment may be described as a systematic process of collecting relevant,
valid and timely information about the outcomes of schooling used for making decisions
about the development and learning of students. According to Kellaghan and Greaney (2001),
it is an exercise designed to describe the level of achievement, not of individual students, but
of a whole education system, or a clearly defined part of it. That is, national learning
assessment is meant to discover how well an educational system is progressing in general and

National Organization for Examinations 9

The Ethiopian Second National Learning Assessment Grade 4 Report

students are acquiring the knowledge and skills delivered by the educational system. Greaney
and Kellaghan (1996) indicate that all national assessments seek answers to one or more of
the following questions:
1. Do particular groups in the population perform differently? Are there disparities
between students’ achievements of different regions, of boys and girls, of rural and
urban locations, and of different language groups?
2. How well are students learning in the education system? (aims of curriculum)
3. Is there evidence of particular strengths and weaknesses in students’ learning?
National learning assessment emphasizes the measuring of students’ acquisition of
knowledge, level of understanding, attitudes, values, skills, satisfactions, actions and
intellectual growth.

2.3 Purposes and Functions of National Learning Assessment

National Learning Assessment seeks to determine how well students are learning, and it is an
integral part of the quest for improved education. National assessments are systematic,
regular measures of learning achievements in a country that are designed to assist policy
making. Indeed, national learning assessment provides feedback not only to policy makers,
but also to educators, parents, students and the public at large about the effectiveness of
educational services and quality of students’ learning.
Information obtained from national learning assessment may be used for a variety of
purposes. According to Greaney and Kellaghan (1996), these may include the following:
ƒ Monitoring standards: Information about students’ learning and achievement in
school subjects collected on a regular basis that help to monitor changes of
achievement over time.
ƒ Informing policy makers: Results of the analysis of the learning outcomes can have
implications for curriculum design, teacher training, planning and allocation of
resource.
ƒ Introducing realistic standards: The results of national learning assessment can foster
a sense of realism in the discussion about the appropriate achievement levels we
should aim for.
ƒ Identifying correlates: Information on relationships between student learning
outcomes and system input and characteristic variables may help policy makers to
identify factors over which they can exercise some control to direct the scarce
resources and materials. Therefore; data, on some of those variables which can
potentially be manipulated, have to be collected along with achievement data at the
time of the national assessment.
ƒ Promoting accountability: Governments spend a substantial part of their public
resources on education. To retain the public support for education, the government
and the people need information which substantiates that the money spent on
education is producing skills that are desirable in the students. So, governments need
access to relevant information on the operation of their education system.
ƒ Increasing public awareness: Results of national learning assessment can be used to
bring an aid of reality, a level of integrity to discussions about the education system
and increase public support for quality.

National Organization for Examinations 10

The Ethiopian Second National Learning Assessment Grade 4 Report

In general, national learning assessment is used to manage expectations, direct the curricula,
and create a culture of continuous improvement, culture of accountability and improve
planning and adjust policy.

2.4 Methodological and Technical Considerations in National

Learning Assessment

National learning assessment is a highly complex process that involves a number of

interrelated dimensions, each guided by theory and practice. Its design and implementation
process requires a coordinated effort and consensus of a wide range of stakeholders and
decision makers.
According to Greaney and Kellaghan (1996), the design and implementation of a successful
national learning assessment requires the involvement and consensus of major stakeholders.
It includes those organizations and individuals who are responsible for administering the
national assessment, those who will consider the results for policy making, those responsible
for funding the exercise, and those who involve in policy making and reform activities.
All these and other concerned bodies have to have a room or chance for participating at the
different levels of the assessment process through various mechanisms. For example, by
establishing national steering committee, organizing a highly technical working group,
selecting item-writing groups and data collectors, etc; each would have several functions
right from the stage of designing the assessment conceptual framework to the levels of data
analysis and report production.
Any national learning assessment program requires collecting and systematizing a host of
empirical data that can be analyzed and interpreted as indicative of learning in specific
curriculum areas at clearly identified levels of students’ performances.
The scope of the data needed to conduct a national learning assessment can determine the
nature and techniques of the sampling process. Most of the national learning assessment
programs so far conducted around the globe are based on a representative sample of both
students and items.
In general, the sampling design has three stages: 1) selection of primary sampling units (eg.
regions); 2) selection of secondary sampling units (Schools), and 3) selection of tertiary
sampling units (students). This type of sampling design is known as a multi (three) stage
stratified sampling procedure, which is too complex to analyze data and interpret findings in
relation with different variables.
The size of the sample on which data are collected depends on the purpose of the national
assessment and the number and homogeneity of students, teachers, schools and regions that
the national assessment plan addresses. For example, the United States NAEP’s sample
consists of approximately 0.4 percent of the total student population for each Grade (NCES,
2003).
The type of information to be collected for conducting a national learning assessment should
cover the issues and subjects based on policy needs. Available literature indicates that most
of the national assessment programs focus on knowledge, conceptual understanding,
investigation and practical reasoning, attitudes, outcomes, and background variables of both
school and non-school factors. For example, the National Assessment of Educational
Progress conducted in USA assessed reading and mathematics every two years in Grades 4
and 8 (NAEP, 2003). In general, language, mathematic and science subjects are the most

National Organization for Examinations 11

The Ethiopian Second National Learning Assessment Grade 4 Report

widely considered areas of curricula by various NLAs. Language, mathematics and science
subjects are emphasized in most of the national assessment programs because:
ƒ the language skills are essential tools not only because they serve as the necessary
bases for further learning and career development but also because they enable the
human spirit to be enriched, foster responsible citizenship, and preserve the collective
memory of a nation,
ƒ mathematics is critical for all students in that it trains the mind to be analytic-
providing the foundation for intelligent and precise thinking, and
ƒ Science is an organized body of knowledge and a method of proceeding to an
extension of this knowledge though hypothesis and experiment.
The instruments used in national assessments should undergo extensive scrutiny, review, and
pre-testing before operational use. National assessment analysis methods combine results
across different test books within an assessment year and produce results that are comparable
across assessment years despite changes in the item pool. The results provide information
about students’ proficiency in subject area and content sub-domains in terms of average
scores and percentages of students at or above the achievement levels for the subject in
addition to providing information about trends.
Data analyses may be designed to compare group performances by location, gender and type
of school or it may be required to provide information about relationships between student
achievement and characteristics of students, schools, and teachers over time (Greaney and
Kellaghan, 1996).
In the United States, for example, National Assessment of Educational Progress (NAEP)
results were reported in terms of predetermined achievement levels because each assessment
reflects current standards of performance. The NAEP reports on student performance with
comprehensive information about what students at Grades 4, 8, and 12 know and can do in
various subject areas. It also describes relationships between achievement and certain
background variables.
In 1999, the Uganda National Examinations Board (UNEB) conducted a sample-based
assessment on Grade six students to measure the mastery levels of reading and writing in
English. The findings of this assessment indicated that 98% of the Grade 6 pupils in Uganda
failed to achieve the advanced grade, and that 35% of the pupils failed to achieve the
minimum acceptable level of competency in English. And, pupils in urban schools have a
distinct advantage over their peers in rural schools (Kellaghan and Greaney , 2001).
Regarding gender disparity, at both Grades 4 and 8, male students scored higher on average
than female students.

2.5 Experiences of Other Countries Regarding National Learning

Assessment

Experiences in conducting national learning assessments vary among countries. The longest
– running and best-known national learning assessments are in the United Kingdom, which
existed since 1948; the United Sates which conducted its first national learning assessment in
1969; and France which introduced the system in 1979. From developing countries, the
longest-running national learning assessment is that of Chile, and it dates back to 1978. Most
economically less developed countries including Colombia, Vietnam, Thailand, and many
African nations started national learning assessments in the 1990s.
There are two basic models for the implementation of national assessments. One is sample
based (analytical view of achievement) derived from USA and the other is a census type

National Organization for Examinations 12

The Ethiopian Second National Learning Assessment Grade 4 Report

(holistic performances) derived from the United Kingdom assessment (Greaney and
Kellaghan, 1996).
In the United States, the National Assessment for Educational Progress (NAEP), which is
mandated by the National Congress, has become a standard feature of the education system
since 1969.The objective of the program is to measure students’ achievements at specified
ages and Grades (4, 8, and12) on 11 instructional areas. Since 1969, assessments have been
conducted periodically in reading, mathematics, science, writing, and other subjects. By
making objective information of student performance available to policymakers at the
national, state, and local levels, NAEP is an important part of the nation’s evaluation of the
condition and progress of education.
England and Wales first applied a large scale survey or national assessment in 1948 at the age
levels of 9, 11, and 15 years. In 1978, an improved assessment system was made on three
main areas; language, mathematics and science at ages of 11, 13, and 15 years. It was given
much weight by politicians in the 1980s and exhibited the various functions of assessment
such as formative, diagnosis, summative, and evaluative (Greaney and Kellaghan, 1996).
National learning assessment in France has been introduced since 1979 using both sample
and census models of USA and United Kingdom respectively. On the sample based, students
are assessed about every five years at the end of Grades 7, 9 and 10 to provide information on
achievements at the system level in relation to the curriculum. In the other method, the total
population of students in Grades 3, 6, and 10 are provided with diagnostic assessment
designed to provide information on individual schools and feedback is sent to schools,
students, and parents with assisting teachers to adapt their pedagogical skills to the needs of
their students.
The assessments conducted in six countries of Latin America -Chile, Argentina, Brazil, Costa
Rica, Mexico, and Colombia- showed similar results in that students scored far below
expectations, and students from urban and private schools scored better than their
counterparts. Chile is the most experienced country in conducting national learning
assessment among these countries. The first national assessment, School Performance
Assessment Program (PER), was carried out from 1978 to 1984 at Grades 4 and 8 with the
general goal of providing information on the extent to which students achieve learning
standards or targets. Achievement tests on Spanish (reading, writing), mathematics, natural
and social sciences were used for 90% of the population, and data on contextual variables
including home, teacher and classroom, principal and school and instructional practices were
collected (Himmel, 1984). After 4 years of interruption, national learning assessment was
reintroduced in 1988 under the National Program for Quality Assessment of Basic Chilean
Education (SIMCE). Findings showed that the performances of students were poor on
curriculum objectives in general, but better in urban schools than in rural, in large schools
than in small, and in private schools than in government schools. As a result of the findings,
low scoring schools were given special fund, curriculum reform was attempted and percentile
rank system was changed to percentage.
Colombia is another country in Latin America that conducted an assessment in 1991 on
Grades 3 and 5.The purpose was to find out to what extent students attained the minimum
standards set in mathematics and language. Findings showed negative correlation for grade
repetition, absenteeism, time spent getting to school and family size. Number of courses
teachers took was not a significant predictor of achievement .The final result was released
through mass media for discussion at national level, at local workshops, newsletters,
brochures and friendly documents.
In Asia, learning assessment was made in a number of countries including Sri Lanka,
Bangladesh, Nepal, Pakistan, Cambodia, Thailand and India. Notable among these

National Organization for Examinations 13

The Ethiopian Second National Learning Assessment Grade 4 Report

assessments was the one conducted in Thailand. After two years of abolishing the public
certification examination, Thailand introduced national assessment in 1983 at Grade 12. The
main purpose of the assessment was to determine the national education standards schools
reach and encourage schools to broaden their objectives and instructional practices. Students
were assessed on mathematics, the sciences and career education in both cognitive and
affective areas. The task was given to the Office of Educational Assessment and Testing
Services in the Department of Curriculum and Instruction Development. The assessments
continued in subsequent years. Every second year, samples of Grades 6 and 9 were assessed
and expanded to include measures of school process. The result was provided to each school,
region and province and for the public too. As Prawalpruk (1996) indicated, the findings of
national learning assessment in Thailand had been used for school and provincial planning,
monitoring of achievement levels of students over time and increasing the interests of
teachers in learning outcomes.
Greany and Kellanghan (1996) identified 4 major categories or approaches to the
implementation of national learning assessments in Africa: Monitoring Learning
Achievement (MLA), South Africa Consortium for Monitoring Educational Quality
(SACMEQ), Francophone countries (PASEC), and others.
The Monitoring Learning Achievement (MLA) project was initiated in 1992 by a joint
UNESCO/UNICEF support. It was an attempt to monitor the extent to which students
achieved in literacy, numeracy and life skills for Grade 4, and the knowledge, ability and
skills in mathematics and the sciences for Grade 8 students. In both Grades, data were
collected through tests and background questionnaires for students on school characteristics
and home issues. MLA was conducted twice: MLA I focused on Grade 4 and MLA II on
Grade 8. About 47 African countries were involved and by 2003 reports were published on
MLA I assessments of 18 Sub-Saharan countries, and MLA II of 2 countries. Separate MLA I
reports were prepared in 11 countries (Botswana, Madagascar, Malawi, Mali, Mauritius,
Morocco, Niger, Senegal, Tunisia, Uganda and Zambia). Results indicate that only 4
countries had met the learning target set in Jomtien (i.e., that 80 percent of learners should
attain the defined learning competencies) for Grade 4 pupils in life skills, only two countries
met the target in literacy, and one in numeracy. In all the countries, gender difference was not
found to be insignificant. The ability of parents to assist their children with their schoolwork
positively correlated with student achievement.
The other approach to the national learning assessment is that of SACMEQ. The Southern
African Consortium for Monitoring Educational Quality (SACMEQ) is a collaborative effort
of 15 Ministries of Education and the International Institute for Educational Planning (IIEP).
The first SACMEQ assessment was conducted between 1995 and 1998, and only 8 Ministries
( Kenya, Malawi, Namibia, Tanzania/Zanzibar, Zambia, Mauritius and Zimbabwe) collected
information on baseline indicators for educational inputs, general conditions of schooling,
equity assessments for human and material resource allocations, and literacy levels among
Grade 6 students. The second SACMEQ study was conducted between 1999 and 2002, and
15 countries in the region participated. One of the major aims was also to promote capacity
building by equipping educational planners in member countries with the technical ability to
monitor and evaluate schooling and the equity of education.
Like in the MLA project, findings from SACMEQ indicated that education systems were
failing to meet performance standards of ministries. For instance, in Namibia and Zambia,
fewer than 30 percent of Grade 6 pupils met the specified minimum literacy standards.
According to these findings, some problems such as lack of equipment and facilities
(Zambia, Zimbabwe), lack of text books (Zimbabwe), unqualified teachers (Zambia and
Zanzibar), lack of books and writing materials (Zimbabwe), high repeaters, un-repaired
buildings and high inequity (Zanzibar), high pupil-teacher ratio and lack of seats

National Organization for Examinations 14

The Ethiopian Second National Learning Assessment Grade 4 Report

(Zimbabwe), low frequency of tests/ homework (Zambia, Zimbabwe), district variation

(Mauritius), and students’ absenteeism were observed as the major determinants of students'
performance.
The third type of national learning assessment in Africa is PASEC (Programme d’ Analyse
des Systems Educates des Pays de la CONFEMEN). PASEC was established as a network
for sharing information on educational evaluation, instruments and results among 12
Francophone countries (Burkina Faso, Cameroon, The Republic of Congo, Cote d’Ivoire,
Djibouti, Guinea, Niger, Central Africa Republic, Mali, Senegal and Togo). It encourages the
involvement of senior decision makers and stakeholders to identify policy issues, base
decisions on data and follow up decisions for actions. At first, the assessment began on
Grades 2 and 5 for literacy (French) and mathematics. Starting from 1994, the study included
all Grades from 2 to 6 and same tests were used in all countries. Other data on school and
background factors were also collected from schools. This network differs from the others in
that the assessment is made twice in an academic year, at the beginning of November and at
the end of May. As a result, this approach reflected a strong research orientation of the
program and it attempted to identify causal relationships on which to base policy that seeks to
improve the efficiency and effectiveness of education. Efforts have also been made to
determine the impacts of in-school factors (teacher training, class size, textbooks availability)
and out-of-school factors (parents education, home language, distance from school).
The 4th group includes some countries which attempted national assessments of their own:
Burundi, Eritrea, Mali, Senegal, Uganda and Zambia. For example, in Namibia, National
Learner Baseline Assessment measured students’ achievements in English and mathematics
at Grades 4 and 7. In Eritrea, students were assessed in six regions to determine if they had
mastered basic skills in the official curriculum, basic skills in their mother tongue (Grade 1)
and English and mathematics (Grade 4). The assessment identified particular areas of the
curriculum causing problems. It found that boys outperformed girls, and identified
implications for teacher education and teaching. The experiences of some counties can be
found in Greaney and Kllanghan (1996).

2.6 Background and Community Factors that Influence Student

Achievement
Various studies emphasize a host of factors, both external and internal to schools, to have
influences on student achievement. Among the external factors, the influence of student home
background, personal and community characteristics are widely acknowledged.
With reference to student home background, well known influences to student achievement
relate to parental socio-economic status (SES), family configuration and parental support.
Among the indices of SES are parental occupation, parental level of education, parental
income and occupational prestige. Husen (1967) reported that in the first IEA mathematics
study, the total mathematics score at 13–year-old level correlated 0.16 with father’s education
and 0.12 with mother’s education over 12 countries. The correlations between father’s
occupational status and the mathematics scores were 0.25 at the same age level of the same
populations. In the IEA six-subject studies, it was found that achievements in science, reading
comprehension, literature and the cognitive side of civic education were linked to the socio-
economic background of the home. In family configuration, family size was found to have a
negative relationship with student achievement. Increased spacing between children’s birth
reduced the decrease in scholastic performance, although this did not hold true for developing
countries. With regard to parental support, studies indicated that parental attitudes towards
education, their interests in their children’s education, and their beliefs in the values of
schooling are related to students’ scholastic performance.

National Organization for Examinations 15

The Ethiopian Second National Learning Assessment Grade 4 Report

When it comes to students’ personal characteristics, it has been observed that the relationship
between gender and achievement varies based on the subject and the level of students under
investigation. Parelius and Parelius (1987) noted that girls’ gradess in both elementary and
high schools tend to be higher than that of boys. However, boys often scored higher on
standardized tests. Fagerlind and Saha (1989) observed that in most developed countries, girls
performed equally with boys at primary school level, but at secondary school, girls began to
do more poorly. Subjects also make a difference in achievement of boys and girls. For
instance, a comparison of the reading performance of boys and girls in many countries has
frequently shown differences in favor of girls (Thorndike, 1973). However, girls fall behind
boys in mathematics and science. In the second IEA science study in 17 countries, it was
found that boys scored higher than girls at all levels, and the gap increased from the 10-year-
olds to the 14-year-olds (Chen, 1996). TV watching, as the other student characteristics, is
inversely associated with student achievement (Comstock, 1994). However, leisure or
voluntary reading as well as the average amount of time spent on homework of all countries
have shown positive relations with student achievement.
Just as students may be influenced by their classmates, their teachers and the norms of their
schools, they are influenced by the communities in which they live. Some of this influence
can be transmitted through families, but some may happen through direct influence of
communities. Variations among communities can take various forms and this has a great deal
of influence on student achievement. Traditionally, communities are divided in geographical
neighborhoods taking urban, rural or suburban forms. Several studies indicate that such
divisions have impact on student achievement, attainment, aspirations, and school climate
(Stockard and Mayberry, 1992). Just as the geographical space of schools influences learning,
it also affects the resources that a school can offer its students. The resources available to
schools depend to a large extent on the economic basis of its inhabitants. Parental and
community involvement in schools, community participation in school control, and
community values all have their influences on student achievement. In developing countries,
demand for child’s labor and opportunity costs of attending schools also exert a significant
impact on academic achievement.
2.7 School Based Factors that Influence Student Achievement
The analysis of contextual or external factors provides only a partial view of the influences
on student learning. To fully understand how higher achievement is facilitated, it is important
to examine internal factors at school and classroom levels. Stockard and Mayberry (1992)
discuss these factors under two generic divisions: school and classroom environment, and
school resources. In the school climate/environment, they include academic expectations and
excellence, school leadership, orderly environment, and school coherence, teacher and
student morale, and effective teaching. School resources comprise school facilities and per-
pupil expenditures, teacher qualifications, classroom size, and school size. Fuller (1986)
analyzes these factors under the generic concept of school quality. This construct comprises
material inputs, teacher quality, teaching practices and classroom organization, school
management and structure. Chen (1996) discusses internal factors influencing student
achievement as school background factors. In this category, he included school structure,
instructional facilities, teachers’ instructional time, and the curriculum.
With reference to school environment, Stockard and Mayberry (1992) refer to various
materials to indicate the significance of 4 main dimensions. These are the ecology (the
physical environment), milieu (the social characteristics of individuals and groups
participating in organizations), social system (patterned relationships of persons and groups),
and culture (the collectively accepted beliefs, values, and meanings of the group). School
environments are commonly measured using indices containing a variety of items based upon

National Organization for Examinations 16

The Ethiopian Second National Learning Assessment Grade 4 Report

the above factors. Various studies in the area demonstrate that school environment is linked
to school achievement.
According to McDill et.al, (1967), McDill and Rigsby (1973), schools with teachers and
students who see or expect higher achievement and academic excellence as a real and
attainable goal actually do have higher achievement. Among the most important attitudinal
factors related to increasing school achievement in this respect is staff commitment to
improved academic performance, high or increasing expectations of teachers about students,
high opinions of students’ abilities and peer norms emphasizing academics.
Strong school leadership and management is another internal characteristic of schools which
determines student achievement. Several investigations indicate that principals, who are
involved in instruction, communicate high expectations, promote good feelings and
collegiality between faculty and administrators and among faculty members. They also
encourage teacher participation in the school’s decision-making processes which is effective
in promoting success. Effective administrators promote cohesive relations within schools
performing a balancing act. They promote higher achievement and develop ethos or culture
that enhances morale, mutual trust and respect, and shared norms and values (Ellet and
Walberg, 1979; Fullan, 1990).
Orderly environments and teacher and student morale are important ingredients for success in
school. Edmonds (1979) indicates that student achievement is enhanced in schools that
maintain a consistent set of rules and values that clearly map out school goals and policies
while also promoting purposefulness and pleasure in learning. Teacher’s morale in terms of
job satisfaction and students’ recognition of teachers’ morale in their work exhibit high level
of attendance and achievement (Brookover and Lezotte, 1979). Schools that can nurture high
morale among students and staff maximize the chances of developing attitudes about
individual abilities and learning environment that promotes higher level of achievement.
Specific material inputs enhance student learning. According to Fuller (1986), material inputs
that consistently relate to student achievement in developing countries include instructional
materials (texts, desks and reading materials), instructional media, and school building
quality, library size and quality and nutrition and feeding programs. His report indicates that
class size, school size and laboratories do not consistently relate to student achievement.
Teachers’ characteristics and quality have great effects on students’ achievement. As Fuller
(1986) mentions, teacher quality elements that consistently impact student achievement in
developing countries include total years of teacher’s schooling, years of tertiary and teacher
training, in-service teacher training, teacher’s verbal proficiency, and teacher’s social
background. Among the elements of teacher quality that show no consistent relationship with
student learning in developing countries are individual teacher’s salary level, total years of
experience and punctuality. Studies in developed countries show mixed results for the impact
of teacher qualifications on student learning. According to Stockard and Mayberry (1992),
studies which focus on the impact of teacher qualification on student achievement show
only 50 % confirmation rate, while those studies that investigated the impact of teacher’s
attending prestigious schools show very little relation with student achievement. A little
more than 50% confirmation rate was obtained from studies which examined the impact of
teacher’s experience. However, a significant positive relation was observed between teacher
verbal proficiency and longevity on the one hand, and student achievement on the other.
Again, teacher’s salary has not shown a consistent or significant impact on learning.
Effective teaching practice makes a difference in student achievement. For developing
countries, the most important influences come from the length of instructional time,
frequency of homework given, teacher’s expectation of pupil performance, and time spent by
teachers on classroom preparation (see Fullan, 1986). In developed nations, studies suggest

National Organization for Examinations 17

The Ethiopian Second National Learning Assessment Grade 4 Report

that the quantity and pacing of instruction (i.e. curriculum covered and time spent on
instruction), the way in which teachers give information (i.e. well-organized, well structured
presentations), the way in which teachers question students and wait for responses, and the
way they handle seat work and homework all influence student achievement (see Puff, 1978;
Klitgard and Hall, 1973; Rosenshine, 1983).
The length of instructional time is an important indicator/school factor that boosts
achievement. Instructional time varies from the number of school days in the school year to
the hours the subject is studied during a week. This instructional time is bounded by the
variety of material resources and management practices. Teacher’s knowledge of the subject
is strongly and consistently related to student performance. Although there is no specific
teaching practice that is universally effective, teachers without a limited repertoire of
teaching skills appear to be more effective than those with a limited repertoire (Chen, 1996).
Active teaching - learning interaction in the classroom and the frequency of evaluating
students produced better effects on student achievement (Brook, et al., 1979). Clear
explanation of the materials by the teacher and more time spent on concrete learning also
affected student learning positively (Heyneman and Loxily, 1983). A focus on students’
perceptions rather than on teachers’ is likely to be more productive; it attempts to improve
and understand classroom learning. Students’ perceptions vary with such factors as teacher
personality, class size, Grades level, student gender, subject matter, the school level,
environment, and the type of school.

2.8 Historical Overview of Learning Assessments in Ethiopia

2.8.1 ESR and ERGESE

In Ethiopia, very limited attempts have been made to review the education sector since the
introduction of modern education in 1908. The Education Sector Review (ESR), the
Evaluative Research of the General Education System in Ethiopia (ERGESE) in 1983 to
1986, and the Ethiopian National Baseline Assessment (ENBA) on Grades 4 and eight
students’ achievement launched in 1999/2000 are the major attempts made to review the
education sector in the history of the Ethiopian education system.
The Education Sector Review (ESR) did not investigate the quality of education system,
rather it devoted its attention to strategies for a rapid expansion of primary education with
the view of achieving universal literacy before the year 2000 (Tekeste Negash, 1990). The
ERGESE project was initiated by the Ministry of Education based on the resolution passed by
the government in 1983 to review the education sector. The purpose of the project was to
investigate the quality of primary and secondary schooling in Ethiopia and to make
recommendations for improvement (MOE, 1986). The findings of this study indicated that
while educational opportunities had widened, there had not been a corresponding
improvement in quality.
These two projects ESR and ERGESE, therefore, could not be considered as national learning
assessment programs of the Ethiopian education system. This was because these two studies
did not collect and analyze data on the actual learning of students systematically and
thoroughly. Rather such attempts could be considered as survey studies of the overall status
of general education system of the country at that particular time.

National Organization for Examinations 18

The Ethiopian Second National Learning Assessment Grade 4 Report

2.8.2 EBNLA (1999 – 2000)

The Ethiopian National Baseline Learning Assessment was launched in 1999/2000 (1992
E.C.) initiated by the Ministry of Education (MOE) with the need to evaluate and improve the
results of the ongoing reform in the education system over certain periods of time. In the
process, the National Organization for Examinations (NOE), and the Basic Education System
Overhaul (BESO I ) project played major roles.
The main objective of this National Baseline Learning Assessment was to determine the
various levels of students’ performances at both Grades 4 and eight in 4 key academic subject
areas. Grade 4 students were tested in English, basic reading, mathematics and environmental
science subjects, all prepared in the different instructional languages; and Grade 8 students
were assessed in English, mathematics, chemistry, and biology. Moreover, teachers and head
teachers and the overall conditions of school compounds, in addition to students, were
considered as major sources of the data collected for the study (NOE, 2000).
A three stage stratified random sampling design was used to select sample regions, schools
and students at both grade levels (4 and 8). Accordingly, 256 schools for Grade 4, and 136
schools for Grade 8 studies were sampled. About 10,506 students for Grade 4, and 5099
samples for Grade 8 studies were tested across ten regions of the nation (NOE, 2000).
Information on the background characteristics of students and teachers were also collected
and analyzed. The implications of the findings for participating regions were indicated.
According to the findings of the study, all participating regions have room for improvement
in all key subject areas since no one region scored above the acceptable minimum level of
50% achievement (NOE, 2000).

The results also indicated that schools with high achievement at Grade 4 level also tended to
obtain high achievement at Grade 8. The results of the study were also discussed in a
workshop and constructive recommendations were provided for the improvement of the
quality of students’ learning in the nation. The reports were disseminated to regions and other
stakeholders.

National Organization for Examinations 19

The Ethiopian Second National Learning Assessment Grade 4 Report

CHAPTER THREE: FRAME OF ANALYSIS AND

METHODOLOGY
3.1 Frame of Analysis

The main purpose of the Second National Learning Assessment was to find out the extent to
which learning takes place in the Ethiopian primary education, and determine the main
conditions that influence the learning outcomes of students. It was also to generate
information on school improvement from the first national learning assessment conducted in
1999/2000.
The dependent variable, student learning, was measured by achievement tests. Another
dependent variable, student’s attitude towards socially relevant issues, was measured by
questionnaires. The independent variables that refer to factors which affect the achievement
of student learning in this study included school condition/environment, teachers behavior,
school management, school structure and supply, instructional support, language of
instruction, students’ background, and community opinions. The following table shows their
relations and descriptions.
Table 1. Description of variables

Variables Description
Total achievement mean score for Grades 4 and 8
Students Mean score of basic reading comprehension, English, mathematics and
achievement environmental science for Grade 4
Mean score of English, mathematics, biology, physics, and chemistry for
Dependent Grade 8
Variables
Students’ attitude Attitude of Grade 4 and 8 students towards socially relevant issues included
health, environment, civics and ethics, and the school.

Students’ home Family size, parents’ education and occupation, language at home, distance
background from school, attendance, learning support.
Student personal Students’ gender, students' sense of ownership of school property, time spent
characteristics listening to the radio, time spent watching television, leisure ( reading non-
textbook materials), interest in English, interest in mathematics, interest in
chemistry, interest in biology, interest in physics and the use of school
library
School structure Location (urban-rural), school program, level, instructional language, class
and curriculum size, program, classroom condition, supplies, facilities and equipment,
materials provision of instructional materials, period allotment, school construction,
Independent class size
Variables Instructional inputs Textbook-pupil ratio, availability of basic instructional materials (other than
textbooks), availability and use of pedagogical center, lab., library
Teacher variables Teacher’s qualifications, years of experience, knowledge of subject matter,
distance from school, teaching load, attendance or absenteeism
School School directors’ qualifications, organization, evaluation of teachers,
management meetings conducted with staff and community
Parent/community Attitude towards students’ behavior, learning and schools, collaboration
views with the school to solve problems, parents’ involvement in school affairs

National Organization for Examinations 20

The Ethiopian Second National Learning Assessment Grade 4 Report

3.2 Design

In order to obtain the required information for the proposed research questions, both
quantitative and qualitative research approaches were used. In the quantitative approach, a
cross-sectional data using achievement tests were collected to determine the extent to which
learning takes place in primary schools. Moreover, a baseline study design which compared
student achievement results of the Ethiopian Baseline National Learning Assessment with the
Second National Learning Assessment was used in order to monitor school improvement.
Since the tests were given in different years to two independent samples of the same Grade,
threats to internal validity due to history or maturation effects were controlled. A qualitative
study design was used to supplement the quantitative study. It mainly involved a collection of
cross-sectional data on similar issues addressed by the quantitative study. Wherever
necessary, a historical approach has been used to understand the change of events over time
in the qualitative design.
The Second National Learning Assessment has been carried out in three stages. The first
stage involved institutional arrangement for carrying out the study. In addition to mapping
out the mission of the entire study, this task led to establishing the National Advisory
Council, the Technical Working Group and the Secretariat of the Second National Learning
Assessment as responsible key structures for its implementation. Terms of Reference for
each of these structures were prepared. Accordingly, the National Advisory Council (NAC)
was responsible for providing overall leadership to the study. The chairman of the NAC was
the Vice Minister for General Education and the Manager of NOE was the secretary. The
members of NAC included leaders of central institutions of the Ministry of Education (MOE)
and the Heads of Regional Education Bureaus, AED/BESO II and USAID. The Technical
Working Group (TWG) was established to provide leadership and direction on everyday
basis to the development and implementation of The Ethiopia Second National Learning
Assessment. The secretariat coordinated the required logistic and other support to the study.
The second stage involved planning, development and field testing of the Second National
Learning Assessment activities, materials and instruments respectively. Planning the second
national learning assessment activities included the identification of preparatory activities,
making decisions on the design of the main study, fieldwork as well as dissemination.
Development activities included reviewing the literature and the previous assessment
documents, review of curricular profiles, initial revision of instruments, pilot testing of
instruments (in three languages), translation, printing and packaging of instruments.
The third stage involved field work, data analysis and interpretation. The field work, which
began by selecting and training of data collectors, focused on the collection of both
quantitative and qualitative data. This was followed by data organization, cleaning, analysis
and reporting.

3.3 Sampling Procedures

In order to provide national and regional estimates of student achievement results and
attitudes on selected curriculum based topics with group comparisons across gender, location
of school ( i.e., urban/rural), and language of instruction, all students of Grades 4 and 8 in the
Ethiopian primary schools were taken as the target population. To be able to obtain data on
independent variables using quantitative methods, teachers and head teachers were targeted.
The target population for the qualitative study included the purposefully sampled schools
focusing on students, teachers, head teachers and community representatives.

National Organization for Examinations 21

The Ethiopian Second National Learning Assessment Grade 4 Report

3.3.1 School Sampling

The sample size used in the Ethiopian National Baseline Learning Assessment provided the
minimum number of schools to be sampled as a starting point. In that study, the total number
of schools included was 272. By taking into consideration the expansion of primary schools
since this assessment in 1999, an assumption to consider 300 sample schools was initially
accepted. After using a simulation procedure to find out the acceptable minimum number of
samples to make strong estimation of achievement results from the national sampling frame,
and the representation of regions considered, the number of sample schools was determined
at 407. The selection of samples was done for all regions except for Gambella.
To meet the goal of sound statistical estimates of performance for the nation, a multi-stage
stratified random sampling technique was used. The number of sample schools in each region
was determined based on the relative proportion of its school population. Prior to selection,
the sampling frame comprising the number of schools by regions, levels and location was
collected from EMIS (2002/3) statistical document. Following this, schools were stratified
based on region. The decisions to use a stratified sampling procedure were due to the
following reasons.
1. To accurately represent individual regions and the geographical locations
(urban/rural)
2. To have a good representation of various linguistic groups
3. To ensure adequate representation of specific groups of the target population in the
sample
4. To obtain reliable estimates for each stratum, if required
As language, in general, is tied to geographic region or zone, the stratification on zone was
assumed. Emphasis was given to the proportion of schools per region (and zone to SNNPR
only).
A minimum of 25 primary schools were randomly selected from each region. This sample
size was determined from confidence intervals estimated using population statistics from
baseline data. Using simulation procedures, a 90% confidence interval of +/- 5% from the
mean was calculated for a minimum regional sample of 25 schools. A 90% confidence
interval of +/- 8% of the mean was calculated for a minimum number of 15 schools in a
region, the minimum number that would be tolerated for selecting the set of full primary
schools where Grade 8 tests were conducted.
To make school based paired comparison or to investigate what changes have been made
from the Ethiopian Baseline National Learning Assessment study to the Ethiopian Second
National Learning Assessment, it was decided to consider 10 of the sample schools from each
region. This sampling procedure was not aimed at making generalizations, but to determine
school improvement using the schools under investigation as units of analysis.
When stratifying the minimum number of schools, regions with less than 25 schools were
“topped up” in order to meet the minimum representation of regions by a sample of 25
schools. In order to ensure representation in the context of extreme diversity of nationalities
in the SNNPR, the sample size for this region was “topped up” to 80 schools.
The randomization of sample schools for each region was conducted by assigning a random
number to each of the schools in the region, and then by sorting the schools according to the
assigned random number. The target number of schools was counted from the top of the
sorted schools in the spreadsheet.

National Organization for Examinations 22

The Ethiopian Second National Learning Assessment Grade 4 Report

Based upon the above sampling procedure, a total of 407 sample schools from 12471 national
totals was selected. Oromia contributed the highest proportion of the sample schools with
38.15% followed by Amhara 24.14% and SNNPR 19.45%.
Table 2. Sample schools planned and visited by region
Sample schools by Grade
Total Topped up Data Planned for Actual Data Collected
Actual
Primary Sampled collection
Region Proportion
Schools* Schools
Grade-4 Grade- 8 Grade- 4 Grade- 8
Tigray 967 7.75% 25 25 18 25 12
Afar 156 1.21% 25 25 14 25 11
Amhara 3,011 24.14% 50 50 30 47 26
Oromia 4,758 38.15% 77 77 53 76 52
Somali 301 2.41% 25 25 10 24 9
Ben-Gumuz 275 2..21% 25 25 14 25 12
SNPPR 2,426 19.45% 80 80 61 80 43
Gambella 149 1.19% 25 25 14 X X
Harari 48 0.38% 25 25 16 24 11
Addis Ababa 313 2.50% 25 25 21 25 21
Dire Dawa 67 0.54% 25 25 16 25 16
Total 12,471 100% 407 407 267 376 213
*Source: EMIS 2002/3

The above table shows that it was planned to cover 407 primary schools from all regions. Of
these schools, it was planned to include 267 schools having Grade 8 and 407 or all primary
schools having Grade 4. However, the actual number of schools from which data were
collected in the field was 376 for Grade 4, and 213 for Grade 8 (excluding Gambella Region).
Furthermore, the number of sample schools planned by location was 136 for urban and 271
for rural but data were actually collected from 126 and 256 respectively.
The selection of schools for the qualitative study was accomplished on the filed. Prior to this,
however, a decision was made to include up to 50 schools overall. It was managed to
nationally cover 44 schools for this purpose with varying degree of coverage from one region
to the other. In Tigray, Harari, Addis Ababa and Dire Dawa the number of schools included
was 2 each. In Afar and Beneshangul Gumuz 4 schools each were included while in Oromia
and Amhara the number was 6 each. In Somali, only 1 school was used for collecting
qualitative data while the number was 15 for qualitative data. The details are given in
Table 2.

3.3.2. Sampling of School Directors, Teachers and Students

The sampling of schools was followed by another decision to select students, teachers and
directors. It was decided to randomly select only one section from each grade. Within each
section the maximum number of students randomly selected was 40 and less in the cases
where there were fewer students to meet the maximum number. The decision to include a
maximum of 40 students was made by the study team assuming this to be an average number
that can be managed during the test administration.
The school director and all teachers of the sampled students in the respective grades (4 and 8)
were also selected to fill in the questionnaires. The following table provides a summary of
sample teachers, students and school directors by region.

National Organization for Examinations 23

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 3. Sample students, teachers and directors by region

Teachers Grade4 Students Grade 8 Students

Region School
Grade4 Grade8 Collected Cleaned Collected Cleaned
directors
Tigray 25 74 60 953 951 480 479
Afar 25 42 52 658 648 385 384
Amhara 47 48 73 1721 1721 1028 1027
Oromia 76 102 222 3014 2967 1964 1957
Somali 24 77 42 694 687 310 307
Ben-Gumuz 25 42 27 776 776 448 446
SNNPR 80 21 215 3025 3004 1787 1736
Harari 24 76 96 727 723 412 412
Addis Ababa 25 28 105 987 983 825 823
Dire Dawa 25 47 94 791 791 488 488
Total 376 751 988 13346 13248 8127 8059

The above table indicates that the national sample of students who took part in the Ethiopian
Second National Learning Assessment was 13346 in Grade 4, and 8,127 in Grade 8.
Similarly, 751 Grade 4 teachers, 988 Grade 8 teachers and 375 school directors were included
in the provision of the required information for the Second National Learning Assessment.
For the qualitative study, it was decided to include school directors, teachers, Grade 8
students, parents and community representatives in focus group discussions and/or
interviews. For focus groups, it was decided to form different groups comprising 5-10
individuals at least in two school areas from each data collection route. The following table
shows the number of respondents who participated in these discussions by region.
Table 4. Participants of focus group discussions by region

Participants
Region Number of Schools
Students Teachers Parents
Tigray 22 15 6 2
Afar 15 22 14 4
Amhara 33 44 38 6
Oromia 70 59 53 6
Somali 5 9 3 1
Ben-Gumuz 26 34 22 4
SNNPR 111 109 76 15
Harari 16 10 16 2
Addis Ababa 13 15 10 2
Dire Dawa 10 9 8 2
Total 321 326 246 44

Table 4 above shows that the number of Grade 8 student participants in focus group
discussions was 321 while the number of teachers and directors was 326 and 246

National Organization for Examinations 24

The Ethiopian Second National Learning Assessment Grade 4 Report

respectively. The largest number of schools and respondents was from the SNNPR due to the
larger size included from this region.

3. 4 Instruments of Data Collection

In this study, all the previous instruments developed during the National Baseline Learning
Assessment were adopted with some modification and inclusion of other three new
instruments (Physics test and attitude survey for Grade 8, and background questionnaire for
Grade 4 students). In the achievement test, adequate number of items (60-80) was prepared
and organized in two parallel forms for each subject to be piloted. The main rationale for pre-
testing parallel tests was to replace items that require replacement from the first study. The
various types of instruments used for data collection are described below.
Achievement tests: The achievement tests for both Grades contain 40 items each (except for
reading comprehension which is 20) and cover different content areas of the respective
subject. The tests include reading comprehension, English, mathematics, and environmental
science for Grade 4 and English, mathematics, physics, biology and chemistry for Grade 8
students.
Attitude survey: This instrument was prepared for students of both Grades and it contains
two parts meant to find out the background information and attitude of students towards some
socially relevant issues. The background information deals with students’ personal
characteristics, family size and education, provision of textbooks, learning support provided,
learning and assignments at school, distance walked to school and interests in subject areas.
School checklist form: This instrument was used for recording the general conditions of the
school level, location, size, number of classrooms, school facilities, equipment, services and
the classroom conditions of grades 4 and 8.
Questionnaire for teachers: This instrument focused on general background information of
Grade 4 and 8 teachers; sex, qualification, experiences, family education, their opinions
towards the teaching profession, school management, curriculum materials and students,
teachers’ load and provision of instructional materials, and meetings made with parents and
others.
Questionnaire for school directors: This questionnaire in similar to that of the teachers and
seeks data about background information of the directors, manpower in the school, evaluation
of teachers’ performances, conducting meetings with the staff and parents, provision of
curricular materials, and improvements made on the school.
School semester result: This is a form used to collect the first semester result of sample
students from each sample school on the selected subject areas for comparing school results
and results from the Second National Learning Assessment.
Group discussion guide: This instrument was used to conduct discussion with school
teachers about the strength and weakness of the school on various issues: characteristics of
students and teachers, availability of facilities and instructional materials, the school’s
program and period allocation, organization and administration of the school, and
involvement of different groups in supporting the school and other environmental constraints
that affect the teaching learning process.

National Organization for Examinations 25

The Ethiopian Second National Learning Assessment Grade 4 Report

3.5 Validity and Reliability of Instruments

Taking into account the decentralization of the curricula, one of the methods used for the
validation of instruments involved a national workshop which brought together regional
curriculum experts, specialists from ICDR and NOE to review the extent to which the
instruments measure student learning in each region. Prior to this, all instruments mentioned
above were critically reviewed by the TWG members and test developers, and this was
followed by comments from an international consultant. In order to ensure the
representativeness of the contents of the instruments, the workshop weighted items that were
prepared on the basis of the table of specification following the syllabi of primary education.
In other words, the national workshop ensured the content validity and relevance of the test
items prepared by subject experts or the item writers by relating the items to the curriculum
objectives.
Another measure used to check the predictive validity of the Second National Learning
Assessment was collection of teacher evaluation of students from rosters. A comparison
between the composite achievement tests from the Second National Learning Assessment and
school teachers’ evaluation had shown strong correlation. Since samples were randomly
selected and they were representatives of the Grades 4 and 8 Ethiopian student population,
the internal and external validity of the assessment was ensured.
The Second National Learning Assessment was conducted in 14 nationality languages in
Ethiopia. The instruments were first developed in the English language after which they were
translated into other languages. In fact, the process of translation was made two times, first
for pre-testing and second for the final instruments. Based on the agreement made with the
Region Education Bureaus during the NAC meeting held at MOE, the pre-testing of
instruments was made in Addis Ababa, Adama, Debre Birhan and Mekele in three major
language areas (Afan Oromo, Amharic and Tigrigna). The translation of instruments was
made by subject teachers and curriculum experts who have experience in dealing with the
languages in the respective places. On translation, two persons were assigned for different
tasks, one to translate direct from the original version to the respective language and the other
to do the backward translation.
Pre-testing of instruments was made in sixteen schools of 4 regions (Amhara, Addis Ababa,
Oromia and Tigray) to identify item clarity, difficulty and the problem areas of each item.
The 4 regions were first identified by the study team at the centre. Next, the regions were
contacted for selecting 4 schools (primary and secondary) to carry out the tests on Grades 5
and 9 students (for grades 4 and 8 instruments respectively). Questionnaires or instruments
were pre-tested only in primary schools. In each school, the piloting was a two-day task by
two persons (data collectors). After the data were collected, item analyses were carried out in
order to make additional improvements on items. Before decisions were made to modify,
accept or reject items, the difficulty level, discrimination index and point bi-serial correlation
were thoroughly investigated.
Finalized instruments, in particular, the achievement tests, background questionnaires and
attitude surveys of both Grades were translated into additional local instructional languages:
Harari, Somali, Welaita, Keficho, Kembata, Hadiya, Gofa, Gamo, Dawaro, Sidama, Gedeo,
and Himitigna. At last, the translated instruments were sent for printing in booklet forms.
Two subjects were arranged in one booklet and packaged for the sample schools and the field
routes.

National Organization for Examinations 26

The Ethiopian Second National Learning Assessment Grade 4 Report

Prior to adopting instruments from the National Baseline Learning Assessment and
developing new ones for the Second National Learning Assessment, however, a thorough
item analysis was carried out for item difficulty and discrimination power. Based on the
analysis, certain items were modified or totally replaced by new ones and more new sets of
items were prepared for each subject area.
Similarly, the study team revised the instruments for background information which included
the previous attitude survey instruments for grades 4 and 8. The questionnaires for directors
and teachers and the school checklist used in the EBNLA were also improved in line with the
identified variables. A new background questionnaire for Grade 4 and an attitude survey
questionnaire for Grade 8 were also developed based on suggestions made by the
international consultant.

3.6. Data Collection and Administration of Instruments

A systematic and planned approach was used for the field work. The collection of data was
organized to take place in 4teen routes. Each route had two selected training centers in which
the training of data collectors was carried out. One route leader or trainer was assigned from
the center (NOE) to manage the activities of each route. Prior to data collection or the field
work, a consultative workshop was held with regional educational representatives to discuss
the program of data collection, the sample schools and training centers, the assignment of
centre coordinators, and recruiting data collectors in each region. In line with this, data
collectors were selected by regions from the respective Woreda education offices and from
non-sampled schools. One data collector was assigned to each Grade level (4 and 8) in each
sample school. Based on the prepared guideline by the study team, a two day intensive
training was given for data collectors by the route leader /trainer in each of the 27 centers.
After the training, the data collectors were provided with the list of sample schools,
instruments and working schedule at each school. The data collection was conducted at the
same time in all sample schools from April 15-30/2004. Table 5 below summarizes
participants of the field work by routes.

National Organization for Examinations 27

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 5. Number of participants involved in field work by route

Routes Route leaders Centre facilitators Data collectors

Mekele and Axum 1 2 43
D/Tabor and Gonder 1 2 17
D/markos and G/Beles 1 2 44
Ataye and Woldia 1 2 31
Nekemete and Assossa 1 2 45
Wolkite and Ambo 1 2 54
Bonga and Bedele 1 2 36
Shashemene and Adama 1 2 51
Sodo and Arbaminch 1 2 70
Awassa and Negele 1 2 43
Asebeteferi and Diredawa 1 2 51
Harari and Jigiga 1 2 67
Addis Ababa 1 1 46
Awash and Dubti 1 2 40
Total 14 27 638

Before students sat for the examinations, they were given a short-training on how to write or
fill in the answers. They were also sensitized on the goal of the Ethiopian Second National
Learning Assessment and how significant their participations could be for the successful
accomplishment of the task. In order to reduce boredom in the course of taking long
examinations, a limited form of motivation was administered to students. At the end or return
from the field, reports were made by various groups; the field workers to the route leader, the
centre coordinators to the respective region, and the route leaders to the centre or the study
team. Data for the qualitative analysis were collected by center coordinators. There were
center representatives from region and zone education bureaux who assisted in discussions,
selecting participants and in translations.

3.7 Methods of Data Analysis

Two separate data files were created for grades 4 and 8, and the data were organized into
these data bases at school and student levels. Before encoding the collected data into the
computer, the instruments were first organized by region, type of instruments, grade level,
subject area, school and respondents’ code number. The data were first entered to access and
then transformed to SPSS. Twenty five encoders from different departments of NOE were
involved in entering the data into the computer. Parallel to this, other 20 persons were
assigned to check and re-check the encoded printouts of each instrument. The encoding and
checking task took beyond the expected time due to micro-management problems. The
encoding was followed by the process of data cleaning, analysis and interpretation. The
common statistical procedures applied for data analysis included descriptive statistics, t-tests,
ANOVA, homogeneity tests, correlation and regression analyses.

National Organization for Examinations 28

The Ethiopian Second National Learning Assessment Grade 4 Report

CHAPTER 4: FINDINGS OF THE STUDY

4.1 Findings on Grade 4 Achievement Tests

The following section presents the analysis on the 4 achievement tests given to Grade 4
students across the nation. The tests were reading comprehension, English, mathematics, and
environmental science. The composite average score of the 4 tests was also analyzed
separately. The raw scores of each subject were converted into percentages. Each test was
analyzed primarily at national and regional levels and then by gender of the students, and the
location of the schools. The analysis of each test is accompanied by up to seven tables which
show summary of the descriptive statistics, one-way analysis of variance and homogeneity
subsets grouping whenever appropriate.

4.1.1 Overall Achievement of the Students at National Level

The overall achievement of the students at national level was presented initially prior to
detailed discussion at different levels of analysis. Table 6 shows the summary of the
descriptive statistics for all the subjects and the composite mean score at national level. The
overall performances in all subjects were low with only reading comprehension and
environmental science average reaching 50%. The mean percent score for mathematics
(39.70) and English (38.68%) are notably lower than the national mean (48.48%) and also
from the mean of the other two subjects.
Table 6. Summary of the descriptive statistics for all subjects at national level

Subjects N Mean Std. Error Std. Dev

Reading 12886 64.49 .19 21.97
English 13028 38.68 .15 16.88
Mathematics 13143 39.70 .15 17.51
Env. Science 13005 51.74 .15 17.42
Composite 12507 48.48 .13 15.07
Seen from the perspectives of the Ethiopian Education and Training policy, the overall
achievement of Grade 4 students is less than the minimum expectations. Particularly, the
mean percentage scores for Mathematics and English are by far less than the minimum
requirements of 50%. Ironically, the two subjects are considered as core subjects in the
primary school curricula. The fact that test scores are less than expectations implies that the
quality of schooling is low in Ethiopia. It is now clear from good amount of evidence that the
quality of schooling has an influence upon the speed with which societies can become richer
and the extent to which individuals can improve their own productivity and incomes
(UNESCO, 2004). Thus, low quality schooling in Ethiopia also implies immense influence
on the country’s efforts to poverty reduction and national prosperity.

4.1.2 Overall Achievement of the Students by Gender at National Level

Table 7 shows that the mean differences between males and females in all subjects were
statistically significant in favor of males. The mean difference in mathematics was the
highest. This is an indication that gender gap still exists.

National Organization for Examinations 29

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 7. T-test for equality of means by gender at national level

Std. Error Std.
Subjects Gender N Mean t Sig. MD
Mean Deviation
Reading MALE 7547 65.71 .25 21.78
6.858 0.000 2.71
FEMALE 5221 63.00 .31 22.18
English MALE 7628 40.55 .20 17.39
16.398 0.000 4.90
FEMALE 5282 35.66 .21 15.59
Mathematics MALE 7689 41.68 .21 18.15
16.244 0.000 5.02
FEMALE 5350 36.67 .22 16.13
Env. Science MALE 7590 53.43 .21 17.89
12.376 0.000 3.82
FEMALE 5313 49.61 .22 16.32
Composite MALE 7324 50.14 .18 15.43
14.690 0.000 4.02
FEMALE 5082 46.12 .20 14.31

The above findings in Table 7 are not surprising as many studies in developing countries
show similar trends. For instance, Duncan (1989) refers to unpublished IEA studies which
indicate that girls performed less than boys in science in three African countries namely,
Ghana, Nigeria and Zimbabwe. She further refers to a Zambian study where examination
results show that girls scored considerably below boys in all 4 subjects examined (English,
social studies, mathematics and science). These results even concur with previous findings in
Ethiopia for primary science and mathematics by Biazen and Junge (1988). Reasons and
appropriate solutions for such disparity have been related to numerous factors including
socio-economic, cultural, and institutional/policy, and school situations (see Odaga and
Heneveld, 1995). Very recently, Kellagan and Greany (2004) referred to studies from MLA
and SAQMEC projects in Africa where significant gender differences were not found in
academic performance among female and male students. This shows that it is possible to
overcome the gap in female and male student academic performances provided that the right
approaches are adopted in Ethiopia.

4.1.3 Overall Achievement of the Students by Location at National Level

Table 8 shows that the mean differences between students from urban and rural schools in the
composite score, reading comprehension and mathematics were statistically significant in
favor of urban schools. The mean difference in reading comprehension (10.87) was the
highest. Though not statistically significant, the mean difference in environmental science
was in favor of rural schools.
Table 8. T-test for equality of means by location at national level
Std.
Subjects Location N Mean Std. Error t Sig. MD
Deviation
Reading Urban 4329 67.44 .32 20.73
10.871 0.000 4.43
Rural 8557 63.01 .24 22.42
English Urban 4365 38.87 .25 16.84
.907 0.364 .28
Rural 8663 38.58 .18 16.90
Mathematics Urban 4427 40.38 .26 17.28
3.165 0.002 1.02
Rural 8716 39.35 .19 17.62
Env. Science Urban 4304 51.49 .25 16.64
-1.176 0.240 -.38
Rural 8701 51.87 .19 17.79
Composite Urban 4149 49.28 .22 14.41
4.165 0.000 1.19
Rural 8358 48.09 .17 15.37
Table 8 above indicates that the Ethiopian first cycle primary schooling is characterized not
only by low level of academic achievement, but also the distribution of achievement levels
favors urban schools over their rural counter parts. This finding is consistent with many

National Organization for Examinations 30

The Ethiopian Second National Learning Assessment Grade 4 Report

studies. For instance, Fuller (1986) reports several studies from developing countries
whereby rural school students perform less than their urban counter parts. For Ethiopia which
has been working for equality of educational opportunity between urban and rural schools
over the last decade, the current finding is an alarm that primary school student performance
in rural first cycle primary education requires improvement not only in comparison to urban
primary education, but also in relation to the national standards. The above table also
indicates that the highest difference between rural and urban schools was in reading
comprehension. Given the key role of reading comprehension for learning other subjects and
the fact that most primary schools are rural in Ethiopia, the finding suggests that
strengthening the reading ability of rural students is a precondition for improved quality of
primary schooling at the lower level for the nation.

4.2 Overall Achievement of the Students by Subject

4.2.1 Reading Comprehension
Table 9 shows the summary of the descriptive statistics for reading comprehension. A total of
12,886 students took the test. The mean score at national level was 64.5 with a standard
deviation of 21.97. It can be seen in the table that overall performances were in the minimum
acceptable range with all regions surpassing a 50% correct, on the average. Addis Ababa,
Afar and Amhara scored well above the national mean 76.90, 71.92, and 70.92 respectively.
The common characteristic of these three regions was that they took the test in Amharic
which is a language that has been used as medium of instruction for long time. Somali scored
the least with mean score of 50.53. In comparison to those three regions which scored
highest, the introduction of Somali as a medium of instruction is recent.
Table 9. Summary of the descriptive statistics for reading comprehension by region
REGIONS N MEAN SD
TIGRAY 947 63.37 16.77
AFAR 643 71.92 16.20
AMHARA 1641 70.92 22.08
OROMIA 2773 63.73 24.07
SOMALI 672 50.53 20.07
BEN-GUMUZ 771 60.34 21.89
SNNPR 2980 60.31 21.28
HARARI 722 68.60 19.61
ADDIS ABABA 946 76.90 17.19
DIRE DAWA 791 62.24 21.61
NATION 12886 64.49 21.97

One-way analysis of variance (Table 10) showed that there was a statistically significant
mean difference between regions for reading comprehension (F = 115.481, p < 0.001).
Multiple comparisons using Scheffe Post Hoc Test was conducted to identify homogenous
subset groupings across regions.
Table 10. One- way analysis of variance on reading comprehension by region
Sum of Mean
Df F Sig.
Squares Square
Between Groups 464413.18 9 51601.46 115.481 0.00
Within Groups 5753481.65 12876 446.84
Total 6217894.8 12885

National Organization for Examinations 31

The Ethiopian Second National Learning Assessment Grade 4 Report

The homogenous subset grouping for reading comprehension resulted in 4 distinct groups
(Table 11).The mean difference between the groups was statistically significant but members
of each group did not differ significantly from each other. The highest performing region was
Addis Ababa followed by Afar, Amhara and Harari. Somali’s performance was way behind
and all the other regions fell in group two.
Table 11. Homogenous subset groups for reading comprehension

Subset for alpha = .05

Region Number
1 2 3 4
SOMALI 672 50.53
SNNPR 2980 60.31
BEN-GUMUZ 771 60.34
DIRE DAWA 791 62.24
TIGRAY 947 63.37
OROMIA 2773 63.73
HARARI 722 68.60
AMHARA 1641 70.92
AFAR 643 71.92
ADDIS ABABA 946 76.90
Sig. 1 0.182 0.222 1
Table12 shows the independent t-test for equality of mean scores for reading comprehension
between males and females across regions. There are statistically significant mean differences
in Tigray, Amhara, Oromia, and SNNPR in favor of males. Though not statistically
significant, females in Afar, Harari and Addis Ababa performed better than males. The
highest score for females was in Addis Ababa (77.04) and the least in Somali (49.45).
Table 12. T-test for equality of means for reading comprehension by gender within regions
Region Gender N Mean Std. t Sig. Mean
M 449 66.41 15.85 5.375 0.00 5.78
TIGRAY
F 498 60.63 17.12
M 401 71.32 16.38 -1.208 0.228 -1.59
AFAR
F 242 72.91 15.87
M 904 72.52 22.92 3.257 0.001 3.57
AMHARA
F 733 68.95 20.88
M 1719 66.31 23.45 7.267 0.00 6.78
OROMIA
F 1054 59.53 24.50
M 426 50.50 20.72 0.517 0.605 1.06
SOMALI
F 136 49.45 20.84
M 476 60.43 21.90 0.14 0.889 0.23
BENISHA/GUMUZ
F 295 60.20 21.92
M 1809 62.79 20.57 7.962 0.00 6.29
SNNPR
F 1169 56.49 21.80
M 425 68.36 19.46 -0.35 0.726 -0.52
HARARI
F 296 68.89 19.85
M 433 76.74 17.68 -0.261 0.794 -0.29
ADDIS ABABA
F 513 77.04 16.79
M 505 62.97 21.25 1.274 0.203 2.04
DIRE DAWA
F 285 60.93 22.25
M 7547 65.71 21.78
NATION 0.000 .000 2.71
F 5221 62.99 22.18

National Organization for Examinations 32

The Ethiopian Second National Learning Assessment Grade 4 Report

The mean difference in reading comprehension between urban and rural schools was
statistically significant in Amhara, Oromia, Somali, Harari, and Dire Dawa (Table 13). In all
these regions except Oromia, urban schools performed better than rural ones. Though not
statistically significant, rural schools performed better in Tigray and Ben-Gumuz than urban
schools.
Table 13. T - test for equality of means for reading comprehension by location within regions

Mean
Region Location N Mean Std. T Sig. Difference
U 396 62.16 15.36 -1.892
0.059 -2.09
TIGRAY R 551 64.25 17.67
U 369 72.06 14.16 0.252
0.801 0.33
AFAR R 274 71.73 18.61
U 283 75.72 14.77 4.043
0.000 5.81
AMHARA R 1358 69.92 23.19
U 548 60.21 26.85 -3.833
0.000 -4.39
OROMIA R 2225 64.60 23.26
U 399 54.67 18.72 6.681
0.000 10.21
SOMALI R 273 44.47 20.47
BEN- U 277 60.11 21.08 -0.223
0.823 -0.37
GUMUZ R 494 60.48 22.35
U 298 61.06 19.76 0.639
0.523 0.83
SNNPR R 2682 60.23 21.44
U 428 74.05 18.16 9.565
0.000 13.39
HARARI R 294 60.66 18.94
DIRE U 385 65.48 21.32 4.151
0.000 6.32
DAWA R 406 59.16 21.46
U 4329 67.44 20.73 10.871
NATION 0.000 4.43
R 8557 63.01 22.42
4.2.2 English
Table 14 shows the summary of the descriptive statistics for English. A total of 13,028
students took the test. The mean score at national level was 38.68 with a standard deviation
of 16.88. Somali scored the highest (50.36) and Beninshangul Gumuz and Tigray scored the
least with mean scores of 32.44 and 32.95 respectively. The overall performance in all
regions was very low for English with only Somali reaching 50% correct, on the average.
Table 14. Summary of the descriptive statistics for English by region
Regions N Mean SD
TIGRAY 946 32.95 12.78
AFAR 641 35.30 13.75
AMHARA 1654 37.87 17.49
OROMIA 2962 39.49 17.94
SOMALI 670 50.36 20.98
BEN-GUMUZ 771 32.44 13.65
SNNPR 2928 38.93 15.28
HARARI 721 42.46 16.74
ADDIS ABABA 945 39.94 17.33
DIRE DAWA 790 37.19 15.75
NATION 13028 38.68 16.88

National Organization for Examinations 33

The Ethiopian Second National Learning Assessment Grade 4 Report

One-way analysis of variance (Table 15) shows that there is a statistically significant
difference in the mean score for reading comprehension between regions (F = 72.204, p <
.001). Multiple comparisons using Scheffe Post Hoc Test was conducted to identify
homogenous groups across regions.
Table 15. One way analysis of variance of English across regions

Sum of Mean
Squares Df Square F Sig.
Between Groups 176544.53 9 19616.06 72.204 0.000
Within Groups 3536653.48 13018 271.67
Total 3713198 13027
Table 16 shows that there are five homogenous subset groups. The highest performing group
contains Somali followed by Harari, Addis Ababa and Oromia. From the remaining ones
Beninshangul Gumuz, Tigray and Afar performed the least and fell under Group 1. When the
alpha level is reduced from .05 to .01, the result of multiple comparison table shows only two
distinguished groups. In this case, the five lower performing regions were significantly
different from Somali Region.
Table 16. Homogeneous subset groups of English by region

Subset for alpha = .05

Regions N 1 2 3 4 5
BEN-GUMUZ 771 32.44
TIGRAY 946 32.95
AFAR 641 35.30 35.30
DIRE DAWA 790 37.19 37.19
AMHARA 1654 37.87 37.87
SNNPR 2928 38.93
OROMIA 2962 39.49 39.49
ADDIS ABABA 945 39.94 39.94
HARARI 721 42.46
SOMALI 670 50.36
Sig. 0.108 0.229 0.145 0.074 1
The mean difference between males and females in English was statistically significant in
Tigray, Afar, Amhara, Oromia, Benshangul Gumuz, SNNPR, and Dire Dawa in favor of
males. Though not statistically significant females in Somali and Harari performed better
than males, and males in Addis Ababa performed better than females (Table 17).

National Organization for Examinations 34

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 17. T-test for equality of means for English by gender within regions

Mean
t Sig.
Regions Gender N Mean Std. Difference
M 449 35.12 13.40
5.008 0.000 4.12
TIGRAY F 497 31.00 11.88
M 400 36.49 13.59
2.841 0.005 3.17
AFAR F 241 33.32 13.82
M 905 40.41 18.82
6.511 0.000 5.57
AMHARA F 745 34.84 15.23
M 1824 42.04 18.42
9.99 0.000 6.66
OROMIA F 1138 35.38 16.33
M 425 49.11 22.20
-1.099 0.272 -2.38
SOMALI F 136 51.49 21.41
M 476 33.67 14.21
3.197 0.001 3.21
BEN-GUMUZ F 295 30.46 12.46
M 1788 41.37 15.55
10.993 0.000 6.25
SNNPR F 1137 35.12 14.04
M 424 42.35 16.66
-0.196 0.845 -0.25
HARARI F 296 42.60 16.92
M 432 40.62 18.20
1.107 0.269 1.25
ADDIS ABABA F 513 39.37 16.56
M 505 38.34 16.32
DIRE DAWA 2.796 0.005 3.25
F 284 35.09 14.49
M 7628 40.55 17.39
NATION 16.398 0.000 4.89
F 5282 35.66 15.59

The mean difference between urban and rural schools in English was statistically significant
in Tigray, Afar and Harari (Table 18). In Tigray and Afar, rural schools performed better than
urban ones while in Harari, urban schools performed better than rural ones. The trend shows
that there was no big gap between urban and rural schools.

National Organization for Examinations 35

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 18. T-test for equality of means in English by location within regions

Std. t Sig. Mean

RegionsLocation N Mean Deviation (2-tailed) Difference
U 396 30.13 11.92
-5.876 0.000 -4.86
TIGRAY R 550 34.99 13.00
U 370 33.62 13.09
-3.638 0.000 -3.96
AFAR R 271 37.58 14.31
U 284 36.92 14.13
-1.008 0.313 -1.15
AMHARA R 1370 38.07 18.11
U 588 40.25 19.16
1.156 0.248 0.96
OROMIA R 2374 39.30 17.62
U 397 50.23 19.12
-0.192 0.848 -0.32
SOMALI R 273 50.55 23.46
U 277 32.59 14.09
0.226 0.821 0.23
BEN-GUMUZ R 494 32.36 13.41
U 296 38.23 12.94
-0.827 0.408 -0.77
SNNPR R 2632 39.01 15.52
U 428 44.83 16.64
4.656 0.000 5.83
HARARI R 293 39.00 16.31
U 384 36.22 13.66
-1.684 0.093 -1.89
DIRE DAWA R 406 38.10 17.47
U 4365 38.87 16.84
NATION .907 .364 .28
R 8663 38.58 16.90

4.2.3 Mathematics
Table 19 shows the summary of the descriptive statistics for mathematics. A total of 13,143
students took the test. The mean score at national level was 39.70 with a standard deviation of
17.51. Addis Ababa scored the highest (44.92). Overall performances were very low in
mathematics, with no region reaching a 50% correct level on the average.
Table 19. Summary of the descriptive statistics for mathematics by region
Regions N Mean SD
TIGRAY 945 34.50 13.92
AFAR 643 40.29 16.80
AMHARA 1711 43.27 20.13
OROMIA 2932 38.30 17.33
SOMALI 670 43.09 18.98
BEN-GUMUZ 771 34.34 15.00
SNNPR 2986 39.70 16.87
HARARI 719 40.71 16.47
ADDIS ABABA 981 44.92 18.28
DIRE DAWA 785 37.83 15.77
NATION 13143 39.70 17.51

National Organization for Examinations 36

The Ethiopian Second National Learning Assessment Grade 4 Report

One-way analysis of variance (see Table 20 below) shows that there was a statistically
significant difference in the mean score for reading comprehension between regions
(F = 42.22, p < .001). Multiple comparisons using Scheffe Post Hoc Test was conducted to
identify homogenous groups across regions. The following table shows the results.
Table 20. One way analysis of variance on mathematics by region

Sum of Squares Df Mean Square F Sig.

Between Groups 113367.12 9 12596.35 42.221 0.00
Within Groups 3918117.3 13133 298.34
Total 4031484.4 13142

Table 21 below shows that there are 4 groups. Addis Ababa, Amhara and Somali fell in the
upper most group followed by Afar and Harari. Benshangul Gumuz and Tigray fell in the
lowest group.
Table 21. Homogenous subset groups of mathematics by region
Subset for alpha = .05
Region N 1 2 3 4
BEN-GUMUZ 771 34.34
TIGRAY 945 34.50
DIRE DAWA 785 37.83
OROMIA 2932 38.30
SNNPR 2986 39.70
AFAR 643 40.29 40.29
HARARI 719 40.71 40.71
SOMALI 670 43.09 43.09
AMHARA 1711 43.27 43.27
ADDIS ABABA 981 44.92
Sig. 1 0.143 0.109 0.796

The mean difference in mathematics between males and females in Tigray, Afar, Amhara,
Oromia, Benshangul Gumuz, SNNPR and Dire Dawa was statistically significant in favor of
males (see Table 22). Though not statistically significant, females in Somali and Harari
performed better than males.

National Organization for Examinations 37

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 22. T-test of mean difference for mathematics by gender within regions

Std. Sig. Mean

Region Gender N Mean Deviation t (2-tailed) Difference
M 449 35.12 13.40
5.008 0.000 4.12
TIGRAY F 497 31.00 11.88
M 400 36.49 13.59
2.841 0.005 3.17
AFAR F 241 33.32 13.82
M 905 40.41 18.82
6.511 0.000 5.57
AMHARA F 745 34.84 15.23
M 1824 42.04 18.42
9.99 0.000 6.66
OROMIA F 1138 35.38 16.33
M 425 49.11 22.20
-1.099 0.272 -2.38
SOMALI F 136 51.49 21.41
M 476 33.67 14.21
3.197 0.001 3.21
BEN-GUMUZ F 295 30.46 12.46
M 1788 41.37 15.55
10.993 0.000 6.25
SNNPR F 1137 35.12 14.04
M 424 42.35 16.66
-0.196 0.845 -0.25
HARARI F 296 42.60 16.92
M 432 40.62 18.20
1.107 0.269 1.25
ADDIS ABABA F 513 39.37 16.56
M 505 38.34 16.32
2.796 0.005 3.25
DIRE DAWA F 284 35.09 14.49
M 7689 41.68 18.15
Nation 16.24 0.000 5.02
F 5350 36.67 16.13

The mean difference of mathematics between urban and rural schools was statistically
significant in Tigray, Afar, Amhara, Somali, SNNPR, Harari and Dire Dawa (Table 23). In
Tigray, Afar, Amhara and SNNPR, rural schools performed better than urban ones. In
Somali, Harari and Dire Dawa, urban schools performed better than rural schools.

National Organization for Examinations 38

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 23. T-test of mean difference for mathematics by location within regions

Std. Sig. Mean

Region Location N Mean Deviation t (2-tailed) Difference
U 397 32.66 12.92 -3.18
-3.483 0.001
TIGRAY R 548 35.83 14.47
U 367 38.48 15.88 -3.164 0.002 -4.21
AFAR R 276 42.69 17.70
U 317 40.55 16.48 -2.673 0.008 -3.34
AMHARA R 1394 43.89 20.82
U 581 38.69 17.99 0.602 0.547 0.48
OROMIA R 2351 38.20 17.17
U 396 45.63 19.63 4.217 0.000 6.21
SOMALI R 274 39.42 17.40
U 277 34.78 15.07 0.6 0.549 0.68
BEN-GUMUZ R 494 34.10 14.98
U 300 34.44 13.07 -5.724 0.000 -5.85
SNNPR R 2686 40.28 17.14
U 429 44.64 16.80 8.128 0.000 9.74
HARARI R 290 34.90 14.10
U 382 39.47 15.42 2.85 0.004 3.20
DIRE DAWA R 403 36.27 15.96
U 4427 40.38 17.28
NATION 3.165 0.002 1.02
R 8716 39.35 17.62

4.2.4 Environmental Science

Table 24 shows the summary of the descriptive statistics for environmental science. A total of
13005 students took the test. The mean score at national level was 51.74 with a standard
deviation of 17.42. Amhara scored the highest (60.05), well above the national mean. The
overall regional performances in environmental science were average, with most regions
approximating 50% correct, on the average.
Table 24. Summary of the descriptive statistics for environmental science by region
Regions N Mean SD
TIGRAY 938 45.87 11.88
AFAR 604 51.60 13.77
AMHARA 1708 60.05 17.19
OROMIA 2932 50.73 18.63
SOMALI 632 45.83 22.30
BEN-GUMUZ 768 47.84 16.75
SNNPR 2977 49.75 16.14
HARARI 720 53.35 16.86
ADDIS ABABA 929 54.64 14.02
DIRE DAWA 789 47.86 16.17
NATION 13005 51.74 17.42

National Organization for Examinations 39

The Ethiopian Second National Learning Assessment Grade 4 Report

One-way analysis of variance (Table 25) shows that there is a statistically significant
difference in the mean score for reading comprehension between regions
(F = 72.44, p < .001). Multiple comparisons using Scheffe Post Hoc Test was conducted to
identify homogenous groups across regions.
Table 25. One way analysis of variance on environmental science by region

Sum of Squares Df Mean Square F Sig.

Between Groups 188474.64 9 20941.63 72.435 0.00
Within Groups 3756994.7 12995 289.11
Total 3945469.3 13004

Table 26 shows that there were six groups. Amhara was in the upper most group followed by
Addis Ababa, Harari, Tigray, and Afar. Somali, Beninshangul Gumuz, and Dire Dawa fell in
the lowest group. When the alpha level was reduced to .01, the result of multiple comparison
tests showed Somali as performing significantly lower than the higher performing region of
Amhara.
Table 26. Homogenous subset groups of environmental science by region

Subset for alpha = .05

Region N 1 2 3 4 5 6
SOMALI 632 45.83
BEN-GUMUZ 768 47.84 47.84
DIRE DAWA 789 47.86 47.86
SNNPR 2977 49.75 49.75
OROMIA 2932 50.73 50.73 50.73
AFAR 604 51.60 51.60 51.60
TIGRAY 946 52.50 52.50 52.50
HARARI 720 53.35 53.35
ADDIS ABABA 929 54.64
AMHARA 1708 60.05
Sig. 0.668 0.134 0.195 0.259 0.087 1

The mean difference in environmental science between males and females was statistically
significant in Tigray, Amhara, Oromia, SNNPR, and Dire Dawa in favor of males (see Table
27). Though not statistically significant, females performed better than males in Somali and
Harari.

National Organization for Examinations 40

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 27. T -test of mean difference for environmental science by gender within regions

Std. t Sig. Mean

Region Gender N Mean Dev. (2-tailed) Difference
M 448 55.80 15.70
6.377 0.000 6.26
TIGRAY F 498 49.54 14.49
M 375 52.22 14.84
1.423 0.155 1.64
AFAR F 229 50.58 11.77
M 938 63.24 17.47
8.669 0.000 7.11
AMHARA F 766 56.12 16.05
M 1804 53.77 18.84
11.419 0.000 7.90
OROMIA F 1128 45.87 17.22
M 410 46.91 22.69
-1.268 0.205 -2.91
SOMALI F 126 49.82 21.93
M 473 47.90 17.47
0.139 0.889 0.17
BENISHA/GUMUZ F 295 47.73 15.56
M 1804 51.62 16.18
7.881 0.000 4.72
SNNPR F 1172 46.89 15.65
M 425 52.55 16.91
-1.544 0.123 -1.97
HARARI F 295 54.52 16.74
M 409 54.91 15.04
0.513 0.608 0.48
ADDIS ABABA F 520 54.43 13.17
M 504 49.27 15.95
3.27 0.001 3.90
DIRE DAWA F 284 45.37 16.32
M 7590 53.43 17.89
NATION 12.38 0.000 3.82
F 5313 49.61 16.33

The mean difference in environmental science between urban and rural schools was
statistically significant in Tigray, Afar, Amhara, Somali, and SNNPR (see Table 28). In
Tigray, Afar, Amhara and SNNPR rural schools performed better than urban ones. In Somali
and Harari, urban schools performed better than rural schools.

National Organization for Examinations 41

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 28. T-test for mean difference in environmental science by location within regions

Std. Sig. Mean

Region Location N Mean Dev. T (2-tailed) Difference
U 397 47.39 14.42
-9.061 0.000 -8.82
TIGRAY R 549 56.20 15.01
U 330 49.43 12.04
-4.304 0.000 -4.77
AFAR R 274 54.21 15.22
U 316 57.48 14.38
-2.942 0.000 -3.14
AMHARA R 1392 60.63 17.72
U 583 51.40 19.62
0.973 0.33 0.84
OROMIA R 2349 50.56 18.38
U 358 48.97 23.54
4.094 0.000 7.24
SOMALI R 274 41.73 19.89
U 277 47.70 15.64
-0.17 0.865 -0.21
BEN-GUMUZ R 491 47.91 17.36
U 300 45.21 14.14
-5.167 0.000 -5.06
SNNPR R 2677 50.26 16.27
U 429 57.60 16.27
8.626 0.000 10.52
HARARI R 291 47.09 15.73
U 385 48.19 14.84
0.565 0.572 0.65
DIRE DAWA R 404 47.54 17.36
U 4304 51.49 16.64
NATION -1.176 0.240 -.38
R 8701 51.87 17.79

4.2.5 Composite Score

Table 29 shows the summary of the descriptive statistics for the composite score. A total of
12, 507 students had mean composite score. Only those students who sat for all the 4 tests
had mean composite score. The mean score at national level was 48.48 % with a standard
deviation of 15.07. Addis Ababa, Amhara, and Harari scored well above the national mean
53.69%, 52.67%, and 51.29% respectively. Benshangul Gumuz scored the least, 43.69%.

Table 29. Summary of the descriptive statistics for composite mean by region

Regions N Mean SD
TIGRAY 938 45.87 11.88
AFAR 594 49.64 12.01
AMHARA 1597 52.67 15.86
OROMIA 2724 47.83 16.59
SOMALI 613 47.18 15.80
BEN-GUMUZ 768 43.69 13.67
SNNPR 2886 47.27 14.40
HARARI 716 51.29 14.53
ADDIS ABABA 888 53.69 13.65
DIRE DAWA 783 46.20 14.58
NATION 12507 48.48 15.07

National Organization for Examinations 42

The Ethiopian Second National Learning Assessment Grade 4 Report

One-way analysis of variance (see Table 30 below) shows that there was statistically
significant difference in the mean composite score between regions in the mean composite
score (F = 47.07 at p < .001). Multiple comparisons using Scheffe Post Hoc Test was applied
to identify homogenous groupings across regions.
Table 30. One way analysis of variance on composite mean by region

Sum of Squares df Mean Square F Sig.

Between Groups 93063.104 9 10340.35 47.07 0.000
Within Groups 2745642.9 12497 219.70
Total 2838706 12506
Table 31 shows that there were five groups. Addis Ababa, Amhara and Harari fell in the
upper most group. Benshangul Gumuz, Tigray and Dire Dawa fell in the lowest group.
Table 31. Homogenous subset groups of mean composite score by region

Subset for alpha = .05

Region N 1 2 3 4 5
BEN-GUMUZ 768 43.69
TIGRAY 938 45.87 45.87
DIRE DAWA 783 46.20 46.20
SOMALI 613 47.18 47.18
SNNPR 2886 47.27 47.27
OROMIA 2724 47.83 47.83
AFAR 594 49.64 49.64
HARARI 716 51.29 51.29
AMHARA 1597 52.67
ADDIS ABABA 888 53.69
Sig. 0.155 0.521 0.174 0.767 0.21
The mean difference in composite score between males and females was statistically
significant in Tigray, Amhara, Oromia, SNNPR, Benshangul Gumuz and Dire Dawa in favor
of males (Table 32).

National Organization for Examinations 43

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 32. T -test of mean difference for composite mean by gender within regions

Std. T Sig. Mean

Region Gender N Mean Dev (2-tailed) Difference
M 446 48.82 11.93
7.46 0.000 5.63
TIGRAY F 492 43.19 11.20
M 368 50.31 12.58
1.74 0.082 1.76
AFAR F 226 48.55 10.96
M 876 55.41 16.39
7.77 0.000 6.10
AMHARA F 717 49.32 14.56
M 1693 50.37 16.89
10.42 0.000 6.70
OROMIA F 1031 43.67 15.20
M 398 46.60 16.63
-1.37 0.172 -2.39
SOMALI F 120 48.99 17.27
M 473 44.49 14.27
2.03 0.042 2.10
BEN-GUMUZ F 295 42.43 12.57
M 1759 49.62 14.48
11.18 0.000 6.02
SNNPR F 1126 43.60 13.50
M 421 51.05 14.43
-0.54 0.587 -0.60
HARARI F 295 51.65 14.69
M 390 54.09 14.25
0.77 0.440 0.71
ADDIS ABABA F 498 53.37 13.16
M 500 47.30 14.59
2.83 0.005 3.10
DIRE DAWA F 282 44.24 14.39
M 7324 50.13 15.43
Nation 14.69 0.000 4.02
F 5082 46.12 14.31

National Organization for Examinations 44

The Ethiopian Second National Learning Assessment Grade 4 Report

The mean difference in composite score between urban and rural schools was statistically
significant in Tigray, Afar, Somali, SNNPR, Harari, and Dire Dawa (Table 33). In Tigray,
Afar, and SNNPR rural schools performed better than urban schools. In Somali, Harari and
Dire Dawa urban schools performed better.
Table 33. T-test of mean difference for composite mean by location within regions

Std. T Sig. Mean

Region Location N Mean Dev (2-tailed) Difference
U 392 43.10 10.52
-6.16 0.000 -4.76
TIGRAY R 546 47.86 12.40
U 327 48.09 10.45
-3.52 0.000 -3.45
AFAR R 267 51.54 13.47
U 279 52.44 12.29
-0.26 0.792 -0.28
AMHARA R 1318 52.72 16.52
U 540 47.30 17.88
-0.84 0.403 -0.67
OROMIA R 2184 47.97 16.25
U 342 49.72 16.89
4.54 0.000 5.75
SOMALI R 271 43.97 13.68
U 277 43.79 13.07
0.15 0.88 0.15
BEN-GUMUZ R 491 43.64 14.02
U 295 44.72 11.25
-3.21 0.001 -2.83
SNNPR R 2591 47.56 14.69
U 428 55.28 14.17
9.49 0.000 9.91
HARARI R 288 45.37 12.97
U 381 47.26 13.16
2.00 0.046 2.08
DIRE DAWA R 402 45.19 15.75
U 4149 49.28 14.41
NATION 4.17 0.000 0.29
R 8358 48.09

National Organization for Examinations 45

The Ethiopian Second National Learning Assessment Grade 4 Report

4.3 Grade 4 Pupils’ Background and Academic Achievement

A student background questionnaire was used as an instrument to gather the student

background information. It contained a number of different items that helped the study team
to collect general information about distance to school, watching television, listening to the
radio, doing home work, interests in school subjects, etc. Findings of the study substantiate
but sometimes challenge the conventional wisdom concerning the relationships between
student background characteristics and academic achievement. For example, parents’ level of
education, tutorial support, teachers’ qualification, teachers’ teaching experiences and many
other characteristics had either negative or no relationships with students’ academic
achievement. Some detailed analyses of such relationships, therefore, are indicated in this
part of the study.
Table 34. Time needed to reach school and academic achievement.
Composite Std. % of
N
Mean Deviation Total N
up to 60 min. 10370 49.21 .1471 85.6%
Above 60 min. 1745 45.72 .1537 14.4%
Total 12115 48.70 .1485 100.0%

As Table 34 summarizes, students who travel up to 60 minutes scored a better mean (49) in
the Grade 4 composite mean score than those who travel above 60 minutes (45.72). This
implies that the longer the distance a student traveled from home to school the lower the
score she/he achieved in academic subjects. This finding is not surprising as it is obvious that
students who travel long distances spend their study time on the way. In fact upon arriving at
schools, they may find it difficult to participate properly and attend class lessons as they are
already tired.

Table 35. ANOVA on time required to reach school

Source of variations SS Df MS F Sig.

Between Groups 1.819 1 1.819 82.971 0.001
Within Groups 265.495 12113 .02192
Total 267.313 12114

In Table 35, it is indicated that the F-ratio calculated on the observed mean difference
between the two groups (F (1, 12113) = 82.971, P < .001) was statistically significant. This
implies that distance between home and school can influence Grade 4 pupils’ academic
achievement in Ethiopia. Moreover, the study identified the correlation between Grade 4
composite mean and distance from home to school (r = -.082) to be significant at the 0.01
level (2-tailed).

National Organization for Examinations 46

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 36. Watching non-educational television and student achievement

Number of Days Grade 4
Std. % of Total
Devoted to Watching Composite N
Deviation N
TV Mean
Not at all 50.94 6348 .1478 55.6%
1-2 days a week 48.68 1977 .1404 17.3%
3-4 days a week 44.84 801 .1400 7.0%
5-7 days a week 44.65 2284 .1474 20.0%
Total 48.86 11410 .1483 100.0%
Students were asked how frequently they watch television programs per week. The data
indicated in Table 36 show that 55.6% of the sampled students selected the “Not at all”
alternative, while 20% of them said they watch TV programs for 5 to 7 days a week. The
finding regarding this item is negatively correlated (r = -.177) with the Grade 4 composite
achievement. This correlation is significant at the 0.01 level (2-tailed) implying that as the
number of days for watching TV per week increases, the level of students’ performance is
likely to decrease (see Table 36 above). The negative relationship between TV viewing and
student achievement is well recorded in many studies. As indicated in the review of literature
in this report, it has been found out that TV viewing for leisure, as the other student
characteristics, is inversely associated with student achievement. Obviously, spending time in
front of a television screen takes away the time a student may use for studying or doing home
work.
Table 37. Listening to the radio (other than educational radio) per week
Number of Days Devoted
Composi Std. % of Total
to N
te Mean Deviation N
Listening to the Radio.
Not at all 49.22 2743 .1488 23.5%
1-2 days a week 50.21 2539 .1419 21.8%
3-4 days a week 49.99 1964 .1508 16.8%
5-6 days a week 47.30 4423 .1479 37.9%
Total 48.84 11669 .1479 100.0%

Students were also asked to rate how frequently they listen to radio programs per week at
home. Out of the total number, 23.5% of the sampled students replied that they did not waste
any single day listening to the radio; nevertheless, their achievement level (49.22%) was
lower than the score (50.21%) of those students who reported that they devote 1-2 days per
week to listening to the radio. In general, when the findings summarized in Table 37 are
related to performance, for every group except the “Not at all” option raters, the greater the
amount of time spent on listening to the radio per week, the poorer the performance in 4 of
the Grade 4 subjects.

National Organization for Examinations 47

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 38. Absenteeism and achievement

% of Total
Frequency Mean N Std. Dev
N
None 50.24 5157 .1419 42.8%
1-5 days 49.78 4546 .1449 37.7%
6-10 days 46.66 1014 .1597 8.4%
11-15 and more days 41.28 1338 .1477 11.1%
Total 48.77 12055 .1479 100.0%
Students’ absenteeism is directly related to the level of their achievement as shown in
Table 38. In this particular case, the relationship between absenteeism and performance was
(r = -.172) significant but negatively correlated. Unless a student attends classroom
instruction, it is obvious that not much can be expected in terms of academic achievement. In
Ethiopia, students may miss classes due to a number of reasons. For instance, the qualitative
study of this report indicates that parental demands for children’s labor at home plays an
important role. Students from economically low families may quit classes for doing some
business on local market days to support themselves in school. Health and other factors may
also contribute to student absenteeism in Ethiopia.
Table 39. Frequency at which attendance is taken
Std.
Frequency Mean N % of Total N
Deviation
Never 35.83 619 .1247 5.1%
Sometimes 44.49 1961 .1527 16.2%
Always 50.40 9506 .1432 78.7%
Total 48.70 12086 .1486 100.0%
To minimize the negative impact of pupils’ absenteeism, taking frequent attendance by
teachers is identified as a solution to the problem. Table 39 indicates that as the frequency of
checking students’ attendance increases, performance is likely to increase. These two
variables correlate significantly in the positive direction with r = .247. This implies that
taking attendance regularly has a positive contribution to students’ achievement.
Table 40. Daily meals and achievement
Std.
Daily meals Mean N % of Total N
Deviation
Once 41.97 1007 .1487 8.4%
Twice 49.01 4175 .1479 34.6%
Three or more times 49.60 6877 .1458 57.0%
Total 48.76 12059 .1482 100.0%

From the Table above, it can be observed that a considerable number (43%) of the sampled
students did not get sufficient number of meals. However, more than half of the students
reported fed themselves three times a day. These responses are quite associated with the
levels of students’ overall performance in school subjects. The correlation between these two
variables was (r = .109) significant at .01 level of P value, indicating that getting sufficient
number of meals is a contributing factor for children to achieve well in school. Fuller (1986)
refers to studies in Guatemala, Egypt and Uganda to show how malnutrition brought about
lower student achievement (in verbal skills, reading, and mathematics and comprehensive
exams) by causing illness in students. In Ethiopia, the current finding is extremely relevant
because schools do not provide feeding services under normal circumstances. School feeding

National Organization for Examinations 48

The Ethiopian Second National Learning Assessment Grade 4 Report

is significant because little children can not grow physically and intellectually to their full
potential unless they eat well. In this respect, gaining experiences from other developing
countries seems essential to learn how they addressed this important dimension in lower
primary education.
Table 41. Interest in learning English
Level of Mean N Std. Min Max % of Total
Interest Deviation N
Low 38.20 738 .1251 .06 .82 6.1%
Average 46.05 3060 .1358 .03 .89 25.5%
High 50.80 8203 .1486 .01 .92 68.4%
Total 48.81 12001 .1480 .01 .92 100.0%
With regard to the level of students’ interest towards learning the English language, Table 41
provides the responses of three different groups; “Low”, “Average” and “High”. Of the
respondents, 68.4% said that they had “High” interest in the subject while 25.5% rated their
interest as “Average”; 6.1% put their interest as “Low”.
These different levels of students’ interest positively correlated (r = .227) with their
achievements in the subject, and the correlation was significant at the 0.01 level (2-tailed).
That is, those who showed high interest in the subject achieved a high composite mean score
of 50.80%, while those with an “Average” level of interest achieved the next higher mean
score (46.05%). Those who reported the level of their interest in English as “Low” achieved
the least mean score (38.20%). This implies that achievement in a particular subject can, to
some extent, be gauged by the degree of interest in that particular subject.
Table 42. Interest in learning mathematics
Level of Composite Std. % of
N
Interest Mean Deviation Total N
Low 37.30 667 .1225 5.6%
Average 45.57 2652 .1432 22.1%
High 50.76 8679 .1449 72.3%
Total 48.87 11998 .1477 100.0%
Background information gathered from the sampled students with regard to their attitudes
toward learning mathematics, indicated that 72.3% had a high degree of interest. However,
22.1% of them rated their interest as “average” and 5.6% as “low”. These responses
positively correlated (r = .236) with performances in the subject, and the correlation was
significant at the 0.01 level (2-tailed). That is, those who reported to have high interest
achieved a composite mean score of 51%, which is higher than the scores of the other two
groups. This implies that performance in a particular subject can meaningfully be influenced
by the level of interest in learning that subject.
Table 43. Interest in learning environmental science
composite Std.
Level of Interest N % of Total N
Mean Deviation
1 Low 37.43 622 .1244 5.2%
2 Average 45.76 2202 .1524 18.5%
3 High 50.35 9062 .1441 76.2%
Total 48.82 11886 .1482 100.0%

National Organization for Examinations 49

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 43 shows the interest of Grade 4 sampled students in the learning of environmental
science. Of the total number of respondents, 76.2% exhibited “High” interest in the subject
while 18.5% showed “Average” interest. 5.2% of the sampled students rated their interest as
“Low.” The levels of’ interest positively correlated (r = .213) with performances in the
subject, and the correlation was significant at the 0.01 level (2-tailed). Students who reported
to have high interest in the subject achieved a high composite mean score of 50.35% while,
those with an “Average” level of interest achieved the next higher mean score (45.76).
Students with“Low” interest achieved the least mean score (37.43%).

Table 44. English textbook distribution

Textbook Ratio Composite N Std. % of Total
Mean Deviation N
1:1 49.65 5445 .1447 45.3%
1:2 48.29 1887 .1476 15.7%
1:3 48.01 863 .1655 7.2%
1:4 45.89 1251 .1543 10.4%
No textbook available 48.78 2587 .1439 21.5%
Total 48.74 12033 .1480 100.0%

Table 44 indicates that 45.3% of the students had an English textbook each. This group had
the highest mean scores in relation to others. The majority of students, however, had one
textbook for two or more and they scored lower than those who had a textbook each.

Table 45. Mathematics textbook distribution

Mathematics textbook ratio composite N Std. % of Total N
Mean Deviation
1:1 50.25 6064 .1447 50.6%
1:2 47.78 2157 .1479 18.0%
1:3 43.77 673 .1532 5.6%
1:4 44.02 901 .1504 7.5%
No textbook available 48.84 2182 .1465 18.2%
Total 48.71 11977 .1480 100.0%

In Table 45, slightly over half of the students said that mathematics textbook- student ratio
was 1:1, and this group scored higher than others. Yet a considerable proportion uses a
textbook for two or more. These ones scored lower than their privileged peers.

Table 46. Environmental science textbook distribution

Envi Science Textbook composite N Std. % of Total
ratio Mean Deviation N
1:1 50.05 5957 .1453 49.4%
1:2 48.40 2242 .1483 18.6%
1:3 44.68 772 .1533 6.4%
1:4 45.80 902 .1532 7.5%
No textbook available 48.02 2188 .1442 18.1%
Total 48.71 12061 .1477 100.0%

National Organization for Examinations 50

The Ethiopian Second National Learning Assessment Grade 4 Report

The above table shows that about half of Grade 4 students do not have their own
environmental science textbooks. In fact, these students are also the ones who are at the
disadvantage when it comes to academic achievement.
In summary, it can be concluded from the three tables that students’ performance increases as
the student-textbook ratio increases. The correlation (r) between Grade 4 English,
mathematics, and environmental science textbook-student ratio and overall average score was
.043, .073, .075 respectively, indicating that the better access to learning materials makes a
difference in pupils’ achievement. However, a considerable proportion of students do not
have sufficient textbooks although some reports indicate that much work has been done in the
sphere in Ethiopia. Referring to different studies, UNESCO (2004) indicates that lack of
textbooks can result from an inefficient distribution system, and malpractice or corruption. In
Zambia, for instance, it notes that not even 10% of the books procured had reached
classrooms. UNESCO’s suggestion is that liberalizing textbook markets and strong
government coordination may contribute to drop textbook prices and higher availability.
4.4 Grade 4 Pupils’ Attitudes in Socially Relevant Issues
The following shows the response frequency distributions for a health care, environmental
protection, civics and ethics, cultural conditions, and education subscales. For analyzing
mean differences, student responses depicting favorable attitude were assigned a score of 3
those depicting unfavorable attitude1, and neutral responses 2. Negatively phrased items
were shaded. Overall, attitudes tend to be in the favorable direction, suggesting that Ethiopia's
social development curriculum is making a difference in positively shaping students' attitudes
around health, environmental protection, civics and ethics, cultural issues and education
values. Total number of students that responded to each of the attitude questions varied,
ranging from 12616 to 12843.

National Organization for Examinations 51

The Ethiopian Second National Learning Assessment Grade 4 Report

4.4.1 Grade 4 Pupils’ Attitudes in Socially Relevant Issues at National

Level
Table 47. Attitude towards health care
Percentage of Respondents
Items Agree Neutral Disagree
We can prevent the spread of diseases through personal and
environmental hygiene. 79.27 10.52 10.21
I would like my parents to give birth over and over again so that I
can have many brothers and sisters. 45.46 12.17 42.37
As far as we are aware of the modes of transmission of HIV/AIDS
we will not be infected by the virus. 68.83 10.79 20.38
Blood tests cannot help identify HIV/AIDS infection. 34.62 16.67 48.71
Helping HIV/AIDS infected people is a social obligation. 63.09 11.54 25.37
Learning with those with HIV/AIDS causes no problem. 50.17 12.21 37.62
Eating meat regularly is a proper dietary practice. 35.79 12.40 51.81

Majority of the students (79.27%) agreed that with sanitation diseases could be prevented.
The second highest percentage (68.83) showed agreement that if they knew how HIV/AIDS
is transmitted and took care, they would not be infected. On the other hand, 51.81 %
disagreed with the statement ‘Eating meat regularly is a proper dietary practice’ and 45.46 %
agreed with the statement ‘I would like my parents to give birth over and over again so that I
have many brothers and sisters’; 48.71% disagreed with the statement ‘Blood test can not
help identify HIV/AIDS infection’ where the neutrals (16.67%) and the agreed (34.62 %)
made the 51.29%.

Table 48. Attitude towards environmental protection

Items Agree Neutral Disagree

Drinking river water can expose one to water borne diseases. 58.46 10.76 30.78
Visiting historical places is wasting both money and time. 42.73 11.98 45.29
Planting and taking care of trees is pleasurable. 84.97 6.25 8.78
One should plant trees to prevent erosion. 80.71 6.46 12.83
Keeping the environment clean is everyone’s duty. 77.78 7.81 14.41
It is possible to prevent the consequences of drought. 65.82 11.66 22.52
It is necessary to burn forests to drive away wild animals that
destroy crops. 29.59 10.05 60.36
84.97 % agreed that “Planting and taking care of trees is pleasurable,” 80.71% agreed that
one should plant trees to prevent erosion and 77.78 % agreed that “keeping the environment
clean is every one’s duty”; 45.29 % disagreed with the statement “visiting historical places is
wasting time and money” as opposed to the 42.73% who agreed on the same statement.

National Organization for Examinations 52

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 49. Attitude towards civics and ethics

Items Agree Neutral Disagree

Women are equal to men in knowledge and thinking. 57.03 11.00 31.97
It is appropriate to return property found lost either to its owner or
to the school administration. 79.55 7.41 13.03
I believe that group work is important for academic success. 79.66 7.44 12.90
Gathering information about our surrounding is necessary to have
knowledge about it. 77.17 8.76 14.07
Solving differences through dialogue is good culture. 79.52 7.38 13.10
Boys and girls should equally participate in household chores after
school. 79.19 7.08 13.73
While 57.03% agreed to the statement “women are equal to men in knowledge and thinking,
31.97% disagreed. The statements “Lost property should be returned” (79.55%), “Group
work is important for academic success” (79.66%), “solving differences through dialogue is
good (79.52%) and “Boys and girls should equally participate in household chores” (79.19%)
won high percentages.
Table 50. Attitudes towards harmful cultural practices

Items Agree Neutral Disagree

Early marriage is a practice that should be abolished. 56.38 9.25 34.38
Female circumcision is a harmful traditional practice that should be
abolished. 62.55 9.61 27.84
56.38% of the students agreed to the statement “Early marriage should be abolished” while
34.38% disagreed. “Female circumcision is a harmful traditional practice that should be
abolished” won the agreement of 62.55% of the respondents.
Table 51. Attitude towards education

Items Agree Neutral Disagree

Absenteeism and coming late to school are signs of laziness. 71.04 8.05 20.91
Student participation in teachers’ evaluation is appropriate. 79.41 6.80 13.80
Going to school and staying at home are the same to me 29.77 10.45 59.78
Learning is a means to earning a higher status. 84.85 5.37 9.79
Because I am a student, I have more knowledge and maturity than
those not in school. 81.28 6.70 12.03
“Learning is a means to earn a higher status” (84.85%), “Students have more knowledge and
maturity than those not in schools” (81.28%) and “Students should be involved in teacher
evaluations “(79.41%) were statements that won extensive agreement. Those who disagreed
with the statement “Going to school and staying home are the same to me” (59.78%) were
relatively fewer than those in the former groups.
4.4.2 Grade 4 Pupils’ Attitudes towards Socially Relevant Issues at Regional
Level
Analysis of subscale mean differences and region-by-response cross-tabulations provided
insight into social awareness differences across regions. For analyzing mean differences,
student responses depicting favorable attitude were assigned a score of 3 while those
depicting unfavorable attitude were assigned a score of 1. Neutral responses were assigned a
score of 2. The Scheffe paired comparisons procedure (ANOVA post-hoc tests) was applied
to identify homogeneous groupings across regions using an alpha level of 0.005. The alpha

National Organization for Examinations 53

The Ethiopian Second National Learning Assessment Grade 4 Report

level was set at 0.005 to compensate for inflation of Type I error that can be expected when
conducting more than one statistical test on the same population.
The following tables show the within region distribution of responses for individual attitude
survey questions.
Table 52. We can prevent the spread of diseases through personal and environmental hygiene
Region Agree Neutral Disagree
Tigray 66.63 19.79 13.58
Afar 85.71 7.54 6.75
Amhara 91.28 4.30 4.42
Oromia 73.42 12.38 14.20
Somali 81.95 5.30 12.75
Ben-Gumuz 81.72 9.01 9.27
SNNPR 76.11 13.67 10.22
Harari 81.69 9.53 8.77
Addis Ababa 90.96 4.21 4.83
Dire Dawa 75.06 11.05 13.88
NATION 79.27 10.52 10.21
Table 52 shows that big majority of respondents in Amhara region (91.28%) and Addis
Ababa (90.96%) agreed that personal and environmental hygiene could help prevent the
spread of diseases. Respondents from Tigray stood last in order of agreement (66.63).
Table 53. I would like my parents to give birth over and over again so that I can have many brothers and
sisters.
Region Agree Neutral Disagree
Tigray 50.05 10.96 38.99
Afar 64.77 7.42 27.80
Amhara 44.87 5.37 49.76
Oromia 43.70 19.56 36.74
Somali 76.49 7.62 15.89
Ben-Gumuz 56.61 7.21 36.18
SNNPR 40.91 12.42 46.67
Harari 39.03 13.01 47.96
Addis Ababa 27.26 9.77 62.96
Dire Dawa 42.45 14.71 42.84
NATION 45.46 12.17 42.37
Table 53 shows the between region differences of students’ attitude regarding family
planning. The overall national percentage of those who agreed to the statement and those
who disagreed was very close (45.46% and 42.37% respectively). For most of the regions
there was no great difference in the percentage between those who agreed and those who
disagreed. The highest percentage difference was between respondents from Somali region
(76.49% agreed, and 15.89% disagreed). In the Afar region too, the gap between those who
agreed and the two groups was considerable (64.77% agreed and 27.80% disagreed). A
relatively smaller percentage of respondents in Addis Ababa (27.26%) agreed with the
statement.

National Organization for Examinations 54

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 54. As far as we are aware of the modes of transmission of HIV/AIDS and take care, we
will not be infected by the virus.
Region Agree Neutral Disagree
Tigray 58.49 14.52 26.99
Afar 69.50 8.99 21.51
Amhara 71.03 6.93 22.04
Oromia 67.02 13.04 19.94
Somali 74.96 6.84 18.20
Ben-Gumuz 70.03 9.41 20.56
SNNPR 66.91 12.50 20.59
Harari 74.32 8.36 17.33
Addis Ababa 78.04 5.41 16.55
Dire Dawa 67.79 14.55 17.66
NATION 68.83 10.79 20.38
Table 55. Blood tests cannot help identify HIV/AIDS infection
Region Agree Neutral Disagree
Tigray 30.78 20.04 49.18
Afar 32.10 15.32 52.58
Amhara 32.47 13.00 54.53
Oromia 36.73 18.19 45.08
Somali 47.83 12.54 39.63
Ben-Gumuz 37.23 14.52 48.25
SNNPR 32.69 21.30 46.00
Harari 33.54 10.87 55.59
Addis Ababa 26.95 11.24 61.81
Dire Dawa 43.27 15.56 41.18
NATION 34.62 16.67 48.71

Table 56. Helping HIV/AIDS infected people is a social obligation

Agree Neutral Disagree
Tigray 50.81 14.70 34.49
Afar 75.12 8.87 16.01
Amhara 65.63 6.61 27.75
Oromia 60.30 15.81 23.89
Somali 64.74 8.24 27.02
Ben-Gumuz 62.92 9.50 27.58
SNNPR 52.64 13.86 33.51
Harari 74.55 9.55 15.91
Addis Ababa 89.98 3.20 6.82
Dire Dawa 67.36 12.95 19.69
NATION 63.09 11.54 25.37

National Organization for Examinations 55

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 57. Learning with those with HIV/AIDS causes no problem

Region Agree Neutral Disagree
Tigray 42.75 13.65 43.60
Afar 61.07 6.91 32.03
Amhara 53.08 6.58 40.34
Oromia 43.04 18.79 38.17
Somali 49.75 9.09 41.16
Ben-Gumuz 54.65 8.38 36.97
SNNPR 42.38 13.43 44.19
Harari 62.97 9.71 27.31
Addis Ababa 74.77 5.74 19.49
Dire Dawa 53.49 14.73 31.78
NATION 50.17 12.21 37.62
Tables 54, 55, 56 and 57 show the attitude of the respondents regarding HIV/AIDS issues.
These tables show that respondents from Addis Ababa have better awareness of the issue.
The number of respondents undecided about the issue was, however, not insignificant.
Table 58. Eating meat regularly is a proper dietary practice.
Region Agree Neutral Disagree
Tigray 28.48 14.45 57.07
Afar 44.69 8.68 46.62
Amhara 34.36 6.21 59.43
Oromia 35.02 16.89 48.09
Somali 54.93 6.74 38.32
Ben-Gumuz 40.38 9.96 49.66
SNNPR 35.63 15.05 49.31
Harari 35.66 8.80 55.54
Addis Ababa 21.16 8.38 70.46
Dire Dawa 42.34 14.29 43.38
NATION 35.79 12.40 51.81

Table 58 shows that 70.46% of the respondents from Addis Ababa disagreed with the
statement while only 38.32% disagreed from Somali Region. Those from the other regions
varied between the two limits.
Table 59. Drinking river water can expose one to water borne diseases.
Region Agree Neutral Disagree
Tigray 49.95 11.26 38.79
Afar 58.15 10.19 31.66
Amhara 60.01 6.81 33.18
Oromia 59.80 14.23 25.97
Somali 65.07 8.90 26.03
Ben-Gumuz 63.13 6.76 30.11
SNNPR 51.28 12.89 35.83
Harari 58.01 9.82 32.18
Addis Ababa 70.21 6.08 23.71
Dire Dawa 63.86 10.47 25.67
NATION 58.46 10.76 30.78

National Organization for Examinations 56

The Ethiopian Second National Learning Assessment Grade 4 Report

According to Table 59, 70.21% of the respondents from Addis Ababa agreed with the idea
that drinking river water causes water borne diseases while the figure was 49.95% from
Tigray. Others spead between the two extremes.
Table 60. Visiting historical places is wasting both money and time.
Region Agree Neutral Disagree
Tigray 41.02 13.85 45.13
Afar 45.58 8.83 45.58
Amhara 42.65 7.79 49.55
Oromia 47.39 16.10 36.50
Somali 44.26 8.49 47.25
Ben-Gumuz 45.60 8.67 45.73
SNNPR 41.12 13.39 45.50
Harari 43.47 11.25 45.29
Addis Ababa 28.57 7.56 63.87
Dire Dawa 44.69 13.08 42.23
NATION 42.73 11.98 45.29
Table 60 shows that 63.87% of respondents from Addis Ababa and 36.50% from Oromia said
that visiting historical places is not a waste of time and money. The other regions fall
differently between the two.
Table 61. Gardening gives me pleasure.
Agree Neutral Disagree
Tigray 77.55 9.76 12.69
Afar 90.75 4.08 5.17
Amhara 92.54 2.60 4.85
Oromia 80.27 7.88 11.85
Somali 83.93 4.59 11.48
Ben-Gumuz 84.71 6.78 8.51
SNNPR 84.36 7.48 8.16
Harari 84.85 5.61 9.55
Addis Ababa 94.22 2.48 3.30
Dire Dawa 81.75 7.20 11.05
NATION 84.97 6.25 8.78
Table 61 shows that the majority of students in all regions get pleasure from gardening.
Significantly high in descending order were Addis Ababa (94.22%), Amhara (92.54%) and
Afar (90.75%). Tigray, in this regard, was the lowest (77.55%).

National Organization for Examinations 57

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 62. One should plant trees to prevent erosion.

Region Agree Neutral Disagree
Tigray 62.98 8.69 28.33
Afar 87.15 4.23 8.62
Amhara 90.37 2.95 6.67
Oromia 75.23 8.92 15.85
Somali 79.93 5.22 14.85
Ben-Gumuz 83.31 5.07 11.62
SNNPR 81.78 7.59 10.63
Harari 81.21 6.21 12.58
Addis Ababa 88.62 3.00 8.38
Dire Dawa 79.43 6.86 13.71
NATION 80.71 6.46 12.83
Table 62 shows that majority of the respondents believe that planting trees could prevent
erosion. 90.37% those from Amhara agreed with this statement while only 62.98% from
Tigray did the same. The percentage for respondents from other regions was between these
two points.
Table 63 below indicates the attitude of the sampled students towards environmental
protection. 91.19% of the respondents from Addis Ababa agreed that protecting the
environment from pollution is the duty of everybody, while pertinent figure for Oromia was
only 64.85%. For the rest of the regions the percentage lay between the two boundaries.
Table 63. Keeping the environment clean is everyone's duty.
Region Agree Neutral Disagree
Tigray 74.81 8.65 16.54
Afar 85.17 4.57 10.25
Amhara 86.28 3.03 10.69
Oromia 64.85 13.20 21.95
Somali 84.62 3.76 11.62
Ben-Gumuz 81.65 4.26 14.10
SNNPR 78.34 8.63 13.03
Harari 81.97 6.21 11.82
Addis Ababa 91.19 2.80 6.01
Dire Dawa 72.47 11.30 16.23
NATION 77.78 7.81 14.41
Table 64. It is possible to prevent the consequences of drought.
Region Agree Neutral Disagree
Tigray 58.36 12.94 28.69
Afar 68.76 7.54 23.70
Amhara 71.68 6.90 21.42
Oromia 58.85 17.17 23.98
Somali 71.62 8.25 20.13
Ben-Gumuz 68.21 9.93 21.85
SNNPR 61.01 12.00 26.98
Harari 75.04 9.53 15.43
Addis Ababa 81.11 7.43 11.46
Dire Dawa 69.51 13.82 16.67
NATION 65.82 11.66 22.52

National Organization for Examinations 58

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 64 shows that the majority of respondents agree that it is possible to prevent the
consequences of drought. The percentage of students who agree, however, differs from
region to region. For instance, 81.11% of the respondents from Addis Ababa agreed with the
statement while only 58.36% agreed from Tigray. Others lay between these two points.
Table 65. It is necessary to burn forests to drive away wild animals that destroy crops.
Region Agree Neutral Disagree
Tigray 30.36 11.48 58.15
Afar 31.71 7.22 61.07
Amhara 27.20 5.64 67.16
Oromia 26.36 14.76 58.89
Somali 34.60 8.11 57.28
Ben-Gumuz 32.62 7.59 59.79
SNNPR 35.07 10.75 54.18
Harari 29.51 7.95 62.54
Addis Ababa 19.86 5.35 74.79
Dire Dawa 28.33 12.29 59.38
NATION 29.59 10.05 60.36

Table 65 shows the percentage of respondents who agreed, disagreed remained neutral in
each region with regard to the statement. 74.79% of the respondents from Addis Ababa
disagreed while the percentage for SNNPR was only 54.18%. The percentage of those who
disagreed with the statement in other regions is between these two.
Table 66. Women are equal to men in knowledge and thinking.
Region Agree Neutral Disagree
Tigray 48.26 11.41 40.33
Afar 55.05 8.68 36.28
Amhara 68.58 5.15 26.27
Oromia 45.86 18.35 35.79
Somali 70.02 5.77 24.22
Ben-Gumuz 51.33 5.59 43.09
SNNPR 58.51 10.67 30.82
Harari 61.52 10.76 27.73
Addis Ababa 69.88 6.97 23.16
Dire Dawa 54.73 14.07 31.20
NATION 57.03 11.00 31.97

As seen in Table 66, majority of the respondents agreed that women are equal to men. The
percentage of students who agreed or disagreed, however, differs from region to region. For
instance, 70.02% of the respondents in Somali region agreed with the statement, while only
45.86% agreed in Oromia. The percentage of students who agreed in other regions spreads
between these two points.

National Organization for Examinations 59

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 67. It is appropriate to return property found lost either to its owner or to the school
administration.
Region Agree Neutral Disagree
Tigray 72.60 10.36 17.04
Afar 87.44 4.08 8.48
Amhara 88.46 3.47 8.07
Oromia 67.03 12.64 20.34
Somali 89.02 3.33 7.65
Ben-Gumuz 83.95 4.08 11.97
SNNPR 80.56 7.68 11.76
Harari 81.39 6.20 12.41
Addis Ababa 90.57 3.28 6.15
Dire Dawa 77.17 7.65 15.18
NATION 79.55 7.41 13.03
Table 67 shows that respondents from Addis Ababa stood at the top in terms of agreement
(90.57%) agreed to the statement above. The percentage of those who expressed agreement
from Oromia region was lower (67.03%) than that of all the other sampled students.
Table 68 (below) shows that the majority of students in all regions believe that group work is
important for academic success. 90.08% of respondents from the Somali region agreed to the
statement while only 72.41% did from Dire Dawa. The percentage of respondents in other
regions who agreed to the statement spread between the two limits.
Table 68. I believe that group work is important for academic success.
Region Agree Neutral Disagree
Tigray 77.88 7.26 14.85
Afar 82.78 6.16 11.06
Amhara 87.33 3.61 9.06
Oromia 72.93 11.08 16.00
Somali 90.08 4.63 5.29
Ben-Gumuz 81.18 5.13 13.68
SNNPR 78.41 8.31 13.28
Harari 83.82 5.65 10.53
Addis Ababa 84.79 4.52 10.69
Dire Dawa 72.41 10.09 17.50
NATION 79.66 7.44 12.90

National Organization for Examinations 60

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 69. Gathering information about our surrounding is necessary to have knowledge about it.
Region Agree Neutral Disagree
Tigray 77.52 9.85 12.63
Afar 81.42 6.46 12.13
Amhara 84.08 4.93 10.99
Oromia 66.36 14.22 19.42
Somali 86.14 5.61 8.25
Ben-Gumuz 79.00 6.30 14.70
SNNPR 77.69 8.84 13.48
Harari 79.20 5.81 14.98
Addis Ababa 88.33 4.44 7.23
Dire Dawa 71.47 10.15 18.38
NATION 77.17 8.76 14.07
In Table 69 we see that the percentage of respondents who agreed that gathering information
is important for acquiring adequate knowledge about their surrounding lay between 88.33%
(Addis Ababa) and 66.36% (Oromia). Even at the national level, it could be seen that the
majority of students agreed to this idea (77.17%).
Table 70. Solving differences through dialogue is good culture.
Region Agree Neutral Disagree
Tigray 75.84 8.85 15.32
Afar 84.91 4.87 10.22
Amhara 87.46 3.09 9.45
Oromia 70.54 11.72 17.74
Somali 87.19 4.93 7.88
Ben-Gumuz 82.12 4.64 13.25
SNNPR 80.08 7.55 12.37
Harari 81.89 6.09 12.02
Addis Ababa 86.96 4.24 8.80
Dire Dawa 73.19 10.10 16.71
NATION 79.52 7.38 13.10
As can be seen from Table 70, the percentage of respondents who agreed to the idea in the
statement came between 87.40% for Amhara region and 70.54% for Oromia. The figure for
the national level is 79.52% which means the majority agreed to the idea.

National Organization for Examinations 61

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 71. Boys and girls should equally participate in household chores after school.
Region Agree Neutral Disagree
Tigray 71.54 10.10 18.37
Afar 85.96 4.37 9.67
Amhara 86.40 2.66 10.94
Oromia 72.15 10.72 17.13
Somali 74.51 6.09 19.41
Ben-Gumuz 81.53 5.54 12.93
SNNPR 80.29 6.45 13.27
Harari 80.18 7.87 11.95
Addis Ababa 89.72 3.29 6.99
Dire Dawa 76.12 10.73 13.15
NATION 79.19 7.08 13.73

Table 71 shows that the majority of students in all regions agreed to the statement that boys
and girls should participate in household activities equally. Highest agreement was secured
in Addis Ababa (89.72%) while the lowest was in Tigray (71.54). The rest speed between
these extremes.
Table 72. Early marriage is a practice that should be abolished.
Region Agree Neutral Disagree
Tigray 45.75 10.87 43.38
Afar 55.13 4.74 40.13
Amhara 63.87 4.40 31.73
Oromia 48.73 15.06 36.21
Somali 55.39 7.63 36.98
Ben-Gumuz 60.40 6.36 33.25
SNNPR 58.27 9.23 32.50
Harari 55.91 7.27 36.82
Addis Ababa 69.16 4.92 25.92
Dire Dawa 55.90 11.41 32.69
NATION 56.38 9.25 34.38
It can be seen from Table 72 that although the majority of respondents in many regions agree
that early marriage is something that should be abolished, substantial percentage of
respondents in other regions disagree to its abolition. It is only 45.74% of the respondents in
Tigray and 48.73% in Oromia that agreed to this statement. 69.16% of respondents from
Addis Ababa who agreed to the idea make comparatively the highest group.

National Organization for Examinations 62

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 73. Female circumcision is a harmful traditional practice that should be abolished
Region Agree Neutral Disagree
Tigray 56.09 10.22 33.70
Afar 62.30 6.78 30.91
Amhara 74.93 3.39 21.69
Oromia 51.17 17.48 31.35
Somali 74.88 7.88 17.24
Ben-Gumuz 67.77 5.15 27.08
SNNPR 60.70 8.81 30.49
Harari 63.53 8.97 27.51
Addis Ababa 76.72 4.82 18.46
Dire Dawa 58.61 11.31 30.08
NATION 62.55 9.61 27.84

Table 73shows the majority of students in all regions believe that female circumcision is a
harmful traditional practice. Large percentages of respondents from Addis Ababa (76.72%)
agreed to the statement making the highest in the sample. The smallest percentage of
respondents who agreed to this statement, in comparison with other regions, was seen in the
Oromia region (51.17%).

Table 74. Absenteeism and coming late to school are signs of laziness.
Region Agree Neutral Disagree
Tigray 65.38 8.87 25.75
Afar 83.20 3.45 13.34
Amhara 82.98 3.50 13.52
Oromia 59.08 14.67 26.25
Somali 77.72 3.63 18.65
Ben-Gumuz 76.19 5.16 18.65
SNNPR 67.54 8.81 23.65
Harari 69.53 5.58 24.89
Addis Ababa 87.90 2.15 9.95
Dire Dawa 68.43 9.66 21.91
NATION 71.04 8.05 20.91
We understand from Table 74 that Addis Ababa agreed the most (87.90%) and Oromia
agreed the least (59.08). Even at the national level, it could be seen that the majority of
students agreed to the suggestion that laziness of a student is manifested by being absent from
or coming regularly late to school (71.04%).

National Organization for Examinations 63

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 75. Student participation in teachers’ evaluation is appropriate.

Region Agree Neutral Disagree
Tigray 73.02 10.17 16.81
Afar 83.33 4.76 11.90
Amhara 87.90 2.73 9.37
Oromia 74.44 9.42 16.15
Somali 85.09 6.03 8.88
Ben-Gumuz 81.73 3.47 14.80
SNNPR 76.73 7.05 16.22
Harari 80.64 6.40 12.96
Addis Ababa 87.32 3.71 8.97
Dire Dawa 74.68 10.98 14.34
NATION 79.41 6.80 13.80
Table 75 shows the majority of the students in all regions believe that they should participate
in the evaluation of teachers. 87.90% of respondents from the Amhara Region and 73.02%
from Tigray agreed to the statement making the highest and the lowest respectively. The
percentage for respondents from the other regions was between these two lines.
Table 76. Going to school and staying at home are the same to me.
Region Agree Neutral Disagree
Tigray 26.20 11.98 61.82
Afar 28.64 6.01 65.35
Amhara 27.74 6.25 66.01
Oromia 29.78 14.89 55.34
Somali 31.51 5.80 62.69
Ben-Gumuz 30.69 7.28 62.04
SNNPR 28.95 11.86 59.20
Harari 39.42 11.57 49.01
Addis Ababa 27.09 5.89 67.01
Dire Dawa 35.10 12.00 52.90
NATION 29.77 10.45 59.78
In Table 76 we see that the percentage of the national average for respondents who disagreed
with the idea that spending time in school and at home are the same was 59.78%. The
highest percentage of disagreement was 67.01% (Addis Ababa) and the lowest was 49.01%
(Harari). The rest of the regions come between the two.
Table 77 below shows that the majority of students in all the regions believe that learning
enables people to attain higher status. 94.64% of respondents from Addis Ababa, 92.36%
from Amhara, and 91.97% from Afar have agreed to the statement. The lowest percentage of
respondents who agreed to this statement, in comparison with other regions, was from
Oromia (76.25%).

National Organization for Examinations 64

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 77. Learning is a means to earn a higher status.

Region Agree Neutral Disagree
Tigray 82.53 7.35 10.12
Afar 91.97 3.15 4.88
Amhara 92.36 2.25 5.39
Oromia 76.25 8.50 15.26
Somali 88.63 2.64 8.73
Ben-Gumuz 86.02 3.83 10.16
SNNPR 87.10 5.43 7.47
Harari 80.45 6.02 13.53
Addis Ababa 94.65 1.23 4.12
Dire Dawa 76.70 8.19 15.11
NATION 84.85 5.37 9.79

Table 78. Because I am a student, I am more knowledgeable and mature than those not in school.
Region Agree Neutral Disagree
Tigray 80.17 8.42 11.41
Afar 88.21 4.56 7.23
Amhara 91.85 2.97 5.18
Oromia 69.77 10.54 19.70
Somali 82.83 4.67 12.50
Ben-Gumuz 84.66 4.50 10.85
SNNPR 82.85 6.20 10.94
Harari 80.66 7.25 12.08
Addis Ababa 90.73 3.81 5.46
Dire Dawa 75.06 8.87 16.07
NATION 81.28 6.70 12.03
Table 78 shows the majority of the students in all the regions believe that they are better than
their peers who do not go to school. Significant percentages of respondents of Amhara
(91.85%) and Addis Ababa (90.73%) agreed to the statement. Comparatively, the least
agreement was observed with respondents from Oromia (60.77%).

National Organization for Examinations 65

The Ethiopian Second National Learning Assessment Grade 4 Report

4.5 Relationships between Grade 4 Pupils’ Achievement and Factors

4.5.1 Correlations with Grade 4 Composite Score
Table 79 shows correlates of factors that affect achievement at different levels. The variables
are organized into five blocks: School Structure and Supplies, Teacher Behaviour, School
Management, Student Behaviour and Instruction / Support. Those factors which show
statistically significant relationships (p < .05) with the composite score are highlighted.
Table 79 Spearman Rho Correlation with Grade 4 Composite Score
Correlation with
Group of Variables Sig.
G4 Composite
Block 1: School Supplies
Availability of Mother Tongue Syllabus 0.065 0.22
Availability of the English Syllabus 0.046 0.38
Availability of Mathematics Syllabus 0.125 0.02*
Availability of Environmental Science Syllabus 0.155 0.003**
Pupil-Text Ratio in Mother Tongue Learning -0.038 0.25
Pupil-Text Ratio in English Learning -0.020 0.70
Pupil-Text Ratio in Environmental Science Learning -0.060 0.25
Average of Syllabi Availability 0.122 0.02*
Availability of Mother Tongue Teacher's Guide 0.107 0.04*
Availability of English Teacher's Guide 0.096 0.07
Availability of Mathematics Teacher's Guide 0.064 0.22
Availability of Environmental Science Teacher's Guide 0.087 0.10
Total Teacher's Guide 0.128 0.01*
Teacher's Perception of Student's Learning Attitudes 0.206 <.001***
Block 2: Teachers Quality and Behavior
Teacher Workload 0.017 0.74
Teacher Training on new teaching techniques 0.127 0.02*
Teacher Training on new assessment techniques 0.129 0.02*
Teacher Training average 0.106 0.05*
Teacher's Age -0.039 0.45
Teacher Qualifications -0.049 0.35
Total teaching experience -0.059 0.26
Teacher Distance to school -0.081 0.12
Other social commitments -0.015 0.78
Block 3: School Management
Students' Reporting Teachers taking Attendance 0.392 <.001***
Director Qualifications 0.069 0.20
Social obligations outside of job -0.020 0.68
Director experience 0.104 0.05*
How often are teachers supervised? 0.104 0.05*
Block 4: Students Behavior
Interest in subjects 0.298 <.001***
Meals a day 0.118 0.02*
Agreement between home and instructional language 0.160 0.002**
Students' sense of ownership in school property 0.204 <.001***
Time spent listening to the radio -0.171 0.001**
Time spent watching TV -0.134 0.01*
Distance from Home to school -0.082 < .001**
Absences from school -0.223 <.001***

National Organization for Examinations 66

The Ethiopian Second National Learning Assessment Grade 4 Report

Block 5: Instruction
Director's perception of English instruction quality 0.101 0.05*
Director's perception of Mathematics teacher's instruction quality 0.108 0.04*
Director's perception of Science instructional quality 0.094 0.08
Director's perception of M/tongue instruction quality 0.121 0.03*
Director's perception of all instruction quality 0.140 0.007**

As to the availability and conditions of curriculum materials in schools under the “school
supplies block”, there are 12 items considered important in bringing impact on the overall
level of student achievement. Among these, Mathematics Syllabus, Environmental Science
Syllabus, average of all syllabi and Teacher's Guide for Mother Tongue have significant
simple correlation with composite score. These curriculum materials are associated with
higher achievement.
With regard to teachers’ behavior, the variables having positive correlation with composite
score are teacher training on new teaching techniques, training on assessment techniques, the
use of teacher's guide, and teacher's perception or expectation about the learning of their
students.
On the other hand, some characteristics of teachers behavior such as “teacher age”, “teacher
qualifications”, “total teaching experience”, “distance teachers travel from home to school”
and “other social commitments of teachers” have negative but small relationships with levels
of achievement as shown in Table 4.4.1
Concerning the school organization and management block, the degree to which head
teachers focus on school matters, taking students’ attendance frequently, directors’
experience, and supervising teachers are considered as elements of the overall quality of
management having positive relationships with achievement. Whereas, some factors like
“directors’ qualifications and social obligations outside of school” have weak or negative
relationships with achievement.
Factors grouped under students’ background result in high association with achievement.
Some of these factors such as “students’ interest in school subjects, the amount of meals
students get in a day, similarity between home and instructional language, and students' sense
of ownership in school property have strong positive correlation with achievement, while
other factors like “time spent listening to the radio and watching TV programs, distance from
home to school, and being absent from school” are associated with low level of performance
having strong negative correlations with achievement.

National Organization for Examinations 67

The Ethiopian Second National Learning Assessment Grade 4 Report

4.5.2. Results of Multiple Regression Analysis

This section deals with the exploration of the overall influence of the factors that showed
strong and significant correlation with the composite score by using multiple regression
analysis at the school level. The variables were organized into five blocks. Initially each
block was analyzed separately using the standardized method (ENTER) of multiple
regression analysis. The main purpose of this was to examine the extent to which each block
and each variable in each block influences or explains the variability of student achievement.
This means not all variables examined in the correlation analysis strongly and significantly
determine students overall achievements. Table 80 summarizes the result.

Table 80. School-level multiple regression results on the Grade 4 composite achievement score details on
reduced models

Predictor models and Std.

Variables coefficient Sig. R R-sq F df.
School Structure and curricular
material supplies 0.266 0.071 10.215*** 2,268
Availability and making use of syllabuses 0.337 0.002
Availability and making use of teacher's
guides 0.385 0.004
Teachers’ Behavior 0.253 0.064 8.147*** 3,356
Teachers' perception of student learning 0.532 <.001
Teachers’ traveling long distances from
home to school -0.102 0.054
Teachers’ use of textbooks 0.095 0.022
School Management 0.414 0.171 35.992*** 2,349
Teachers taking attendance 0.16 <.001
Directors’ overall managerial experience 0.086 0.041
Students’ Home Background and
Behavior 0.435 0.189 16.533*** 5,355
Average interest in subjects 0.316 0.009
Students' sense of ownership of school
property 0.233 0.007
Time spent listening to the radio -0.284 0.008
Absence from school -0.345 0.01
Similarity of home and school language 0.0041 <.001
Instruction Support 0.162 0.026 8.361** 1,312
All teachers' instructional quality 0.342 0.007

The variables under school structure and curriculum materials have explained 7.1% of the
variation in learners' achievement. There are only two variables under this block, availability
and making use of syllabuses and teacher's guides. Both variables showed strong positive
contribution to achievement.
The variables under the block teacher's characteristics and behavior accounted for 6.4% of the
variation in achievement. Under this block 'teachers' perception of student learning' showed
very strong relation with achievement. Teachers' use of textbooks also contributed positively.

National Organization for Examinations 68

The Ethiopian Second National Learning Assessment Grade 4 Report

On the other hand the variable 'teachers' traveling long distance from home to school' showed
negative relation.
The school management block explained 17.1% of the variation observed in achievement
score. Taking attendance and the director's managerial experience both contributed positively.
The variables under students' home background and behavior explained 18.9% of the
variation observed in achievement. Schools in which students demonstrate a sense of
ownership towards school properties and where high level of interest in their learning is
developed tend to exhibit higher achievement. The alignment of home and instructional
language also contributed positively. On the other hand absenteeism and time spent listening
to non-educational radio programs contributed negatively.
Finally the director's perception of instructional quality which is the only variable under the
block instruction and support explained 2.6% of the variation observed in achievement.
Table 81 shows the Model Summary which resulted from the Hierarchical Multiple
Regression Analysis. In the first column the numbers in bracket show the number of variables
each model contains. The maximum total variance explained (shown under the column R
Square) by a single model is 27.8 % (Model 5) which contains all the variables listed in Table
80 lumped together.

Table 81 School level multiple regression results on the composite score model summary
Std. Error
Adjusted of the R Square Sig. F
Model R R Square R Square Estimate Change F Change df1 df2 Change
1 (2) .157 .025 .018 9.38 .025 3.790 2 300 .024
2 (5) .287 .083 .067 9.14 .058 6.255 3 297 .000
3 (7) .465 .216 .198 8.48 .134 25.189 2 295 .000
4 (12) .525 .275 .245 8.23 .059 4.704 5 290 .000
5 (13) .527 .278 .246 8.22 .003 1.122 1 289 .290

Table 82 shows the contribution of each block to the final model when entered in different
orders. It can be seen from the table that all the blocks have significantly contributed to the
explanation of the variation in achievement of learners when they are entered first. Their
contribution, however, dwindles and becomes insignificant in the case of blocks one and five
when they are entered last. This fact shows that the other three blocks explained most of the
variations observed in achievement scores.

National Organization for Examinations 69

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 82 Contributions of each block to R-square of completely reduced model

R square
contribution R square
if entered contribution if
Predictor models and Variables first entered last
1. School Structure and Supplies .021* .004 (n.s.)
Availability and making use of syllabuses
Availability and making use of teacher's guides
2. Teacher Variables .068*** .029**
Teachers' perception of student learning
Teachers’ traveling long distances from home to school
Teachers’ use of textbooks
3. School Management .164*** .060***
Student perception of the role of teachers taking attendance
Director overall managerial experience
4. Student Home Background and Behavior .172*** .057***
Average interest in subjects
Students' sense of ownership of school property
Time spent listening to the radio
Absence from school
Similarity of home language and school language
5. Instruction .029** .002 (n. s.)
Director's perception of all teachers' instructional quality

The variables grouped under Student Background and Behavior explained the highest portion
(17.2%) of the observed variation in student achievement at school level before other factors
were considered and 5.7% after controlling or considering all other factors in the model. This
means, of the 28.7% variations explained by the factors, Student Background and Behavior
ranked highest in both cases of entry order for analyses. This block contained five variables
Among these, interest in the subject matters, students’ sense of ownership of school property
and alignment of home and instructional language showed positive relationship with
achievement while absenteeism and time spent listening to the radio showed negative
relationship.
School Organization and Management block explained 16.4% of the observed variation in
student achievement before controlling other factors and still a significant 6.0% with other
factors controlled. This means, of the 27.8% variations explained by the factors in the current
study, School Organization and Management ranked as the second highest in both cases of
entry order for analyses. This block contained two variables, namely frequency of taking
attendance by the teachers and director’s managerial experience. Both factors showed
positive relationship with the student achievement.
Teachers' Variables explained the observed variation in achievement by 6.8% before other
factors were controlled and 2.9% after controlling all other factors. This means of the 27.8%
variations explained by the factors in the current study, Teachers’ Variables ranked as the
third highest in both cases of entry order for analyses. This block contained variables which
asked about teachers’ perception (judgment) of students, the distance they traveled from
home to school, and the use of the textbooks. The distance the teacher traveled from home to
school affected student achievement negatively while the other two variables were related
positively with achievement.

National Organization for Examinations 70

The Ethiopian Second National Learning Assessment Grade 4 Report

School Supplies and Structures explained the observed variation in achievement by 2.1%
before other factors were considered and by 0.4% after controlling all other factors in the
model. This block contained two variables, namely availability and use of syllabuses and
teacher’s guides. Both factors showed significant positive relationships with achievement.
Finally the Instruction and Support block, which includes factors concerning the role of the
director in supervising instructional activities, accounted for 2.6% of the total variation in
student achievement before other factors were controlled and 0.2% after controlling all the
other factors.

National Organization for Examinations 71

The Ethiopian Second National Learning Assessment Grade 4 Report

4.6 Comparative Analysis between EBNLA and ESNLA

4.6.1 General Comparison
One of the major purposes of this study was to make comparative analysis on the findings of
EBNLA and ESNLA. In this part, an attempt is made to discuss the major findings obtained
by the study. The following tables show the mean achievements of both EBNLA and ESNLA
assessment measures followed by a short analysis.
Table 83. EBNLA mean scores at national level
Envi.
Reading English Mathematics Composite
Science
Mean 64.25 40.46 39.31 48.14 47.94
N 10411 10297 10388 10495 10281

Table 84. ESNLA mean scores at national level

Envi.
Reading English Mathematics Composite
Science
Mean 64.49 38.67 39.69 51.74 48.48
N 12886 13028 13143 13005 12507

Table 83 and Table 84 contain the summary report of the national mean scores on the
baseline and the second national learning assessments of the country respectively.
The tables show that the sampled pupils performed better in the two basic reading
comprehension tests than in the remaining three subjects in both cases. It is also true that
their performances in the two environmental science tests were slightly higher as compared to
the results of both English and mathematics tests.
Moreover, there is an acute shortage of textbooks since only less than half of the sampled
students reported that the textbook-pupil ratio was one to one while more than half of them
used either one textbook for many or did not have any. The correlations between textbook-
student ratios in English, mathematics, and environmental science and the Grade 4 composite
mean score are 0.043, 0.073, and 0.075 respectively. This indicates that the level of textbook
distribution can bring about meaningful differences in pupils’ performance.
Grade 4 students have also reported that the levels of their interests towards the instruction of
these subjects vary considerably. Based on the reported levels of students’ interest in their
learning (HIGH, AVERAGE and LOW), a correlation matrix was produced and relationships
between Grade 4 composite mean score and interest in the environmental science (r = .213),
English (r = .227) and mathematics (r = .236) have been determined. This finding depicts that
the level of one’s interest in learning a subject determines one’s performance more in Grade 4
mathematics and English than in environmental science. This might be partly due to the
typical nature of the subjects and partly due to the inadequacy of human and material inputs
made available to schools. There were also problems of large class size and lack of
instructional facilities as reported by the samples.
Therefore, these and other related problems in schools might have contributed to the low
performance of pupils, especially, in the English and mathematics subjects. In addition to
this, the finding with reference the mathematics for all sampled pupils is similar to those
reported by other researchers.

National Organization for Examinations 72

The Ethiopian Second National Learning Assessment Grade 4 Report

Finally, it is quite imperative to note that in both EBNLA and ESNLA, the obtained average
mean scores in Grade 4 English and mathematics were far below the 50% minimum
requirement to pass from one level to the next higher level as stipulated by the Education and
Training Policy (ETP) of our country. Another point, quite worth noticing, is that ETP from
the start claims much emphasis to be given to the instruction of these subjects at all levels of
the education system.
Table 85. Learning Progress: Comparison between the two national learning assessments

Subject Reading English Mathematics Envi. Science Composite

Time EBNLA ESNLA EBNLA ESNLA EBNLA ESNLA EBNLA ESNLA EBNLA ESNLA

Means 64.25 64.49 40.46 38.67 39.31 39.69 48.14 51.74 47.94 48.48
Progress
(T2-T1) 0.24 -1.77 0.38 3.60 0.54

Table 85 compares the levels of pupils’ achievement obtained in 4 academic subjects on the
two NLAs conducted in 1992 E.C. and 1996 E.C. respectively.
The overall composite mean scores of EBNLA (47.94) and ESNLA (48.48) are almost
similar with an increment of 0.54 points in the ESNLA, which is negligible. This implies that
there is no gain in students’ level of achievement between the baseline and the current
assessment measures. In other words, it can be concluded that during the last 4 years, school
instructions in Ethiopia didn’t register progress in the academic performances of Grade 4
students.
When we consider the specific results gained in the individual subjects, the current status of
student performance is not satisfactory. For example, taking the two English tests, there is a
considerable drop in gain (-1.86%) from the baseline. And there is no significant change
obtained in the results of the two readings and mathematics tests from EBNLA. Nevertheless,
a 3.6% gain was achieved in the environmental science tests from the baseline assessment.
4.6.2 Specific Comparison
The sampling strategy called for carry over of 10 schools for each region that had participated
both in the Ethiopian Baseline National Learning Assessment (EBNLA, 1992 E.C.) and the
Ethiopian Second National Learning Assessment (ESNLA, 1996 E.C.). This strategy would
provide for a comparison of baseline and current performance levels for the nation overall
and, with substantially less power for detecting statistical significance, within regions. A
logical alternative for evaluating change in performance from baseline would be to conduct
statistical tests that are designed to detect mean differences from independent samples. In the
latter case, the EBNLA sample and the ESNLA sample would be considered independent
samples of schools taken at different years. This alternative to the paired comparisons
analysis was, however, problematic due to the difference in the sampling strategies that were
applied for the EBNLA and the ESNLA. Therefore, it was decided to include a set of schools
in the current sample that had participated in the EBNLA and to conduct a repeated measures
analysis to investigate change from baseline.
Though the paired comparisons analysis is, in theory, a more powerful test and can often
yield very meaningful results because the subject (i.e., the school in this situation) serves as
its own control, the limited number of matched schools in combination with the large
variances in difference of scores from the Baseline to the ESNLA reduced the power for
detecting statistically significant differences. In addition, since five statistical tests were
conducted, statistical significance was tested at the 0.01 alpha level in order to compensate

National Organization for Examinations 73

The Ethiopian Second National Learning Assessment Grade 4 Report

for the inflation of Type I error that can be expected when conducting multiple test statistics
on the same sample population.
The statistical results presented below should be considered only in conjunction with careful
review and analysis of the descriptive results. It is often the case that when there are limited
numbers of schools sampled (e.g., less than the 10 matched schools originally planned for the
paired comparisons analysis) the sample population tends to be very skewed and precludes
the valid conduct of parametric statistical tests (e.g., the dependent t-test). However, analysis
of the normality of the data indicated that the samples, even within regions, were relatively
normal. Using the standard error of skewness as a diagnostic statistical tool, it was
determined that there was no marked deviation in spread that would justify using analysis
alternatives to the paired t-test. This is confirmed in the relatively similar mean and median
measures of central tendency given in the tables below.
The compiled descriptive statistics for the difference scores are given below for the nation
overall and for each region. The difference score was calculated for each school by
subtracting average school (percent correct) score obtained at Baseline from the average
school (percent correct) score obtained at the ESNLA. These difference scores were then
summarized for the Nation and for each Region as shown below. To support the
interpretation of results, schools were grouped into high and low groups and these groupings
were then crossed with responses on the six school reform variables from the director's
questionnaires.
Progress in the Basic Reading Subject
Results below show that for the 83 matched schools sampled across the nation there was
considerable variability in change from baseline. The worst performing school fell 24
percentage points while the highest showed a gain of approximately 20 percentage points.
Though results were not statistically significant at the 0.01 alpha level, the performance of
Oromia is noteworthy in that on the average, schools in this region fell in their performance
by 7 percentage points.
Table 86 Reading: average change in percent correct scores from EBNLA to ESNLA
REGION Number Mean Standard Median Minimum Maximum Test of Mean Differences
of Deviation from Baseline to Second
Matched NLA (Paired t-test)
Schools T Statistic P Value
NATION 83 -1.56 9.04 -0.25 -24.78 19.99 -1.58 0.119
TIGRAI 9 1.25 7.20 4.25 -11.00 13.25 0.52 0.616
AFAR 7 4.02 6.48 6.00 -6.38 13.13 1.64 0.152
AMHARA 13 -1.67 5.32 -1.75 -10.88 8.17 -1.13 0.279
OROMIA 8 -7.09 6.51 -8.19 -16.38 2.50 -3.08 0.018
BEN-GUMUZ 6 -4.61 8.49 -3.63 -17.34 4.50 -1.33 0.241
SNNPR 17 0.10 11.46 2.91 -18.83 19.99 0.04 0.972
HARARI 6 -0.30 13.38 2.03 -24.78 13.60 -0.05 0.959
ADDIS ABABA 9 -0.27 6.74 2.25 -15.57 6.13 -0.12 0.907
DIRE DAWA 8 -7.55 10.16 -6.50 -20.43 7.00 -2.10 0.074

Progress in ENGLISH
Results below show that for the 83 matched schools sampled across the nation, there was a
significant drop in performance from Baseline to the ESNLA.

National Organization for Examinations 74

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 87. English: average change in percent correct scores from EBNLA to ESNLA
Test of Mean
Number
Differences from
of Standard
REGION Mean Median Minimum Maximum Baseline to Second NLA
Matched Deviation
(Paired t-test)
Schools
T Statistic P Value
NATION 83 -3.35 8.01 -3.26 -25.59 16.23 -3.81 0.000*
TIGRAI 9 -1.49 6.52 -1.63 -11.94 7.93 -0.69 0.511
AFAR 7 -4.58 2.72 -4.69 -8.87 -0.06 -4.45 0.004*
AMHARA 13 -3.84 5.76 -3.88 -13.38 8.46 -2.41 0.033
OROMIA 8 -13.02 8.32 -14.38 -24.25 3.31 -4.43 0.003*
BEN-GUMUZ 6 -4.95 5.98 -3.72 -13.57 2.22 -2.03 0.099
SNNPR 17 -1.40 6.89 0.53 -21.81 7.94 -0.84 0.415
HARARI 6 -0.92 10.20 -0.34 -13.61 10.69 -0.22 0.833
ADDIS ABABA 9 -2.49 12.22 -2.63 -25.59 16.23 -0.61 0.558
DIRE DAWA 8 0.36 6.76 0.65 -7.97 9.04 0.15 0.884
*Significant negative change from baseline (p<0.01)

It can be seen in the above table that though there was considerable variability in the
magnitude of the negative change, on the average, performances fell in every region for
Grade 4 English. Attention is drawn to the maximum difference scores given in the table
which underscores the fact that some schools did improve in English performance. Further
research investigating the reasons for performance regression and for progress in English is
recommended. Two regions stand out as having significant (p< 0.01) drops in English
performance: Afar and Oromia. It can be seen that in these regions even the maximum scores
fail to reflect significant gains.
Progress in Mathematics
The results below show that for the 83 matched schools sampled across the nation, there was
no significant change in performance from Baseline to the SNLA. Performance change
varied considerably with a minimum of drop in performance by approximately 15 percentage
points to a 20 percentage points gain. As in English and environmental science, Oromia
showed a significant drop (p<0.01) in performance from Baseline. All of the matched schools
in Oromia fell in performance on Grade 4 mathematics.
Table 88. Mathematics: average change in percent correct scores from Baseline to SNLA
Test of Mean
Number
Differences from
of Standard
REGION Mean Median Minimum Maximum Baseline to Second NLA
Matched Deviation
(Paired t-test)
Schools
T Statistic P Value
Total 83 -1.57 7.72 -1.84 -14.77 20.86 -1.85 0.068
TIGRAI 9 -1.73 6.54 -1.84 -13.98 6.60 -0.79 0.450
AFAR 7 -3.30 5.39 -3.43 -12.31 4.96 -1.62 0.156
AMHARA 13 -3.51 6.52 -5.29 -12.38 9.28 -1.94 0.076
OROMIA 8 -8.29 5.71 -8.77 -14.54 -2.14 -4.10 0.005*
BEN-GUMUZ 6 -1.69 8.35 -3.14 -10.21 9.69 -0.50 0.641
SNNPR 17 -0.29 7.80 0.21 -14.77 15.58 -0.15 0.879
HARARI 6 0.16 10.95 3.74 -14.00 9.98 0.04 0.973
ADDIS ABABA 9 2.78 9.80 0.86 -8.80 20.86 0.85 0.420
DIRE DAWA 8 1.22 5.78 2.39 -7.97 9.16 0.60 0.569
*Significant negative change from Baseline
(p<0.01)

National Organization for Examinations 75

The Ethiopian Second National Learning Assessment Grade 4 Report

Progress in the Environmental Science Subject

Results below show that for the 83 matched schools sampled across the nation, there was no
significant change in performance from Baseline to the ESNLA. Performance change varied
considerably with a minimum of 14 percentage point drop to a high of approximately 16
percentage point improvement. On region, Oromia, showed a significant drop (p<0.01) in
performance from the baseline. All of the matched schools selected from Oromia fell in
performance in Grade 4 environmental science.
Table 89. Environmental science: average change in percent correct scores from Baseline to SNLA

Test of Mean Differences

Number
Standard from Baseline to Second
REGION of Matched Mean Median Minimum Maximum
Deviation NLA (Paired t-test)
Schools
T Statistic P Value

NATION 82 0.52 6.47 1.83 -14.44 15.95 0.73 0.468

TIGRAI 9 1.96 6.06 3.70 -7.99 11.47 0.97 0.361
AFAR 7 3.59 4.86 3.45 -5.78 9.52 1.96 0.098
AMHARA 13 0.17 5.96 1.79 -9.35 9.81 0.10 0.919
OROMIA 8 -5.55 3.71 -4.51 -11.90 -1.93 -4.23 0.004*
BEN-GUMUZ 6 -0.99 6.38 0.85 -10.72 5.63 -0.38 0.719
SNNPR 17 1.28 7.40 1.86 -13.74 15.95 0.72 0.485
HARARI 6 2.60 7.16 5.37 -8.09 8.55 0.89 0.415
ADDIS 8 1.91 7.01 3.13 -10.57 12.08 0.77 0.466
ABABA
DIRE DAWA 8 -0.58 6.49 -0.05 -14.44 7.49 -0.25 0.807
*Significant negative change from baseline p<0.01)

Compiled Performance Composite Score across Regions

The following table presents the descriptive statistics for the average difference score from
EBNLA to the ESNLA. Only Oromia showed a significant change, a significant (p<0.01)
drop in the compiled achievement change score. With the exception of Oromia some schools
in all regions made noteworthy improvements, with maximum gains ranging from 4.3 to
14.22 percentage points.
Table 90. Average difference in percent correct scores across Grade 4 achievement tests
Test of Mean
Number
Differences from
of Standard
REGION Mean Median Minimum Maximum Baseline to Second
Matched Deviation
NLA (Paired t-test)
Schools
T Statistic P Value
NATION 83 -1.35 6.82 -0.05 -15.99 14.22 -2.45 0.017
TIGRAI 9 0.09 5.81 -0.16 -11.23 7.86 0.04 0.966
AFAR 7 0.02 3.37 1.35 -4.12 4.33 0.01 0.999
AMHARA 13 -1.87 5.75 -0.05 -9.01 10.56 -1.17 0.265
OROMIA 8 -8.28 5.18 -9.21 -15.99 -0.05 -4.52 0.003
BEN-GUMUZ 6 -3.06 6.88 -2.93 -12.96 4.30 -1.09 0.326
SNNPR 17 -0.07 6.89 2.72 -12.44 8.55 -0.04 0.969
HARARI 6 0.44 9.87 2.92 -14.11 10.35 0.11 0.918
ADDIS ABABA 9 0.96 8.55 1.39 -15.13 14.22 0.34 0.744
DIRE DAWA 8 -1.73 6.47 -1.13 -12.07 6.63 -0.75 0.474

National Organization for Examinations 76

The Ethiopian Second National Learning Assessment Grade 4 Report

4.7 Analysis of Qualitative Data

4.7.1 Overview
This study was set to complement the quantitative study, which used statistical tools to
determine the extent to which learning takes place in primary schools, using interviews and
focus group discussions. It also explored the main explanatory conditions of student learning
using basically similar questions used in the quantitative study from different sources.
The qualitative study was carried out in all the regions where quantitative data has been
collected. The source of information included teachers, Grade 4 students and parents. Each
of these respondents was asked relatively similar questions so that the acquired information
could be cross-checked using the triangulation of sources of data. Large numbers of
languages were used for interviews and focus group discussions. Findings from Grade 4
students’ qualitative study were included in this report because their views are considered
valid as they could provide cumulative information acquired in the entire primary education
system. The analysis of the collected data has been carried out at national and regional levels.
Student learning was gauged by asking respondents what students learn in schools and if the
learning was relevant to their needs. Explanatory conditions were conceptualized around
such categories as school supplies, teacher behaviors, curricular and system level issues,
school administration, parental behavior and student personal characteristics. In the following
section the national and regional level of analyses and interpretation of data are presented.

4.7.2 Characteristics of Student Learning

Teachers believe that primary school children gain basic knowledge, skills and attitudes.
However, in many regions teachers expressed dissatisfaction with the extent to which
students learn. As it can be seen from the summary given in Table 90, teachers used such
metaphors as “schooling is better than remaining at home” and “ they learn how to read and
write” to show the minimal learning taking place in schools. Repeatedly, teachers expressed
“the lack of student interest or vision” in learning to show how little is going on in academic
achievement. In few cases, it has been expressed that students gain useful knowledge,
particularly in environmental understanding and personal hygiene.
Parents perceive that students’ learning did not make a significant change. They underline
that their children come from school without sufficient and appropriate knowledge. There is
still an agreement with teachers in many cases that “students lack interest and motivation to
learn”. Some even do not acknowledge that learning takes place while others say that
“students lack normal development”. Marginal or minimal knowledge acquired is also
expressed as “those who go to school are better than those who remain at home”. In some
circumstances, there are feelings showing that the “time spent on learning and learning gains
do not commensurate”. In a very particular case, parents acknowledged that students learn
useful knowledge and they “developed communication skills”. The emphasis on the
development of communication skills is interesting, because this particular case is in one of
those regions where the medium of instruction is the mother tongue. Parents expressed
concerns about “pooling factors” that contribute to low level learning in schools.
In most cases, students believe that they have satisfaction with their learning. They
emphasized that they get sufficient knowledge and skills in environmental issues, hygiene,
civics and social skills. Few students perceived “lack of interest of students in learning”.
A comparison of the results of qualitative study with quantitative study indicates that the
findings are complementary to each other. A composite scores analysis of test results for
different subjects and regions at the national level was 48.8% for Grade 4. What this finding
says is that learning indeed takes place in the Ethiopian Grade 4 Primary Education, but it is

National Organization for Examinations 77

The Ethiopian Second National Learning Assessment Grade 4 Report

not satisfactory. This is so because the composite scores of students is less than the expected
average for the nation. In the table below, one of the consistent positive evaluations was
expressed in Harari. If one looks into the Harari quantitative achievement results, it becomes
vivid that the average achievement score is the third highest. This implies that findings are
consistent with each other although the methods used for data collection and analyses are
different.
Teachers, students’ and parents views are almost similar in almost all regions except for
Harari. This case requires a special attention and continued research as to why the region
continues to perform more than others. Table 91 summarizes the findings of this part of the
study below.
Table 91. Student learning as viewed by teachers, students and parents by region.
Regions Respondent views on what students learn and its relevance
Teachers Students Parents
Oromia Basic valuable skills and attitudes. Basic knowledge about Basic knowledge about
Some students have no interest to environment. environment
learn. Have interest to learn. Learning is unsatisfactory and
Schooling is better than non- Schooling is better than non- weak.
schooling schooling Schooling is better than non-
schooling
Amhara Lack of satisfaction and confidence in Acquire important knowledge. Failure and no adequate
students to learn. learning
Only the ability to read and do some Better than those who remain
writing at home.
Tigray Students do not show the required Lack of motivation to learn Lack of motivation to learn
competence. Differentiate between good
Lack of vision. and bad,
Better than those who remain
at home.
SNNPR No satisfaction with student learning Satisfied with their learning Do not know that their
Basic knowledge about the children learn.
environment, and hygiene
Understanding of rights and
responsibilities, social skills
Addis Ababa Low interest to learn Low interest to learn Sufficient learning does not
Good moral make-up, but spoiled by take place
environment Student lack normal
development
Dire Dawa Interest to learn is high Interest to learn is high Children acquire
Insufficient/inadequate knowledge is inadequate/insufficient
gained by students knowledge
Somali Student understanding of their Learned reading and writing Better than those who remain
environment is better. at home.
Keep personal hygiene
Afar Basic learning takes place, but not Low interest to learn Low interest to learn
satisfactory.
Low or no interest to learn.
Students have no clear visions.
Harari Grasped useful knowledge Grasped useful knowledge Grasped useful knowledge
Believe in their changed social Developed communication
attitudes skills
Ben-Gumuz Basic knowledge, skills and attitudes Basic knowledge, skills and Basic knowledge, skills and
Students do not have interest to learn attitudes attitudes
No satisfaction with what they learn Students do not have interest
to learn
No satisfaction with what they
learn
Time spent and learning gains
do not commensurate.

National Organization for Examinations 78

The Ethiopian Second National Learning Assessment Grade 4 Report

4.7.3 Conditions Affecting Student Learning

The analyses of explanatory conditions as expressed by respondents were conceptualized into
issues at school and outside school levels. School level issues consistently revolved around
such factors as school facilities and supplies, teacher behaviors, availability and competence.
Issues related to the curriculum, system characteristics, parents and students were considered
as outside school factors that influence student performance. The analysis of these
discussions are presented in the following sections both using narrations and tables.
1. School Supplies and School Facilities
Table 91 below summarizes that teachers, students and parents have the opinion that lack of
school supplies and facilities impede student learning. In almost all regions, shortage of
textbooks, lack of reading rooms, lack of classrooms, lack of laboratories, insufficient
utilization of the available support services, and malfunctioning of the available resources
were some of the main issues raised in the discussions.
In addition to shortages of supplies and facilities, the main concern of teachers was the poor
utilization of the available support services including tutorials and student counseling.
Reasons for this were related to student involvement in business and other activities for self
sufficiency. This in turn led to the shortage of time to participate in extra-school activities
required for effective learning.
Table 91 also shows that students and parents in all regions shared almost all of the concerns
of schools teachers. If one goes back to the quantitative study, it is obvious that school
supplies have a great influence in student achievement. The regression analysis indicated that
school supplies explain about 7% of student achievement. The correlation analysis also
showed that some school supplies like the availability of the syllabi and teacher’s guides have
significant positive relations with student achievement. These findings lead to the fact that the
more schools are supplied with the necessary facilities the better will be student achievement.
The concern with the shortage of textbooks is a nationwide phenomenon. Given the
dependence of the school system on textbooks the issue has to be taken seriously. What the
quantitative study tells in this respect is even more alarming. The correlation analysis
indicates that the pupil-text book ratio negatively correlated with student achievement. This
means the preparation and distribution of textbooks of the current type makes negative
difference in learning rather than the positive development expected in society. In this sense,
what the finding tells is not the insignificance of textbooks in student learning. Rather, it
notes that if the quality of textbooks that students use is not improved and standardized, it
causes student failure to learn and achieve.

National Organization for Examinations 79

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 92. School supplies and school facilities.

Regions Respondents’ views on the influence of school supplies on student learning
Teachers Students Parents
Oromia Shortage of text books Shortage of text books Large number of students in
Lack of library and reading room Lack of library and reading the classroom
Lack of laboratories room
Large number of students in the Lack of laboratories
classroom

Amhara Inadequate utilization by students of Large number of students Inadequate utilization by

available services (e.g. tutorials) in the classroom. students of available
Shortage of textbooks, blackboards Wastage of school tutorial services (e.g. tutorials)
Lack of laboratories services due to lack of Shortage of textbooks,
Shortage of classrooms. time. blackboards
Lack of laboratories
Shortage of classrooms
Tigray Lack of textbooks Lack of textbooks Lack of textbooks
Shortage of laboratories and Shortage of laboratories Shortage of laboratories
libraries and libraries and libraries
SNNPR Minimal use by students of Lack of textbooks Lack of adequate services
materials and tutorials provided by Shortage of seating places like adequate classrooms,
teachers, Lack of textbooks, in the classroom text books
laboratories and reading rooms,
SPCs
Shortage of classrooms
Addis Shortage of laboratories and Shortage of laboratories Shortage of classrooms
Ababa libraries and libraries Shortage of laboratories
and libraries
Dire Dawa Shortage of textbooks Shortage of textbooks Shortage of textbooks
Lack of library, reading room, Lack of library, reading Lack of library, reading
laboratory room, laboratory room, laboratory
Somali Shortage of basic learning materials Shortage of basic learning Shortage of basic learning
from parents materials from parents materials from parents
Inadequate use of tutorial services Inadequate use of tutorial Inadequate use of tutorial
for students are engaged in business services for students are services for students are
Shortage of libraries and engaged in business engaged in business
laboratories Shortage of libraries and
laboratories
Afar Shortage of textbooks and lack of Shortage of laboratories Shortage of laboratories
teaching materials and libraries and libraries
Harari Lack of reading rooms Lack of reading rooms and Available school services
Malfunctioning of available malfunctioning of those are not properly used by
facilities that are available students
Lack of laboratories Lack of laboratories
Instructional services were not
properly used for students spend
time doing business
Ben- Shortage of textbooks Lack of textbooks for arts Lack of teaching materials
Gumuz Unsatisfactory use of tutorial and and music education including textbooks
counselling services

National Organization for Examinations 80

The Ethiopian Second National Learning Assessment Grade 4 Report

2. Teacher Behaviour, Availability and Competence

Teachers, parents and students showed the influence of the teacher in learning achievement in
many ways. Teachers in almost all regions feel that teachers lack the required qualification
and competence to deal with the curriculum they are supposed to teach. Except in Amhara,
Somali and Harari regions where this issue was not directly reflected, teachers in all other
regions pointed out that there is a gap between their current professional capability and the
demands of the curriculum. Late assignment of teachers in some regions has been identified
to cause a problem for student learning.
Table 93. Teacher behavior, availability and competence
Regions Respondents’ views on the influence of teacher behaviour, availability and competence on
student learning
Teachers Students Parents
Oromia Shortage of qualified teachers Teachers lack adequate Failure to cover yearly portions
Lack of teachers’ competence to preparation to teach in the
teach subjects classroom
Teachers give tutorial services,
but student use is inadequate
Amhara Teacher support exists, but used Teachers do not cover all Lack of qualified teachers
inadequately by students portions in the syllabi Loose relationships between the
Loose relationship between schools and community
teachers and parents Insufficient imparting of
knowledge
Tigray Shortage of qualified teachers Teachers do not care to Shortage of qualified teachers
Late assignment of teachers in help Late assignment of teachers in
schools Teachers do not give schools
feedback for their
learning
SNNPR Teacher support is available, but Failure to cover the Instruction is impeded by
not used adequately. syllabi students participation in
Lack of qualified teachers Inadequate use of teacher business
Assignment of incompetent support
teachers
Addis Teachers do not cover large Teachers are unable to Teachers do not cover large
Ababa number of subjects and portions cover subjects and number of subjects and portions
Teachers lack competence to portions Teachers lack competence to
teach most subjects teach most subjects
Dire Dawa Shortage of qualified teachers Shortage of qualified Insufficient imparting of
teachers, particularly in knowledge in schools
English

Somali Inadequate use teachers’ support __ ___

like tutorial
Afar Lack of qualified teachers Teachers are absent Teachers are absent frequently
Teachers’ dissatisfaction in their frequently
work
Harari Insufficient use of tutorials and Participation in business Self –help activities by students
make up classes does not allow to use limits teachers’ efforts
teacher efforts
B. Gumuz Tutorial services Inadequate use of tutorial Inadequate teaching practices
Counselling and competition services
among students Teachers’ lack of
Shortage of qualified teachers competence or
qualification

National Organization for Examinations 81

The Ethiopian Second National Learning Assessment Grade 4 Report

Teachers also commonly showed that they are not capable of covering all the portions of the
curriculum given for the year. They attribute this to large amount of contents prescribed in
the curricula. In some cases, job dissatisfaction of teachers has been reported to influence
learning. On the positive side, teachers in many regions indicated that they organize student
support services, although it has been insignificantly utilized.
Students agree with teachers on many points. They recognize the lack of qualification and
competence on the part of teachers. Students are aware of the fact that the contents of the
curricula are too much to cover. At the same time, students acknowledge teachers lack of
preparation, absenteeism and lack of willingness to provide feedback in their learning as
impediments to achievement. Students also acknowledge the efforts of teachers in the
provision of support services. However, it has been underutilized due to student engagement
in other activities. Apart from the above, parents and teachers mention that there is a loose
relationship between the school and the community.
A comparison of the findings of the quantitative study with the above findings ascertains that
teacher’s behavior and competence is a strong factor influencing student achievement. In the
multiple regression analysis, it has been found out that teachers’ behavior explains 17.1% of
the variations in student learning. In the correlation analysis, teacher training on the average
has shown a strong significant relation with student achievement. For summary of teacher
behavior that influences student learning, refer to Table 92 above.
In summary, the most important teacher qualities that were identified to influence student
learning are lack of teachers’ qualification and competence, late assignment of teachers to the
work place, inability of teachers to cover portions given in the curriculum, absenteeism and
lack of preparation for teaching.
3. Student Characteristics
Student behaviors that negatively influence academic achievement have been identified in
almost all regions. From teachers’ perspectives to absenteeism, less interest and motivation
for learning, engagement in unwanted and wasteful environment, lack of discipline, lack of
study time, lack of background in kindergarten education, and engagement in business have
been mentioned as the most important influences on student learning.
Student absenteeism has been associated with parental labor demands on children. Since
students achievement very much relates to the time spent on learning, absenteeism of students
can be considered as a national problem demanding attention. Less interest in learning and
lack of motivation of students have been related to the presence of undesirable environments
around the schools as well as the consideration of employment perspectives. Teachers
complain in almost every part of the nation about lack of student discipline. One explanation
for such feeling may be related to the new approaches to teaching-learning in which teachers
require to adopt new attitudes towards student-centered approach. Given the passive nature of
the instruction process in previous circumstances such reflection from teachers is only
normal. Capacity building in this area seems vital.

National Organization for Examinations 82

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 94. Student behaviors influencing academic achievement

Regions Respondents’ views on the influence of student characteristics on student learning
Teachers Students Parents
Oromia Less class attendance due to Attraction towards TV, Lack of discipline due to
parental labour demands radio, movies, etc large number of students in
Undisciplined acts of students Low level of parental the classrooms
support
Amhara Some students have less interest to Most students have low Some students have less
learn material support from interest to learn
Some waste time in unwanted home Some students waste time in
engagements unwanted engagements
Some students lack discipline.
Students lack study time due to
labour work
Tigray Most students do not attend KG Less time for study due to Most students receive low
and that creates problems for later parental need for labour material support from home
learning
Students lack interest in learning
Students have confused
assumptions
SNNPR Engaged in trade for helping Engaged in trade for Most students get low
themselves helping themselves. parental support.
Irregular classroom attendance Do petty trade to help their
schooling
Addis Less interest to learn and study Less interest to learn and Students get low parental
Ababa Student engagement in wasteful study support and follow-up
environment
Dire Dawa Students have interest to learn Less classroom attendance Students have interest to learn
Less classroom attendance because because of business
of business
Somali Lack of study time due to Lack of study time due to Lack of study time due to
engagement in trade engagement in trade engagement in trade
Afar Less interest to learn Lack of study time due Low level of parental support
Engagement in chewing chat labour demands at home
Lack of discipline from students
Harari Students are engaged in trade to Students are engaged in Students are engaged in trade
help themselves trade to help themselves to help themselves.
Students are discouraged by jobless Students are discouraged Students are discouraged by
school leavers by jobless school leavers jobless school leavers.
B. Gumuz Students have less interest to learn Lack of confidence in Students have less interest to
Lack of discipline among students education learn
Arrogance of school children
on the rise

From teachers explanations, it is also understood that students lack competence for primary
schooling due to the fact that the curriculum requires attending pre-primary education. It is
equally and significantly mentioned by teachers that students lack study time due to their
engagement in business and other activities. The findings show that student achievement is
not only conditioned by schools factors, but also by matters outside it as well. If students do
business to help themselves, it is obvious that their living standards need to be improved to
enable them concentrate on their studies. Since this requires a long-term strategy, however,
schools may need to consider these situations of learners by arranging flexible programs to
minimize lost learning.

National Organization for Examinations 83

The Ethiopian Second National Learning Assessment Grade 4 Report

Students and parents agree with teachers concerning student behaviors influencing their
academic performance. A different expression from the students shows that there is lack of
confidence on the part of students to learn. This can be explained in different ways, but as the
next section shows the difficulty with the curriculum and lack of vision of the significance of
learning may play their roles for such behaviors.
Like the qualitative data, the quantitative results summarized in the previous section indicated
that most of the variations in student academic achievement in Ethiopia are explained by
student behavior. In the correlation analysis, significant relations have been observed
between student behavior and student achievement more than any other factor. In the
regression analysis it has been demonstrated that student behavior explains 18.9% of the
variation observed in the current study. This is by far the strongest factor explaining student
learning in this country.
4. School curricula
Issues related to school curricula and the school system have strongly surfaced during the
interviews and focus group discussions. As Table 95 below shows, teachers, students and
parents were unanimous in pointing out some of the features of the curriculum and the
system. These include the difficulty of the ccurriculum both to the students and teachers,
difficulty to cover portions due to the mismatch between contents and time, the impracticality
of self-contained classes, the bare advantage of continuous assessment and the discouraging
influence of the free promotion of students from one Grade to the other.
Teachers indicate that the curriculum is difficult both for themselves and the students. This
has been already indicated in the previous section where teachers’ belief indicated that they
lack the necessary qualifications and competence in relation to the new curricula. It is
alarming that the curriculum is seen as a mismatch between time and content. Even more so
is the perceived difficulty of the subjects for students. Unless students understand the subject
they are supposed to learn, it threatens the whole efforts of schooling.

National Organization for Examinations 84

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 95 The influence of the curriculum on students achievement.

Regions Respondents’ views on the influence of the curricula on student learning
Teachers Students Parents
Oromia Curriculum is difficult both to the Curriculum is relevant. Curriculum does not
students and teachers Learning in the mother prepare children for work.
Difficult to cover portions due to the tongue is relevant. Curriculum misses some
mismatch between contents and time Self-contained classes social values of society.
Self-contained classes are impractical are impractical for a Moral education is lacking.
Continuous assessment is a useful single teacher. Learning in the mother
practice Continuous assessment is tongue is relevant.
Free promotion increases failures a useful practice. Free promotion discourages
repetition. Free promotion hard work.
discourages hard work.
Amhara The curriculum is difficult particularly Curriculum demands The curriculum is difficult
for rural children. unavailable facilities and particularly for rural
Curriculum demands unavailable conditions. children.
facilities and conditions. Self-contained classes Free promotion discourages
Curriculum lacks cohesion. are impractical. students.
Teachers can not implement self- Continuous assessment is
contained classes important.
Free promotion discourages students English at higher levels
Problems of translation and may be difficult.
interpretation of books.
Tigray Teachers were not involved in policy The curriculum is Self contained classes are
formulation. difficult. impractical.
Self-contained classes are not Portions are not covered.
supported by appropriate methods and
materials.
SNNPR Curriculum is relevant, but difficult. Curriculum is relevant, Curriculum is relevant, but
Self-contained classes make learning but difficult. difficult.
difficult. Free promotion advances Self contained classes make
Free promotion advances incompetence. learning difficult.
incompetence Free promotion advances
incompetence.
Addis Curriculum is difficult to students. Curriculum is difficult Curriculum is irrelevant.
Ababa Self contained classes are for teachers and students. Self contained classrooms
inappropriate. Free promotion is are inappropriate.
Free promotion is useless. useless.

Dire Dawa Subjects are difficult. Subjects are difficult. Subjects are difficult.
Self-contained classes are impractical Self contained classes cause
Free promotion advances insufficient learning
incompetence
Somali The curriculum is difficult. Period allotment is The curriculum is difficult.
inappropriate.
Afar Regional promotion policy at lower Subjects are difficult Self contained classes are
Grades is discouraging. not useful
Harari Learning in the mother tongue is good. Learning in the mother Learning in the mother
Subjects are difficult. tongue is good. tongue is good.
Self-contained classes are better than Subjects are difficult. Subjects are difficult.
non-self contained. Curriculum is relevant. Continuous assessment is
Continuous assessment is good. Continuous assessment is good.
Free promotion hurts learning. good.
Ben- Subjects are difficult. Subjects are difficult. Subjects are difficult.
Gumuz Curriculum is irrelevant to life Curriculum helps to Curriculum does not
conditions. solve some problems. prepare for work.
Self contained classes are difficult. . Free promotion discourages
Free promotion discourages hard hard work.
work.

National Organization for Examinations 85

The Ethiopian Second National Learning Assessment Grade 4 Report

Self-contained classes were considered impractical for the reason that teachers are not
competent to teach all subjects. It is also considered boring both for teachers and students.
The use of continuous assessment is taken as a positive side of the curriculum by teachers.
However, the practice of free promotion has been rejected almost in regions due to the
perception that such policy encourages incompetence and the pushing through of children
who do not deserve promotion. Since the policy to use continuous assessment is meant to
follow the progress of students from time to time in order to determine student promotion, the
automatic promotions in schools seem to indicate failure to use the continuous assessment
approach for its intended purpose.
Students and parents reflected similar opinions concerning the curriculum. In this respect,
parental special concerns require special attention. In addition to what has been commonly
pointed out in the above analysis, they say that the curriculum is irrelevant since it does not
prepare children for the world of work. Given the two primary cycles which also have
terminal points in their structure, this concern is valid. Parents also say that some social
values are not embedded in the curricula. There is a suggestion that moral education be
included in the curricula. Parents appreciate that learning takes place in the mother tongue.
Some complained, however, that the current curriculum is particularly difficult for rural
children. This observation has to be understood in relation to the previous finding which
indicated the negative influence of the absence of kindergarten education for children. Since
most rural children have no access to these facilities, there is a high possibility that the
curricula become difficult for learning.
5. Parental Support
All respondents perceive that parental support of students’ learning is very low. Even though
parents say that they provide learning materials, teachers pointed out that once parents send
their children to school, they feel they are relieved and never go to school to follow the
progress of their children. In almost all regions, except Addis Ababa, respondents confirm
that students are self-supporting, and are engaged either in petty business or home labor.
They feel that these have resulted in students’ truancy and low achievement. Parents say that
they do not let their school children attend tutorial classes when it is out of school time.
The fact that parents provide less academic support is not surprising. This is because parents
are neither as educated as their children nor do they understand how schools operate. Follow
up of children and the interest to discuss with teachers help find out what goes on between
their child and the school or teachers. On the other hand, parents do participate in formal
committees, raising funds, and classroom construction or school expansion. This is by itself a
positive trend in school-community relations. Table 95 below is the summary of findings
regarding parental support to school children nation wide.

National Organization for Examinations 86

The Ethiopian Second National Learning Assessment Grade 4 Report

Table 96. The influence of parental support on school achievement in Ethiopia.

Regions Respondents’ views on the influence of parental support on student learning
Teachers Students Parents
Oromia Inadequacy of academic support Demand for children’s Low perception of the
and follow up. labour importance of education
Support limited to basic Poverty of parents
learning materials High Participation in
construction of rooms and
teacher- homes.
Amhara High Participation in schools affairs No time for studying due No support for tutorial time
especially in fund raising. to labour demands at home
Academic support is minimal.
Tigray Rare participation in school affairs. Parents have no Rare participation in
participation student affairs
SNNPR Low parental academic support Low parental academic Lack of awareness on the
High Participation in formal support importance of education
committees Participation in school
expansion.
Addis Parents make no follow-up and do _______ Schools do not invite
Ababa not come to school for discussing parents for discussing
children’s affairs. children’s affairs.
Dire Dawa Fund raising whenever requested Less academic support by Employing teachers
parents Building classes
Somali Parental participation and support is Parental participation and Parental participation and
insignificant. support is insignificant. support is insignificant.
Afar Parents’ academic support is Parents’ academic support Parents’ academic support
minimal. is minimal. is minimal.
Demand for labour Demand for labour Demand for labour
Participation in school
affairs is insignificant.
Harari Low parental academic support Parental academic support Fund raising and
is low. construction of rooms and
Teachers employed at teachers residences
home
Ben-Gumuz Minimal parental academic Participate in school Participate in schools
support committees expansion
Construct teacher
residences

5. School administration
In most of the discussions, the influence of school administration on student achievement has
been implied to have direct or indirect consequences. One of the most important findings is
the fact that school administrations have been able to initiate student support services that
were insignificantly used. Poor use has been ascribed to students’ lack of time to make use of
those services and the lack of parental will to let children to benefit from those support
initiatives like tutorials and guidance and counseling. On the other hand, there are reports that
schools assign less competent teachers to teach in lower primary Grades. In some regions
there are indications that schools operate under severe shortage of budgetary resources.
Although schools were unable to initiate significant parental academic support for children, it
is obvious that their initiatives in terms of generating community support for school
expansion, classroom construction and the construction of residential houses for teachers are
common in almost all regions. However, it is clear that the construction of classrooms and

National Organization for Examinations 87

The Ethiopian Second National Learning Assessment Grade 4 Report

student enrolment do not match since reports from all regions point to the problem of large
classes.
Another interesting finding is the prevalence of student and teacher absenteeism in schools.
In the quantitative study, it has been found out that in schools where the director supervises
student and teacher attendance, student achievement has been higher than others. However,
failure to deal with disciplinary issues has been found out to negatively affect student
learning. In fact, it was one of the responsibilities of the school administration to maintain
discipline in schools. It was also indicated that lack of school facilities and supplies are
grounded in school operations all over the nation. The school administration was unable to
tap community resources to alleviate this problem throughout the nation.

National Organization for Examinations 88

The Ethiopian Second National Learning Assessment Grade 4 Report

CHAPTER 5: SUMMARY, CONCLUSIONS AND

RECOMMENDATIONS
5.1 Summary
The main purposes of the Second National learning Assessment (for Grade 4) was to measure
learning attainments of Grade 4 students, and identify the factors that determine those
attainments. It was also aimed at providing comparative information on school improvement
from the National Baseline Learning Assessment conducted in 1999/2000. In order to obtain
the required information for the study, both quantitative and qualitative research approaches
were used. All Grade 4 students in the country were targeted for the purpose of providing
national and regional estimates of student achievement results. A total of 13,248 randomly
selected sample students from 376 schools in all regions except Gambella have been
included. For the purpose of generating data on factors determining scholastic achievement,
751 teachers and 375 school principals were included. In the qualitative study, 321 students,
326 teachers and 246 parents from 44 schools provided information. The instruments of data
collection included achievement tests, attitude surveys, and school checklist forms,
questionnaires for teachers and school directors, and focus group discussion guides. Findings
from the study show the following main results.

1. The summary of composite achievement results at national level indicates that Grade
4 students achieved less than the expected minimum standards by the Ministry of
Education, but their results are slightly better than the National Baseline Learning
Assessment.
As it has been summarized in Table 96, the overall performances in all subjects were
low with only reading comprehension and environmental science averages reaching
50% correct. The mean percent score for mathematics (39.70) and English (38.68%)
were notably lower than the national mean (48.48%) and also from the mean of the
other two subjects.
2. A summary of composite scores by gender indicates that the achievement of boys was
higher than that of girls.
At the national level, the overall composite mean score found for girls was 46.12%,
while it was 50.14% for boys, indicating a gap of 4.02 percentage points. This same
general pattern appears in all the subject areas (basic reading 2.71%; English, 4.9%,
mathematics, 5.0%, and environmental science 4.82%). These results indicate that
concern for gender equity in schooling needs to extend beyond access, to the
provision of quality education that supports and encourages girls' achievement as
well. Though there is an indication of increment in the achievement of girls since the
baseline study, the gap between male and female student achievement has widened
from baseline.
3. A summary of composite scores by location indicates that urban students achieved
better than rural students.
Students from urban schools achieved a better mean score (49.28%) as compared to the
mean score (48.09%) of students from rural schools at the national level. This indicates
that the urban/rural difference in students’ achievement is 1.19%, which is statistically
significant at p < 0.001 level.

National Organization for Examinations 89

The Ethiopian Second National Learning Assessment Grade 4 Report

4. A summary of composite scores by regions indicates that there were disparities in

Grade 4 student achievement results among regions.
The findings for each achievement test and their composite mean are calculated across
regions and summarized in Table 97.
Table 97. Summary of Grade 4 achievement scores by region
Basic Environmental
Region Reading
English Mathematics
Science
Composite
TIGRAI 63.37 32.95 34.50 52.50 45.87
AFAR 71.92 35.30 40.29 51.60 49.64
AMHARA 70.92 37.87 43.27 60.05 52.67
OROMIA 63.73 39.49 38.30 50.73 47.83
SOMALI 50.53 50.36 43.09 45.83 47.18
BEN-GUMUZ 60.34 32.44 34.34 47.84 43.69
SNNPR 60.31 38.93 39.70 49.75 47.27
HARARI 68.60 42.46 40.71 53.35 51.29
ADDIS ABABA 76.90 39.94 44.92 54.64 53.69
DIRE DAWA 62.24 37.19 37.83 47.86 46.20

From the table above, one can infer that there are also considerable variations in
regional mean scores in each subject. Variations within each region also tended to be
high, with standard deviations ranging from 12 to 24 percentage points in every
subject.
Some regions showed a similar pattern of achievement across subject areas and scored
better composite means. These are Addis Ababa (53.69), Amhara (52.67), Harari
(51.29), and Afar (49.64) regions where each one of them scored above the national
mean percentage point. The first three regions have also achieved above the minimum
requirement of 50% pass mark. The remaining seven regions have regional mean
scores below the national mean. These regions are Afar, Oromia, SNNPR, Somali,
Dire Dawa, Tigray and Benishangul-Gumuz.
5. Student achievements in Grade 4 English were significantly less than the national
composite average for all subjects, and there were disparities among regions and
between sexes.
The average mean score obtained at the national level in the Grade 4 English was 38.68
with a standard deviation of 16.88. This is indeed significantly less than the composite
scores for all subjects at national level (i.e. 48.8%).
The Somali Region scored the highest mean average of 50.36% while Tigray and
Benishangul-Gumuz Regions achieved the least average scores of 32.95% and 32.44%
respectively. As to the performance levels of each region, five of the regions - Somali
(50.36%), Harari (42.46%), Addis Ababa (39.93%), Oromia (39.48%), and SNNPR
(38.93%) scored relatively better mean scores (above the national mean of 38.7%) while
all the remaining regions scored below the national average score ranging from 37.87%
to 32.44%. The one way analysis of variance showed that there is a significant
difference (F = (9, 13018), 72.204, P< .001) among the mean scores of the sample
regions.
Boys at the national level scored higher (40.55%) in the Grade 4 English test as
compared to girls (35.65%) showing a gap of 4.9%. This difference between male and

National Organization for Examinations 90

The Ethiopian Second National Learning Assessment Grade 4 Report

female students’ performance is significant at a P-value less than 0.001. The gender gap
in student achievement has also been observed in all the regions except in Harari where
female students scored slightly better than male students.
Nationally, students from urban and rural schools have achieved nearly the same mean
score (38.87%, and 38.58% respectively) in the Grade 4 English test. This indicates that
there was no statistically significant difference between the performance levels of urban
and rural students.
6. Student achievement results in Grade 4 mathematics was significantly less than the
national composite average for all subjects and there were disparities among regions
and between sexes.
The mean score achieved in Grade 4 mathematics test at the national level was 39.70%
with a standard deviation of 17.2. With regard to the performance levels of each region,
five of the regions: Addis Ababa (44.92%), Amhara (43.27%), Somali (43.09%), Harari
(40.71%), and Afar (40.29%) achieved relatively better mean scores (above the national
mean of 39.70%), and students from the SNNP Region scored (39.7%) which is equal to
the national average. The remaining 4 regions scored below the national average ranging
from 37.3% (in Oromia) to 34.34% (in Tigray and Benishangul-Gumuz regions). The
one way analysis of variance computed on the Grade 4 mathematics test showed that
there was a significant difference (F = (9, 13133), 42.22, P< .001) between the mean
scores of the sample regions.
With respect to gender parity, boys scored better (41.68%) than girls (36.67) in the
Grade 4 mathematics test with a gap of 5.01%. This difference between performances of
male and female students is significant at P-value less than 0.001. The gap has also been
observed in eight regions, except for the Somali Region where female students
performed better than their male counter parts while in Harari the observed gap was very
very small (0.49%).
Students from urban and rural schools have achieved differently in the Grade 4
mathematics test with average scores of 40.38% and 39.35% respectively. Though the
observed mean difference between urban and rural scores seems to be very small
(1.03%), the calculated statistical test shows that this difference was significant at P <
0.001 level.
7. Student achievement results in Grade 4 environmental science indicate that students
scored more than the national composite scores, but this was limited only to few
regions and boys.
The national mean average obtained in the Grade 4 environmental sciences was 51.74%
with a standard deviation of 17.41. In this test, the Amhara Region scored the highest
mean average of 60.05% while the Somali Region achieved the least average score of
45.83%.
Concerning regional performances, only 4 regions - Amhara (60.04%), Addis Ababa
(54.64%), Harari (53.35%), and Tigray (52.50%) scored relatively better mean scores
(above the national mean of 51.74%) while all the remaining six regions scored below
the national average score ranging from 51.60% (Afar) to 45.83% (Somali).
As to the observed gender gap in achievement at the national level, boys (53.43%)
scored better mean than girls (49.61%) with an achievement gap of 3.82% in the Grade 4
environmental science test. This level of difference in performance is tested (F (1, 12901), =
153.166, P < .001) as a significant deviation.

National Organization for Examinations 91

The Ethiopian Second National Learning Assessment Grade 4 Report

With regard to the comparison between school location and achievement, students from
both urban and rural areas achieved very similar mean scores of 51.49% and 51.87%
relatively, with only a 0.38% in favour of the rural schools. From this, it can be
concluded that there is no difference between the performance levels of urban and rural
students in the Grade 4 environmental science test at the national level.
8. Student achievement results in Grade 4 reading comprehension test indicate that
students scored more than the national composite scores, but this was limited only to
few regions and boys.
The national mean score achieved in the basic reading comprehension test was 64.5%
which is almost similar to the result of the Ethiopian Baseline National Learning
Assessment in the same subject (64.3%) with very similar standard deviations of
21.97 and 21.36, respectively. There is a statistically significant and positive
relationship (r = .472) between the scores of the basic reading comprehension test and
the school based measures in the mother tongue subjects of the same Grade level.
Boys at the national level scored high (65.71%) in the basic reading comprehension test
compared to what girls scored (63.00%). This difference between the performances of
male and female students is significant at P-value of less than 0.001. This implies that
there is a considerable amount (2.71%) of gender gap. This trend of gender gap in
students’ achievement is observed in the seven regions except for Afar, Harari and
Addis Ababa where female students scored slightly better than male students. The
gender gap observed in the reading comprehension test is most notable in 4 regions:
Oromia (with gender gap of 6.8%), SNNPR (6.3%), Tigray (5.8%), and Amhara (3.6%).
As to the performance levels of each region, Addis Ababa (76.90), Afar (71.92), Amhara
(70.92), and Harari (68.60) scored a relatively high mean score (above the national mean
of 64.49) while all the remaining regions scored below the national average score
ranging from 50.53 (lowest in Somali) to 63.73 (moderate in Oromia). It shall be
remembered that in the Baseline Assessment, the highest score in the Grade 4 basic
reading comprehension test was registered to the Amhara Region, but not this time.
9. The analysis of students’ background in relation to student achievement shows that
there was a relationship between academic scores and the student background
characteristics.
Students who travel longer distance from home to school achieved lower scores in
academic subjects than those who travel shorter distances. Those who spend more time
watching TV and listening to radio for leisure achieved lower scores than those who do
not spend their time with these media. Frequent absenteeism from school has been found
to relate to less achievement while getting sufficient amount of food was related to
higher achievement. Students who had their own textbooks were better achievers than
those who shared with others or who do not have textsbooks at all. Similarly those who
have shown interests in learning subjects were also the ones who scored better than who
did not have such interests.
10. The analysis of Grade 4 students’ general attitudes indicates that the country’s social
development curriculum is making a positive difference in shaping students’ attitudes
towards health care, environmental protection, civics and ethics, cultural issues and
education values.
In health care, the main issues were personal and environmental hygiene, spacing of
child birth, mode of transmission of HIV/AIDS and proper nutrition. In environmental
protection, students expressed positive inclinations towards drinking clean water,
preservation of historical heritages, planting of trees, environmental hygiene, and

National Organization for Examinations 92

The Ethiopian Second National Learning Assessment Grade 4 Report

prevention of drought and preservation of forests. In civics and ethics, the emphasis was
on equality, honesty, socialization, use of information and culture of dialogue. Items on
early marriage and female circumcision were included to see attitudes concerning
harmful cultural practices while absenteeism, student participation in school affairs,
learning and utilisation of personal earnings were some of the issues in which students
expressed positive attitudes.
11. The availability of curriculum materials has influential relationships with student
achievement.
The correlation analysis indicates that the availability of teacher’s guides and subject
syllabi had significant positive relations with student achievement in Grade 4. That is
the more these materials were available to students, the higher was their achievement.
Though not statistically significant, there is an indication that student textbook had
negative relations with achievements. That is the higher the number of students using
textbooks, the lower is student achievement.

12. Teacher’s personal characteristics and behavior have influences on student

achievement in different directions.
In the correlation analysis, teacher variables that positively and significantly relate to
student achievement were teacher’s perception of students’ learning ability/attitude,
teacher training on new teaching methods, teacher training on new assessment
techniques, and the average level of teacher training for teaching. On the other hand,
teacher’s age, qualifications, total teaching experience, distance from school and
social commitment were negatively correlated with student achievement, although
these relations were not statistically significant.
13. School management is another important dimension that influences student
achievement.
It has been found out that teachers taking classroom attendance, director’s experience
in school leadership, and how often the directors supervise teachers had significant
positive relations with student achievement in Grade 4.
14. The multiple regression analysis indicates that student home background and
behavior were the strongest determinants of academic achievement in Grade 4 in
Ethiopia.
The variables grouped under Student Background and Behavior explained the highest
portion (17.2%) of the observed variation in student achievement at school level
before other factors were considered and 5.7% after controlling or considering all
other factors in the model. This means, of the 28.7% variations explained by the final
model, Student Background and Behavior ranked highest in both cases of entry order
for analyses. This block contained five variables; among these, interest in the subject
matters, students’ sense of ownership of school property and alignment of home and
instructional language showed positive relationship with achievement while
absenteeism and time spent listening to the radio showed negative relationship.
15. School Organization and Management is the second strongest determinant for student
achievement in Grade 4.
School Organization and Management block explained 16.4% of the observed
variation in student achievement before controlling other factors and still a significant
6.0% with other factors controlled. This means, of the 27.8% variations explained by
the factors in the current study, School Organization and Management ranked as the

National Organization for Examinations 93

The Ethiopian Second National Learning Assessment Grade 4 Report

second higher in both cases of entry order for analysis. This block contained two
variables; namely, frequency of taking attendance by the teachers and directors’
managerial experience. Both factors showed positive relationship with student
achievement.
16. Teacher variables are the third most important determinants for student achievement
in Grade4.
Teachers' Variables explained the observed variation in achievement by 6.8% before
other factors were controlled and 2.9% after controlling all other factors. This means
of the 27.8% variations explained by the factors in the current study, teacher
variables ranked as the third higher in both cases of entry order for analysis. This
block contained variables which asked about teachers’ perception (judgment) of
students, the distance they traveled from home to school, and the use of the text
books. The distance the teacher traveled from home to school affected student
achievement negatively while the other two variables were related positively with
achievement.
17. School supply with reference to curriculum materials appeared as the fourth most
important determinant for student achievement.
School Supplies and Structures explained the observed variation in achievement by
2.1% before other factors were considered and by 0.4% after controlling all
otherfactors in the model. This block contained two variables namely availability and
use of syllabi and teacher’s guides. Both factors showed significant positive
relationships with achievement.
18. Instruction and support is the least important influential factor that determines
student achievement in Grade 4.
Finally the Instruction and Support block, which included factors concerning the role
of the director in supervising instructional activities, accounted for 2.6% of the total
variation in student achievement before other factors were controlled and 0.2% after
controlling all the other factors.

National Organization for Examinations 94

The Ethiopian Second National Learning Assessment Grade 4 Report

5.2 Recommendations
Based on the findings and implications of this study, the following specific
recommendations are given.
1. The overall performance of the education system as measured by student
achievement at Grade 4 needs improvement. The low level of overall achievement of
students requires to be improved at least to a reasonable level of 50% as planned by
MOE. Schools where students attain less than half of what they are expected to learn
are in serious trouble.
2. The overall performance of girls in Grade 4 needs improvement. The overall
achievement of girls requires to be improved in relation to that of boys. This is
required in all subjects. The current results indicate that concern for gender equity in
schooling needs to extend beyond access to the provision of quality education that
supports and encourages girls' achievement as well. Eventhough there is an indication
of increment in girls’ achievement since the baseline study, the gap between male and
female students achievement has widened from the baseline.
3. The performance of rural students in Grade 4 needs improvement. The
performance of rural schools in comparison to urban schools requires improvement.
Given that the majority of primary schools are rural, it is an alarming condition that
should be addressed since achievements are already low at the national level.
4. Performance of students in Grade 4 core curricula (English and Mathematics)
requires special attention. Mathematics and English achievement results were
extremely lower than the other subjects. Some regions and girls achieved less than
other regions and boys. These problems require solutions with reference to quality
and equity.
5. Disparity of performance by region and sex of students in other Grade 4 subjects
needs improvement. Although students’ achievement scores were more in the
national mean for environmental science and reading comprehension tests, they were
significantly favoring boys and some regions. Attention must be given to bringing the
equity among both sexes and regions.
6. The improvement of student achievement in Grade 4, first and foremost requires
improvement in school management. School management with particular reference
to teachers taking class attendance, director’s experience in school leadership, and
how often the directors supervise teachers showed significant positive relations with
student achievement in Grade 4. Therefore it is required to encourage teachers to take
class attendance, appoint and retain experienced school principals, and encourage
principal’s supervision of teachers.
7. Improvement of student achievement also depends on the improvement of student
home background and personal behavior. Student interest in subjects, meals per day,
agreement between home and instructional language, and students’ sense of
ownership in school property had significant positive relations with student
achievement while time spent by listening to radio and watching television, distance
from home to school and absence from school had significant negative relations with
student achievement. Therefore, improving student interest towards subjects,
improving the nourishment conditions, the use of home language in schools, and
student participation in school affairs need to be improved. Moreover, the
construction of nearby schools to reduce distance and discouraging wastage of time

National Organization for Examinations 95

The Ethiopian Second National Learning Assessment Grade 4 Report

by listening to radios and watching TV, and encouraging students’ classroom

attendance need special attention.
8. Better supply of school curriculum materials for Grade 4 needs attention to improve
student scholastic performance. This factor is one of strongest predictors of student
achievement in Grade 4. Among these, the availability of teacher’s guides and the
availability of the syllabus to teachers stands top in the degree of order. Therefore, it
is vital the teacher’s guides and subject syllabi are made available to schools.
9. special attention to teacher quality and behavior is the other priority area for
improvement of students’ learning in Grade 4. Teacher training on new teaching
methods and assessment techniques, average teacher training level, and teacher’s
perception of student learning attitudes are important conditions for the improvement
of student learning. Thus, organizing training programs on new methods of teaching,
conducting teacher training on new assessment techniques, improving the average
level of teacher training (pre-service), and encouraging teachers to learn about their
students are some of the measures required.
10. Strengthening the instructional practice of teachers and instructional support in the
classroom are vital to improve student learning in Grade 4. The director’s
perception of the overall instructional quality or his/her knowledge and judgment of
how teachers instruct their students has significant positive relation with student
achievement. This means both the quality of the teachers as instructional leaders and
the director’s position to acknowledge performance as an instructional leader have to
be strengthened. This can happen if the capacity of the principal and the teachers is
developed using different trainings or other programs.
11. Regional disparities in student achievements in Grade 4 need to be improved. There
are considerable variations in regional mean scores in each subject. Variation within
each region also tended to be high, with standard deviations ranging from 12 to 24 in
every subject. This means that regions that achieved less on the test scores need to
improve their learning outcomes as the responsibility to plan and implement primary
educations is theirs.
12. Learning from national learning assessment is required. Regular, reliable and
timely assessments are key to improve learning achievement. Such assessments
provide information about the level of performance of the education system at
different levels and the main factors that influence learning. Therefore, it is a major
requirement for all concerned bodies to learn from the current national learning
assessment and make sure that the recommended improvements are realized in the
years to come.

National Organization for Examinations 96

The Ethiopian Second National Learning Assessment Grade 4 Report

REFERENCES
Anbessu Biazen and Barbara Junge (1988). Problems in Primary School Participation
and Performance in Bahir Dar Awraja. Addis Ababa: Ministry of Education
and UNICEF.
Angelo, Tom. (May, 1999). Doing Assessment as if Learning Matters Most. American
Association for Higher Education Assessment Bulletin.
Brookover, W.B. and Lezotte, L.W. (1979). Changes in School Characteristics
Coincident with Changes in Student Achievement. East Lansing: Michigan
University.

Duncan, W. (1989). Engendering School Learning Science Attitudes and Achievement

among Girls and Boys in Botswana. Stockholm: Stockholm University.
Edmond, R..(1979). Effective Schools of the Urban Poor: Educational Leadership (35):
15-24
Ellet, C.D and Walber, H.J. (1979). Principals’ Competency, Environment and
Outcomes in Educational Environments and Effects. Berkeley, CA:McCutchan.
Fagerlind, I. and Saha, L. (1989). Education and National Development: A
Comparative Perspective. Oxford: Pergamon.
Fullan, M.G. (1990). Change Processes in Secondary Schools: Toward a More
Fundamental Agenda. In M.W. in McLauglin, J.E. Talbert and N. Basica (eds.).
The context of Teacher in Secondary schools: Teacher Realities. New
York:Teachers College Press.
Fuller, B. (1986). Raising School Quality in Developing Countries: What Investments
Boost Learning? Report No. EDT7: Education and Training Dept. World Bank .

Fuller, Bruce: (1987) Review of Educational Research What School Factors Raise
Achievement in Third World . The World Bank.
Greaney Vincent, and Thomas Kellaghan (1995) Equity Issues in Public Examinations
in Developing countries world Bank, D,C.
Heyneman, S.and Loxley.(1983). The Effect of Primary School Quality on Academic
Achievment across Twenty Nine High and Low Income Countries:
Himmel, E. (1984). National Assessment in Chile: National Assessments Testing the
System. Washington ,D.C: Economic Development Institute of the World Bank
Husen, T. (1967). International Studies of Achievement in Mathematicsematics: A
Comparison of 12 Countries. Stocklom: Almqvist and Wikshell International.
Kellaghan, T. and Greaney, V. (2001). Using Assessment to Improve the Quality of
Education. Paris: UNESCO

National Organization for Examinations 97

The Ethiopian Second National Learning Assessment Grade 4 Report

Kellaghan, T. and Greaney, V. (2004). Assessing Student Learning in Africa.

Washington, DC: World Bank.
Keller, D. (2004). Ethiopia: Microsoft ® Encarta ®Online Encyclopaedia http:// Encarta
msn.com
Klitgard, R.E and Hall, G. (1973). Are There Usually Effective Schools? Santa Monica,
CA:Rand. (ERIC.No.ED 085 402).
McDill, E. L. et al. (1967). Institutional Effects on the Academic Behavior of High
School Students, Sociology of Education, 40:181-99.
McDill,E. L. and Rigsby, L.C. (1973). Structure and Process in Secondary Schools.
Baltimore, MD: John Hopkins University Press.
Ministry of Education (1986). Evaluative Research of the General Education System
in Ethiopia, Summary Report. Addis Ababa: MOE.
MOFED (2002). Ethiopia: Sustainable Development and Poverty Reduction Program.
Addis Ababa: Ministry of Finance and Economic Development
NCES (2003). The Nation’s Report Card: Mathematics Highlights USA: National
Centre for Education Statistics..
NCES (2004). The Nation’s Report Card: The NAEP Guide.USA: National Centre for
Education Statistics.
Prawalbruk, K. (1996). National Assessment in Thailand: National Assessment Tesing
the System. Washington,D.C: Economic Development Institute of the World Bank
Parelius, R. J. and Parelius, A. P. (1987). The Sociology of Education. (2nd ed.). New
Jersey: Prentice-Hall Inc.
Puff, F. (1978). Instructional Variables and Student Achievement in Reading and
Mathematicsematics: A Synthesis of Recent Process-Produce Research.
Philadelphia: Research for Better Schools, Inc. (ERIC, No.ED 189 135).
Rosenshine, B.V. (1983). Teaching Functions in Instructional Programs. Elementary
School Journal 83:335-51

Seleshi Zeleke (1995). Gender Differences in Mathematics Achievement as a Function

of Attitudes in Grades 8 through 11 (in Northern Shewa Region). Masters thesis
Addis Ababa University, School of Graduate studies.
Serjiovanni, T.J. (1995). The Principalship. London: Allyn and Beacon.
Stockard, J. and Mayberry, M. (1992). Effective Educational Environment London:
Corwin Press, INC.
Tekeste Negash (1990).The Crises of Ethiopian Education: Some Implications for
Nation Building. Uppsala: Sweden

National Organization for Examinations 98

The Ethiopian Second National Learning Assessment Grade 4 Report

Thorndike,R.L.(1973). Reading Comprehension in Fifteen Countries: An Empirical

Study. New York: John Wiley.

UNESCO, (1971). Comparative Study of Secondary School Building Costs.

World Bank Groups (2003). Ethiopia Data Profile. http:// www.devdata world
bank.org

National Organization for Examinations 99