USAID Tusome Pamoja 
Midline Findings Report 
FINAL 
October 2018 
This document was produced for review by the United States Agency for International 
Development (USAID). It was made possible by the support of the American people through the 
USAID. It was prepared by RTI International for the USAID|Tanzania Tusome Pamoja Program.
USAID Tusome Pamoja
Midline Findings Report 
Contract Period: January 12, 2016 – January 11, 2021 
Contract Number: AID-621-C-16-00003 
Prepared for 
USAID|Tanzania 
United States Agency for International Development 
Office of Acquisition and Assistance 
ATTN: Laura Kikuli, Contracting Officer’s Representative 
 lkikuli@usaid.gov
 
686 Old Bagamoyo Road, Msasani 
P.O. Box 9130 
Dar es Salaam, Tanzania 
Prepared by 
RTI International 
3040 Cornwallis Road 
Post Office Box 12194 
Research Triangle Park, NC 27709-2194 
USAID Tusome Pamoja - Preliminary Midline Report iii 
Table of Contents
Page 
Executive Summary ........................................................................................................ 1
Key Findings ....................................................................................................... 1
1 Introduction ......................................................................................................... 2
1.1 USAID Tusome Pamoja Program Description and Key Indicators ........ 2
1.2 USAID Tusome Pamoja Key Interventions ............................................ 2
2 Midline Assessment Approach ........................................................................... 3
2.1 Overview of the Assessment .................................................................. 3
2.2 Overview of Instruments ......................................................................... 3
2.2.1 Midline Reading Assessment .... Error! Bookmark not defined.
2.2.2 Midline Mathematics Assessment ............ Error! Bookmark not 
defined.
2.3 Instrument Adaptation Process .............................................................. 5
2.4 Assessor Training ................................................................................... 5
2.5 Data Collection ........................................................................................ 5
3 Sample ................................................................................................................ 6
3.1 Population and List Frame ...................................................................... 6
3.2 Sample and Methodology ....................................................................... 6
3.3 Final Sample Counts and Estimates ...................................................... 7
4 Results and Findings .......................................................................................... 7
4.1 Reading Performance ............................................................................. 7
4.1.1 Reading performance indicators ................................................. 7
4.1.2 Reading performance categories ............................................... 8
4.1.3 MRA results in detail ................................................................... 9
4.1.4 Discussion of MRA results ........................................................ 10
4.2 Mathematics Performance .................................................................... 10
4.2.1 Mathematics performance indicators ....................................... 10
4.2.2 Detailed MMA results ................................................................ 11
4.2.3 Discussion of the MMA results ................................................. 12
5 Conclusion ........................................................................................................ 12
Annex A. Final Instruments........................................................................................... 14
Annex B. Sample Methodology, Weighting, and Precision .......................................... 24
Annex C. Inter-rater Reliability and Assessor Accuracy .............................................. 28
iv USAID Tusome Pamoja - Preliminary Midline Report
USAID Tusome Pamoja - Preliminary Midline Report v 
List of Figures 
Figure 1. Comparison of MMA Zero Scores by Subtask ............................................. 12
List of Tables 
Table 1. USAID Tusome Pamoja Baseline Indicator Values ......................................... 1
Table 2. MRA Instrument Subtasks in Tanzania ............................................................ 4
Table 3. MMA Instrument Subtasks in Tanzania ........................................................... 4
Table 4. Sample Methodology Summary ....................................................................... 6
Table 5. Sample Counts and Estimated Population Counts by Region ........................ 7
Table 6. Percentage of Children Meeting Reading Benchmarks ................................... 8
Table 7. Percentage of Children Meeting Reading Benchmarks by Gender ................ 8
Table 8. Reading Categories by Grade .......................................................................... 9
Table 9. Comparison of MRA Performance by Subtask ................................................ 9
Table 10. Overview of Midline MRA Performance by Subtask and Gender ............... 10
Table 11. Percentage of Children Meeting Mathematics Benchmark ......................... 11
Table 12. Percentage of Children Meeting Mathematics Benchmark by Gender ....... 11
Table 13. Student Performance on MMA Subtasks ..................................................... 11
Table B.1. Schools Excluded from the List Frame Prior to Sampling .......................... 24
Table B.2 Population Counts of Schools and Pupils by Tusome Pamoja Region ...... 24
Table B.3. Sample Methodology for the Tusome Pamoja Midline 2018 Assessment: 
Three-Stage Stratified Sample of Schools, Classrooms, and Pupils............... 25
Table B.4. Final Sample Counts of Assessments, Interviews, and Observations 
Collected for the Tusome Pamoja 2018 Midline Evaluation ............................ 26
Table B.5. Weighted Sample (Estimated Population) and Sample Counts by Each 
Stage Sampled (School, Classroom, Pupil) ..................................................... 26
Table B.6. Reading Fluency Means and Precision Estimate by Standard .................. 27
vi USAID Tusome Pamoja - Preliminary Midline Report
List of Abbreviations 
3Rs reading, writing, and arithmetic 
AAM assessor accuracy measure 
cwpm correct words per minute 
MMA Midline Mathematics Assessment 
MRA Midline Reading Assessment 
IRR inter-rater reliability 
MOEST Ministry of Education, Science, and Technology 
MOEVT Ministry of Education and Vocational Training 
ORF oral reading fluency 
USAID United States Agency for International Development 
USAID Tusome Pamoja - Preliminary Midline Report 1 
Executive Summary 
Key Findings 
The purpose of this midline evaluation is to measure the impact of the US Agency for 
International Development (USAID) Tusome Pamoja Program by comparing midline 
results against the baseline values reported in 2016. According to the Program’s 
Performance Monitoring and Evaluation Plan, there are three key reading and 
mathematics performance indicators to be reported for Standard 2 at midline: 
• P-001: Percent of learners who demonstrate reading fluency and 
comprehension of grade-level text at the end of Standard 2 with US 
Government (USG) assistance (oral reading fluency [ORF] 45 correct words 
per minute [cwpm]) 
• P-002: Number and proportion (%) of learners who, after two years of 
schooling, can solve grade-level arithmetic problems. As measured based on 
the following subtask scores: Addition and Subtraction (Level 2): 80%; and 
Missing Number: 60%. 
• P-008a: Number and proportion (%) of learners who, after two years of 
schooling, obtain zero scores on oral reading fluency (ORF: 0 cwpm) 
Table 1 lists the indicator values for P-001, P-002, and P-008a. The first column 
shows the definition of the indicator, while the second and third columns show the 
percentage of children meeting the benchmarks at baseline and midline for each 
indicator. 
Table 1. USAID Tusome Pamoja Baseline Indicator Values 
 
Percent of Children Meeting 
Benchmark 
 Baseline Midline 
P-001: Percent of learners who demonstrate reading 
fluency and comprehension of grade level text at the end of 
Standard 2 with USG assistance (ORF: 45 cwpm) 
1.5% 
(± 0.5%) 
3.5% 
(± 2.5%) 
P-002: Number and proportion (%) of learners who, after 
two years of schooling, can solve grade-level arithmetic 
problems (Addition/Subtraction Level 2: 80%, and Missing 
Number: 60%) 
1.4% 
(± 0.6%) 
3.9% 
(± 2.4%) 
P-008a: Number and proportion (%) of learners who, after 
two years of schooling, obtain zero scores on oral reading 
fluency (ORF: 0 cwpm) 
31.0% 
(±2.9) 
17.2% 
(±4.1) 
Improvements were made for all three indicators. While the gains for P-001 and P￾002 were modest, the results for P-008a show a largely significant and meaningful 
reduction in the percentage of students who were unable to read a single word at the 
end of Standard 2. This means that students are beginning to move out of the lowest 
categories of performance, but that it has been difficult to move a large proportion of 
the students into the highest performance category (which is to be expected based 
on the skewed distribution and the large proportion of students at the lower end).
2 USAID Tusome Pamoja - Preliminary Midline Report
1 Introduction 
1.1 USAID Tusome Pamoja Program Description and Key Indicators 
United States Agency for International Development (USAID)Tusome Pamoja is a 
five-year (2016–2021) USAID program designed to provide technical assistance and 
resource to the Government of the United Republic of Tanzania to strengthen the 
3Rs (reading, writing, and arithmetic) performance of Tanzanian learners in 
Standards 1–4 and to pilot pre-primary interventions. The Program is being 
implemented in four regions in Mainland Tanzania—Iringa, Morogoro, Mtwara, and 
Ruvuma—and in Zanzibar. 
USAID Tusome Pamoja’s primary goal is to improve lifelong learning skills, defined 
as mastery of early grade reading, writing, and arithmetic as measured by the 
percentage of children who: (i) after two years of schooling, can read and 
comprehend grade-level text, encode simple sentences, and solve grade-level 
arithmetic problems (measured in all regions); and (ii) after four years of schooling, 
can read and comprehend grade-level text, respond to simple writing prompts, and 
solve grade-level arithmetic problems; and by (iii) the proportion (%) of children 
obtaining zero scores on oral reading fluency (ORF) at the end of Standard 2 primary 
or the equivalent. An additional objective of the Program is to develop, implement, 
and demonstrate the best approaches to strengthen the quality of education in the 
target regions for replication consideration in other, non-target regions. Within the 
target regions, there are estimated to be in excess of 1.4 million children enrolled, 
taught by more than 36,000 teachers in approximately 3,035 schools. 
1.2 USAID Tusome Pamoja Key Interventions 
The three core results, (i) improving reading instruction in the primary grades; (ii) 
improving the policy environment for reading; and, finally, (iii) increasing parental and 
community support for student reading, are broken down to intermediate-result 
activities or interventions. These multiple, intermediate results are designed to 
contribute to the successful achievement of each result area. 
First, the Program will improve reading instruction in primary grades by providing 
quality instructional materials for 3Rs—reading, writing, and arithmetic—teaching, 
including student decodable texts, levelled classroom readers, read alouds, and 
teacher resource materials, as deemed necessary. Target teachers will be 
capacitated on effective use of classroom 3Rs teaching and learning materials 
through school-based professional development model, a community of learning 
approach, and continuous teacher professional development. Besides that, the 
school leadership (head teachers and ward education coordinators) will receive 
training and resource materials to support school-based teacher professional 
development, community engagement, school governance, planning, etc. Head 
teachers, ward education coordinators, quality assurance officers, district education 
officers, and district administrative officers will be trained in supervision and 
mentoring to provide overall support to teachers. 
Second, the Program intends to improve the policy environment for reading by 
supporting the continuous collection, storage, and utilization of data at all levels. This 
entails the introduction of a school information system (paper-based and digital) and 
 
USAID Tusome Pamoja - Preliminary Midline Report 3 
school scorecards, strengthening the education management information system in 
Zanzibar, and designing and carrying out various studies. In addition to these things, 
USAID Tusome Pamoja will establish and support school-based mentoring, 
communities of learning for teachers and mentors, and regional systems for early 
grade reading assessments (using lot quality assurance sampling techniques) that 
will happen periodically. 
Third, USAID Tusome Pamoja will increase parental and community support for 
pupils in lower grades to master the 3Rs. This will be achieved by establishing 
appropriate mechanisms to allow parents and community members to be involved in 
education matters. Such mechanisms include the development of parent–teacher 
partnerships, provision of school notice boards, capacity building of school 
management committees, and implementation of a social and behavior change 
communication campaign. 
2 Midline Assessment Approach 
2.1 Overview of the Assessment 
To establish baseline values against which to measure the impact of USAID Tusome 
Pamoja, a baseline assessment was conducted between August 15 and September 
2, 2016. The study targeted all Standard 2 children attending non-special needs 
education public schools within the five USAID Tusome Pamoja intervention regions 
and all Standard 4 children attending non-special needs education public schools in 
Zanzibar and Mtwara. 
In order to measure the impact of USAID Tusome Pamoja halfway through the 
Program, a midline assessment was conducted in August and September 2018. The 
midline assessment was designed to focus explicitly on changes in reading and 
mathematics performance for Standard 2 students across all intervention regions 
(though Standard 4 students were also tested for internal monitoring purposes). A 
shortened Midline Reading Assessment (MRA) was administered to measure oral 
reading fluency and reading comprehension, while an Midline Mathematics 
Assessment (MMA) was used to measure a range of mathematics skills, as 
discussed below. 
2.2 Overview of Instruments 
2.2.1 Midline Reading Assessment 
MRA measures the basic skills that a child must possess to eventually be able to 
read fluently and with comprehension—the ultimate goal of reading. The higher order 
skills (e.g., fluency and comprehension) build on lower order skills (e.g., letter sound 
knowledge, decoding), and the lower order skills have been shown to be predictive of 
later reading achievement. For the purposes of the midline, only the higher order 
skills were tested. 
Table 2 summarizes the subtasks of the MRA used in this study; the actual 
instruments are provided in Annex A. 
4 USAID Tusome Pamoja - Preliminary Midline Report
Table 2. MRA Instrument Subtasks in Tanzania 
Subtask Skill 
Description 
The child is asked to … 
Oral Reading Fluency Automatic and word recognition … read aloud a grade-level short story 
printed on a page. (Timed subtask) 
Reading Comprehension Comprehension 
… verbally respond to five oral questions 
(four literal and one inferential) that the 
assessor asks about the short story. 
(Untimed subtask) 
2.2.2 Midline Mathematics Assessment 
MMA gathers information about foundational mathematics competencies—those 
competencies that should typically be mastered during the very early grades, and 
without which children will struggle to succeed in mathematics or will potentially drop 
out. Subtasks selected for MMA are drawn from extensive research on early 
mathematics learning and assessment and were constructed by a panel of experts 
on mathematics education and cognition. The MMA protocol systematically assesses 
early numeracy skills, particularly those underlying number sense. These abilities 
and skills are key in the progression toward the ability to solve more advanced 
problems and the acquisition of more advanced mathematics skills. 
Table 3 summarizes the subtasks of the MMA used in this study; the actual 
instruments are provided in Annex A. 
Table 3. MMA Instrument Subtasks in Tanzania 
Subtask Skill 
Description 
The child is asked to … 
Addition and 
Subtraction (Level 1 
[basic facts]) 
This subtask requires knowledge 
of and confidence with basic 
addition and subtraction facts. It is 
expected that students should 
develop some level of automaticity 
and fluency with these facts 
because they need them 
throughout mathematics. 
… mentally solve addition and subtraction 
problems, with sums and differences below 
20. The problems ranged from those with only 
single digits to problems that involved the 
bridging of the 10. There were 10 items per 
addition and subtraction subtask. (Timed 
subtask) 
Missing Number 
(number patterns) 
This subtask requires the ability to 
discern and complete number 
patterns. 
… determine the missing number in a pattern 
of four numbers, one of which is missing. 
Patterns used included counting forward and 
backward by ones, fives, tens, and twos. 
There were 10 items. (Untimed subtask) 
 
USAID Tusome Pamoja - Preliminary Midline Report 5 
Subtask Skill 
Description 
The child is asked to … 
Addition and 
Subtraction (Level 2)a 
This subtask requires the ability to 
use and apply the procedural 
addition and subtraction 
knowledge assessed in the Level 1 
subtask to solve more complicated 
addition and subtraction problems. 
… solve addition and subtraction problems 
that involve the knowledge and application of 
the basic addition and subtraction facts 
assessed in the Level 1 subtask. Students 
were allowed to use any strategy that they 
wanted, including the use of paper and pencil 
supplied by the administrator. The problems 
extended to the addition and subtraction of 
two-digit numbers involving bridging. There 
were five items per addition and subtraction 
subtask. (Untimed subtask) 
a
 The Addition and Subtraction (Level 2) subtasks are more conceptual than the Addition and Subtraction (Level 
1) subtasks because children must understand what they are doing when applying the Level 1 skills. Although the 
(Level 2) subtasks are not purely conceptual, because, with time, children will develop some automaticity with the 
items in these subtasks, they are more conceptual than the Level 1 subtasks, especially so for Standard 2 
children.
2.3 Instrument Adaptation Process 
Language and mathematics experts from the Tanzania Institute of Education and 
ministry officials from the Ministry of Education and Vocational Training; Ministry of 
Education, Science, and Technology; and the Ministry of Education, Science, and 
Technology, along with RTI technical staff and project specialists contributed to the 
development of the USAID Tusome Pamoja MRA and MMA instruments during a 
two-day workshop on June 16 and 17, 2016, in Dar es Salaam. 
The instruments developed during the June 2016 meeting were largely adaptations 
of the instruments already developed for previous 2013 and 2016 national early 
grade reading and mathematics studies. 
The instruments used for this midline evaluation were shortened versions (i.e., 
reduced number of subtasks) of the instruments used at baseline. 
2.4 Assessor Training 
Assessor training was held for 46 assessors during a five-day workshop held August 
13th to August 17th, 2018, in Dar es Salaam. All assessors were trained to administer 
both the MRA and MMA instruments, as well as the brief accompanying 
questionnaires. Assessors participated in two assessor accuracy measures (AAM) 
that evaluated their ability to accurately score subtask items. Problematic items 
identified through the AAM exercise were addressed and improved upon during the 
training. Assessors who performed below the conventional threshold of 90% 
agreement received targeted assistance in order to improve their performance. The 
results of the AAM are presented in Annex C. 
2.5 Data Collection 
Data collection took place August 23rd to August 31st, 2018, following the assessor 
training activities. A total of 41 assessors were deployed to the five regions in which 
USAID Tusome Pamoja operates. Assessors were divided into teams comprised of 2 
to 3 assessors. General oversight and management of the data collection teams was 
provided by Tusome Pamoja monitoring and evaluation staff. Teams uploaded data 
6 USAID Tusome Pamoja - Preliminary Midline Report
regularly. Quality checks and feedback on uploaded data were provided by RTI home 
office statisticians. 
To supplement these data quality checks, the assessors in the field also performed 
daily inter-rater reliability (IRR) checks. For these checks, the assessors sampled 
one additional student. The student was given the MRA as normal with one assessor, 
while another assessor silently marked the students’ responses as well on his or her 
own tablet. Once the student was dismissed, the assessors compared their marked 
responses and discussed any discrepancies, without being able to change their 
original responses. On average, assessors agreed 97% of the time. As such, there 
were no concerns about the quality of the data being collected. The results of the 
field IRR checks are presented in Annex C. 
3 Sample 
3.1 Population and List Frame 
The population of interest was all Standard 2 children attending non-special needs 
education public schools within the five USAID Tusome Pamoja intervention regions. 
The list of schools in the 2016 National Examinations Council of Tanzania Primary 
School Leaving Certificate Examination was used as the sampling frame for the 
schools in Mainland Tanzania, and the 2016 census list of schools collected by the 
Ministry of Education and Vocational Training was used for Zanzibar. See Annex B
for details. 
3.2 Sample and Methodology 
The sample methodology calls for two three-stage samples of schools, classrooms, 
and children. Table 4 below provides a summary of the sample methodology used. 
Table 4. Sample Methodology Summary 
Stage 
Number Item Sampled Stratified by Probability of Selection 
Stage 1 Schools (90) Region (5) Proportion proportional to 
enrollment* 
Stage 2 
Standard 2 Classrooms 
(90) 
Standard 4 Classrooms 
(90) 
Standard (Standards 2 
and 4) Equal 
Stage 3 
Standard 2 Children 
(901) 
Standard 4 Children 
(894) 
Gender (Boys and Girls) 
Equal 
*Proportional to Enrollment: For the Mainland data, enrollment was the total number of children who sat for the 
Primary School Leaving Certificate Examination in 2016. For the Zanzibar data, it was the Standard 2 enrollment 
for the 2016 census data. 
The sample was derived to provide regionally representative estimates of children’s 
performance. The sample was designed to be able to report the estimated mean 
ORF within each grade with a 95% confidence interval band of about ±3.0 words per 
minute. 
 
USAID Tusome Pamoja - Preliminary Midline Report 7 
In the study it was found that all confidence intervals for ORF across the grades were 
tighter than ±3.0, with bands ranging from ±1.9 words per minute to ±2.4. 
3.3 Final Sample Counts and Estimates 
A total of 1,795 Standard 2 and Standard 4 children in all five regions were assessed 
in 90 schools overall. From each of these schools, a Standard 2 and Standard 4 
classroom was selected; the teacher of the selected classroom was interviewed; and 
the classroom inventory was conducted. The sampled counts and estimated 
population counts by region are presented in Table 5. 
Table 5. Sample Counts and Estimated Population Counts by Region 
Sample Counts Estimated Population Counts 
Schools 90 2,883 
Standard 2 classrooms 90 5,046 
Standard 4 classrooms 90 5,046 
Standard 2 children 901 277,102 
Standard 4 children 894 214,855 
Unless explicitly stated, all reported estimates were calculated using the appropriate 
sample weights because the sample weights adjust for any under- or over￾representation in the sample, making the estimates representative of the specified 
population. 
4 Results and Findings 
4.1 Reading Performance 
Children’s performance in reading was assessed using the MRA. The MRA results 
are presented in four parts: reading performance indicators, reading performance 
categories, detailed MRA results, and discussion of MRA results. 
4.1.1 Reading performance indicators 
The key purpose of this midline assessment was to measure the progress made by 
the USAID Tusome Pamoja Program since the baseline estimates in 2016. 
According to the Program’s Performance Monitoring and Evaluation Plan, there are 
two key performance indicators to be reported for Standard 2 reading: 
• P-001: Percent of learners who demonstrate reading fluency and 
comprehension of grade level text at the end of Standard 2 with USG 
assistance (ORF 45 cwpm) 
• P-008a: Number and proportion (%) of learners who, after two years of 
schooling, obtain zero scores on oral reading fluency (ORF: 0 cwpm) 
Accordingly, Table 6 reports the percentage of students for both of these indicators. 
The three columns of the table display the baseline and midline estimates along with 
the target that was set at baseline. 
8 USAID Tusome Pamoja - Preliminary Midline Report
Table 6. Percentage of Children Meeting Reading Benchmarks 
 Baseline Midline Target 
P-001: Percent of learners who demonstrate 
reading fluency and comprehension of grade level 
text at the end of Standard 2 with USG assistance 
(ORF: 45 cwpm) 
1.5% 
(± 0.5%) 
3.5% 
(± 2.5%) 6% 
P-008a: Number and proportion (%) of learners 
who, after two years of schooling, obtain zero 
scores on oral reading fluency (ORF: 0 cwpm) 
31.0% 
(±2.9) 
17.2% 
(±4.1) 20% 
While there was an increase in the percentage of students reading at benchmark (up 
to 3.5% from 1.5%), it was not large enough to meet the 6% target explicitly (though 
the target was within the estimate’s confidence band). For non-readers, on the other 
hand, the percentage was reduced to 17.2%—which is below the 20% target. 
The results for the two reading indicators, disaggregated by gender, are displayed in 
Table 7. This table shows that the margin between girls and boys remained similar 
from baseline to endline for indicator P-001 (approximately 1% difference) but that 
boys narrowed the gap in terms of zero scores. Specifically, boys reduced their 
percentage of zero scores by 16.5%, while girls reduced theirs by about 11%, leading 
to a closing of the gap between the two estimates. 
Table 7. Percentage of Children Meeting Reading Benchmarks by Gender 
 Baseline Midline 
Boys Girls Boys Girls 
P-001: Percent of learners who demonstrate reading 
fluency and comprehension of grade level text at the 
end of Standard 2 with USG assistance (ORF: 45 
cwpm) 
1.0% 
±0.5% 
2.0% 
±0.8% 
2.8% 
±3.2% 
4.1% 
±3.3% 
P-008a: Number and proportion (%) of learners who, 
after two years of schooling, obtain zero scores on 
oral reading fluency (ORF: 0 cwpm) 
34.5% 
±3.4% 
27.6% 
±3.5% 
18% 
±4.5% 
16.4% 
±5.7% 
4.1.2 Reading performance categories 
In order to further understand the shift in student performance from baseline to 
midline, the MRA results were combined into a single, four-category measure of 
readers with regard to their performance on the oral reading fluency and reading 
comprehension subtasks: Non-Readers, Beginning Readers, Progressing Readers, 
and Proficient Readers (Table 8). Proficient Readers were those who were able to 
correctly read 45 or more words of the story in one minute and with 80% or more 
comprehension. Conversely, Non-Readers were students who were unable to read a 
single word of the reading passage. Beginning Readers were correctly reading 
between 1 and 29 words per minute, and Progressing Readers were correctly 
reading at least 30 correct words per minute. Table 8 shows that the large reduction 
in Non-Readers was spread throughout the ensuing three categories (with the largest 
change in Beginning Readers (~7%), followed by Progressing Readers (~4%), and 
finally Proficient Readers (~2%). This shows that significant progress is being made, 
 
USAID Tusome Pamoja - Preliminary Midline Report 9 
but that it will take more time to be able to move a larger proportion of students from 
the lowest categories to the highest. 
Table 8. Reading Categories by Grade 
Category Type of Reader Characteristic Baseline Midline 
1 Non-Readers Unable to read a single word of the 
story 31.0% 17.2% 
2 Beginning 
Readers 
Read between 1 and 29 words of the 
story correctly in one minute 57.3% 64.7% 
3 Progressing 
Readers 
Read at least 30 words of the story 
correctly in one minute 10.4% 14.8% 
4 Proficient 
Readers 
Read at least 45 words of the story 
correctly in one minute and with 80% 
or more comprehension 
1.2% 3.4% 
4.1.3 MRA results in detail 
Detailed results from all of the MRA subtasks, comparing baseline and midline, are 
displayed in Table 9. This table provides three measures for these subtasks: 1) 
percentage of children who were unable to correctly identify a single item; 2) average 
fluency scores (items per minute) for oral reading; 3) the average scores as 
measured by the percentage of correct items out of all the reading comprehension 
items. It is clear from this table the largest impacts are in terms of reducing zero 
scores for oral reading. However, reading comprehension zero scores also dropped 
about 9%, while mean scores (average percent correct) increased by approximately 
the same amount. This shows that performance in terms of oral reading fluency and 
reading comprehension has significantly improved since baseline. 
Table 9. Comparison of MRA Performance by Subtask 
MRA Subtask 
Percentage of Students 
Who Scored Zero Average Fluency Score Average % Correct of 
Items 
Baseline Midline Baseline Midline Baseline Midline 
Oral Reading 
Fluency 31% (±2.9) 17.2% 
(±4.1) 
12.9 correct 
words per 
minute 
(±0.9) 
16.4 correct 
words per 
minute 
(±1.9) 
– – 
Reading 
Comprehension 
43.2% 
(±3.1) 
34.4% 
(±5.1) – – 
24.6% 
(±1.7) 
33.1% 
(±3.7) 
The midline MRA results, disaggregated by gender, are displayed in Table 10. It is 
clear from this table that girls slightly outperformed boys in oral reading and reading 
comprehension (with statistically significant differences in oral reading fluency and 
reading comprehension percent correct). 
10 USAID Tusome Pamoja - Preliminary Midline Report
Table 10. Overview of Midline MRA Performance by Subtask and Gender 
MRA 
Subtask 
Percentage of Students 
Who Scored Zero Average Fluency Score Average % Correct of 
Items 
Boys Girls Boys Girls Boys Girls 
Oral Reading 18.0% 
(±4.5) 16.4% (±5.7) 
15 correct 
words per 
minute 
(±2.1) 
17.8 correct 
words per 
minute (±2.2) 
– – 
Reading 
Comprehension 
38.8% 
(±7.4) 30.2% (±6.6) – – 
30.7% 
(±4.8) 35.5% (±4.3) 
4.1.4 Discussion of MRA results 
Student performance on the MRA subtasks indicates that there was improvement 
from baseline to midline, but there is a continued need for even greater improvement 
in order to meet all targets by endline. While it is possible for students to learn to read 
a passage at the Standard 2 difficulty level with fluency and comprehension, the 
proportions of readers at this level remain low. This is largely a result of the starting 
ability level of students at baseline. In other words, with so few Progressing Readers 
it is difficult to significantly increase the proportion of Proficient Readers. At the other 
end of the spectrum, however, USAID Tusome Pamoja has shown a significant 
decrease in the proportion of Non-Readers (to a level that exceeds the midline 
target). 
4.2 Mathematics Performance 
Performance in mathematics was assessed using the MMA described in detail in 
Section 3.2. The MMA results are presented in three parts: mathematics 
performance indicators, detailed MMA results, and discussion of the MMA results. 
4.2.1 Mathematics performance indicators 
USAID Tusome Pamoja only has one reporting indicator for mathematics. In order to 
define this indicator, the following calculations were used: 
• P-002: Number and proportion (%) of learners who, after two years of 
schooling can solve grade-level arithmetic problems: 
− Addition and Subtraction (Level 2): 80%; and 
− Missing Number: 60% 
The results for indicator P-002 are displayed in Table 11. The table shows that there 
was growth from baseline (1.4%) to midline (4.3%), but that these improvements fell 
shy of the 6% target explicitly (though were once again within the estimate’s 
confidence band). It is promising to note that some improvements were made, 
considering that USAID Tusome Pamoja has only just begun the mathematics 
portion of its intervention. 
 
USAID Tusome Pamoja - Preliminary Midline Report 11 
Table 11. Percentage of Children Meeting Mathematics Benchmark 
 Baseline Midline Target 
P-002: Number and proportion (%) of learners who, 
after two years of schooling, can solve grade-level 
arithmetic problems (Addition/Subtraction Level 2: 
80%, and Missing Number: 60%) 
1.4% 
(± 0.6%) 
4.3% 
(± 2.4%) 6% 
In Table 12 these results are disaggregated by gender. The table suggests that there 
was no gap at baseline, but that slightly more boys were meeting the benchmark at 
midline than were girls. However, it should also be noted that this difference was not 
statistically significant, as it was well within the margin of error of the estimates. 
Table 12. Percentage of Children Meeting Mathematics Benchmark by Gender 
 Baseline Midline 
Boys Girls Boys Girls 
P-002: Number and proportion (%) of learners who, 
after two years of schooling, can solve grade-level 
arithmetic problems (Addition/Subtraction Level 2: 
80%, and Missing Number: 60%) 
1.4% 
±0.8% 
1.5% 
±0.9% 
5.1% 
±3.7% 
3.6% 
±2.9% 
4.2.2 Detailed MMA results 
Complementing the mathematics performance indicator, this section provides an 
overview of the results for all MMA subtasks. Table 13 summarizes baseline and 
midline performance on MMA. This shows that while there were small improvements 
in the rates of the most basic tasks (Addition and Subtraction Level 1), there were 
significantly larger improvements in Addition and Level 2 (for which the mean scores 
more than doubled). 
Table 13. Student Performance on MMA Subtasks 
Subtask Assessment Mean Score 
Addition (Level 1) 
Baseline 6.6 per minute (±0.3) 
Midline 7.5 per minute (±0.7) 
Subtraction (Level 1) 
Baseline 4.9 per minute (±0.3) 
Midline 5.7 per minute (±0.7) 
Missing number 
Baseline 23.3% (±1.0) 
Midline 28.1% (±2.6) 
Addition (Level 2) 
Baseline 16.7% (±1.7) 
Midline 38.8% (±4.4) 
Subtraction (Level 2) 
Baseline 10.5% (±1.4) 
Midline 32.8% (±6.0) 
While mean scores provide an important estimate for average performance, Figure 1 
summarizes the change in performance at the lower end of MMA subtasks (i.e., 
12 USAID Tusome Pamoja - Preliminary Midline Report
reductions in zero scores). It is clear from this figure that progress was made on all 
subtasks, but that the largest reductions in zero scores came in the subtasks of 
Addition Level 2 and Subtraction Level 2. 
Figure 1. Comparison of MMA Zero Scores by Subtask 
4.2.3 Discussion of the MMA results 
There was improvement across all MMA subtasks from baseline to midline. In terms 
of the mathematics reporting indicator, the percentage of students at benchmark 
increased to 4.3%. While this was not equal to the target set at baseline, this level of 
improvement is promising given the fact that the 2019 school year will be the first 
with a comprehensive mathematics intervention across all regions. Additionally, the 
results show that students are making particularly good progress in the higher-order 
skills (such as Addition and Subtraction Level 2). Considering that doing mathematics 
(especially in the early grades) involves being able to read, write, compare, and 
perform basic operations with numbers, it seems that some of these conceptual skills 
may have benefited from the increased reading ability of students. 
In terms of gender, boys performed slightly better than girls, but overlapping 
confidence intervals for all of these tasks makes it impossible to claim that 
differences in performance are statistically meaningful. 
It is recommended that USAID Tusome Pamoja in-service teacher training in 
mathematics focus more on the multi-dimensional nature of mathematical proficiency 
than on more efficient teaching strategies that focus on mathematics as the 
memorization of facts, rules, and procedures. 
5 Conclusion 
The 2018 midline assessment was designed to accomplish one goal: 
• Determine the level of impact from the USAID Tusome Pamoja Program 
across reading and mathematics indicators for Standard 2 by comparing 
program-level estimates from the 2016 baseline and the 2018 midline. 
11%
22%
63%
32%
73%
8%
14%
33%
20%
36%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Missing Number Basic Addition Addition Level 2 Basic
Subtraction
Subtraction
Level 2
Percentage of Zero Scores
Baseline Midline
 
USAID Tusome Pamoja - Preliminary Midline Report 13 
The results of the USAID Tusome Pamoja midline assessment show that gains were 
made across all indicators and subtasks. While the gains for the oral reading fluency 
and mathematics benchmark indicators were apparent, they were not large enough 
to meet the midline targets (though they were not statistically different from them 
based on the confidence intervals for the estimates). While some of the explanation 
for the lower than expected performance may be the result of increased strains on 
resources to due large increases in student enrollment (average Standard 2 
enrollment in sampled school at baseline was 66 students, while it was 117 students 
at midline), it should also be noted that larger than expected gains were made at the 
lower end of the performance spectrum (i.e., in reducing zero scores). This means 
that students are beginning to move out of the lowest categories of performance, but 
that it has been difficult to move a large proportion of the students into the highest 
performance category (which is expected based on the skewed distribution and the 
large proportion of students at the lower end). 
Reading results indicate that it is clear children need more practice to advance from 
laborious, word-by-word reading to reading in meaningful phrases; mathematics 
results show that there must be a shift in teaching from teaching for memorization of 
facts, rules, and formulas to teaching for understanding
14 USAID Tusome Pamoja - Preliminary Midline Report
Annex A. Final Instruments 
Midline Mathematics Assessment – Tusome Pamoja Midline 2018 
Maelekezo kwa ujumla 
Fanya utangulizi/utambulisho rafiki na mwanafunzi kwa namna ya mchezo kwa 
mazungumzo mafupi kati yenu. (ona mfano hapo chini). . Mwanafunzi anatakiwa ahisi 
maswali anayoulizwa ni kama vile anacheza badala ya kujihisi anatahiniwa. Tumia muda huu 
kumuuliza ni lugha ipi atakuwa huru kuitumia. Soma taratibu kwa sauti na kwa ufasaha 
maelezo yaliyomo NDANI ya kisanduku tu. 
Establish a playful and relaxed rapport with the child through a short conversation. The 
child should perceive the assessment almost as a game to be enjoyed rather than a test. 
Use this time to identify in what language the child is most comfortable communicating. Read 
aloud slowly and clearly ONLY the sections in boxes. 
Habari ya asubuhi. Jina langu ni __ naishi _ . Napenda nikueleze kidogo habari 
zangu. [Eleza idadi ya watoto ulionao, umri wao, mchezo unaoupenda, kipindi cha redio 
au luninga unachopendelea, n.k.] Good morning. . My name is ____ and I live in _____. 
I’d like to tell you a little bit about myself. 
[Number and ages of children; favourite sport, radio or television program, etc.] 
1. Unapenda kufanya nini unapotoka shuleni? [ Subiri jibu lake; kama mwanafunzi 
hataki kusema lolote, uliza swali la pili, lakini kama anaonekana angependa kuendelea 
kuongea basi endelea kupata ridhaa yake kwa maneno] 
What do you like to do when you are not in school? 
[Wait for response; if student is reluctant, ask question 2, but if they seem comfortable 
continue to verbal consent]. 
2. Unapenda kucheza michezo ipi? What games do you like to play?
A.Tarehe ya 
tathmini
 Date of 
assessment : 
F. Darasa la 
mwanafunzi: 
Student’s grade level : 
○ 1 = Darasa la 1
 Standard 1 
○ 2 = Darasa la 2 
Standard 2 
○ 3 = Darasa la 3
 Standard 3 
○ 4 = Darasa la 4
 Standard 4 
B. Jina la mtafiti/ 
/namba: 
Assessor 
name/code : 
C. JINA la shule na 
mahali ilipo: 
 NAME and 
location of 
school : 
G. Umri la 
mwanafunzi: 
Student’s age: 
Umri/Age: ___________
D. Namba maalum 
ya Shule:
Database ID: ________
Namba ya utambuzi: ________ 
 
USAID Tusome Pamoja - Preliminary Midline Report 15 
 Unique School 
code : 
E. Namba maalum 
ya Mwanafunz:
 Unique student 
code : 
H. Jinsi ya 
Mwanafunzi: 
Student’s gender : 
○ msichana
/girl 
○ mvulana/
boy 
16 USAID Tusome Pamoja - Preliminary Midline Report
Kazi 1: Namba inayokesekana: Zoezi kwa vitendo / Missing 
number: Practice 
 A1 

 Hapa pana baadhi ya namba 5,10,15. Namba gani itaingia hapa? 
 Here are some numbers. 5, 10, and 15. What number goes here? 


5 10 15 (20)

 Vizuri ni 20. Tuendelee na swali jingine. 
That’s correct, 20. Let’s do some more. 

 Namba 20 itakuwa hapa katika mfululizo huu. Tusome namba hizi kwa 
pamoja. [Onesha kila namba] . . . 5, 10, 15, 20. Tuendelee na swali 
jingine.
The number 20 goes here. Say the numbers with me. [Point to each 
number] 5, 10, 15, 20. 20 goes here. Let’s do some more. 
Kazi 1: Namba inayokosekana / Missing number 
 A2 & A3 

 Hapa kuna namba zaidi. [Onesha katika sanduku] ni namba gani 
inaingia hapa? [Rudia kwa kila namba.]
 Here are some more numbers. [Point to the box] What number goes 
here? [Repeat for each item.] 

• Endapo 
mwanafunzi 
amekosea mara 4 
mfululizo. / If the 
child makes 4 
successive errors 

• Endapo 
mwanafunzi 
hajajibu swali 
baada ya sekunde 
5.
 If the child doesn’t 
respond after 5 
seconds. 
 () 1 = Sahihi / Correct 
 () 0 = Jibu sio sahihi au halijajibiwa / Incorrect or no response 
1 6
 
5 6 7 (8) 1 0 318 319 (320) 321 1 0
2 7
 
13 14 (15) 16 1 0 28 (26) 24 22 1 0
3 8
 
20 (30) 40 50 1 0 45 50 (55) 60 1 0
4 9
(300) 400 500 600 1 0 550 540 530 (520) 1 0
5 10 
2 4 6 (8) 1 0 9 14 (19) 24 1 0
 
USAID Tusome Pamoja - Preliminary Midline Report 17 
Kazi 2a:Kujumlisha: Hatua 1 / Addition: Level 1 
 B1 & B2  Sekunde 60 
seconds

 Hapa kuna maswali ya kujumlisha [Onesha kwa mkono kutoka juu 
hadi chini]. 
 Nitakwambia wakati wa kuanza na wa kumaliza. Toa majibu kwa 
kila swali na kama hujui jibu, endelea na swali linalofuata. Upo 
tayari? 
 Anzia hapa [Onesha swali la kwanza]. 
Here are some addition problems [Glide hand from top to bottom]. 
 I will tell you when to start and when to stop. Say the answer for 
each problem. If you don’t know an answer, move to the next 
problem. Are you ready? 
 Start here [Point to first problem].

• Endapo muda 
umekwisha (sekunde 
60). / If the time on 
the stopwatch runs 
out (60 seconds).

• Endapo mwanafunzi 
hajajibu swali baada 
ya sekunde 5. / If the 
child doesn’t respond 
to an item after 5 
seconds. 
 ( / ) Jibu sio sahihi au halijajibiwa / Incorrect or no response
 ( ] ) Baada ya swali la mwisho kusomwa/ After the last number read
4 + 1 = (5) 12 + 3 = (15) 
2 + 4 = (6) 13 + 4 = (17) 
6 + 2 = (8) 11 + 6 = (17) 
4 + 5 = (9) 6 + 5 = (11) 
3 + 3 = (6) 10 + 4 = (14) 
8 + 1 = (9) 9 + 7 = (16) 
7 + 3 = (10) 6 + 6 = (12) 
2 + 7 = (9) 5 + 6 = (11) 
5 + 5 = (10) 8 + 7 = (15) 
1 + 9 = (10) 8 + 10 = (18) 
 Muda uliobaki (sekunde) / Time left (seconds): 
18 USAID Tusome Pamoja - Preliminary Midline Report
Kazi 2b: Kujumlisha: Hatua 2 /Addition: Level 2 
 B3 
Karatasi na kalamu / Paper and pencil 

 Hapa kuna maswali ya kujumlisha [Onesha kwa mkono kutoka 
juu hadi chini]. 
 Nitakwambia wakati wa kuanza na wa kumaliza. Toa majibu 
kwa kila swali na kama hujui jibu, endelea na swali linalofuata. 
Upo tayari? 
 Anzia hapa [Onesha swali la kwanza]. 
Here are some addition problems [Glide hand from top to 
bottom]. 
 I will tell you when to start and when to stop. Say the answer for 
each problem. If you don’t know an answer, move to the next 
problem. Are you ready? 
 Start here [Point to first problem].

• Endapo mtoto hajajibu 
swali lolote la hatua ya 
kwanza kwa usahihi. . 
If the child did not 
answer any Level 1 
question correctly.
• Endapo mwanafunzi 
amekosea mara 4 
mfululizo. . 
If the child makes 4 
consecutive errors.

• Iwapo mwanafunzi 
atatumia njia 
isiyoridhisha (Mf; 
kuchora chora vimstari) 
muulize mwanafunzi 
kama anaweza kutumia 
njia nyingine? / If the 
child uses an inefficient 
strategy (e.g., tick 
marks), ask the child “Do 
you know another way 
to solve the problem?”
• Endapo mwanafunzi 
ataendelea kutumia njia 
isiyosahihi au 
atasimama/ atakwama 
kwa sekunde tano 
katika swali hilo. . 
If a child continues to 
use an inefficient 
strategy or stops on an 
item for 5 seconds.
 () 1 = Sahihi / Correct
 () 0 = Jibu sio sahihi au halijajibiwa / Incorrect or no response
12 + 6 = (18) *1* *0*
18 + 7 = (25) *1* *0*
11 + 27 = (38) *1* *0*
25 + 35 = (60) *1* *0*
36 + 27 = (63) *1* *0*
Mwanafunzi / The pupil:
** Ametumia vidole au kutali/ Used fingers or tallies 
 
** Ametumia karatasi na penseli / Used paper & pencil 
 
** Amejibu maswali kwa kichwa /Solved the problem(s) in his/her 
head
 
 
USAID Tusome Pamoja - Preliminary Midline Report 19 
Kazi 3A: Kutoa: Hatua ya 1 / Subtraction: Level 1 
 C1 & C2  Sekunde 60/60
seconds
Standards 1, 2, 3 and 4 

 Hapa kuna maswali ya kutoa [Onesha kwa mkono kutoka juu hadi 
chini].
 Nitakwambia wakati wa kuanza na wa kumaliza. Toa majibu kwa 
kila swali na kama hujui jibu, endelea na swali linalofuata. Upo 
tayari? 
 Anzia hapa [Onesha swali la kwanza]. 
Here are some subtraction problems [Glide hand from top to 
bottom]. 
 I will tell you when to start and when to stop. Say the answer for 
each problem. If you don’t know an answer, move to the next 
problem. Are you ready? 
 Start here [Point to first problem].

• Endapo muda 
umekwisha (sekunde 
60). . 
If the time on the 
stopwatch runs out 
(60 seconds).

• Endapo mwanafunzi 
hajajibu swali baada 
ya sekunde 5. . 
If the child doesn’t 
respond to an item 
after 5 seconds.  ( / ) Jibu sio sahihi au halijajibiwa / Incorrect or no response
 ( ] ) Baada ya swali la mwisho kujibiwa / After the last number read
3 – 2 = (1) 12 – 1 = (11) 
5 – 3 = (2) 15 – 3 = (12) 
6 – 2 = (4) 17 – 4 = (13) 
9 – 5 = (4) 14 – 4 = (10) 
8 – 4 = (4) 16 – 10 = (6) 
9 – 1 = (8) 13 – 5 = (8) 
10 – 4 = (6) 14 – 5 = (9) 
9 – 7 = (2) 11 – 7 = (4) 
10 – 5 = (5) 13 – 8 = (5) 
10 – 9 = (1) 16 – 8 = (8) 
 Muda uliobaki (sekunde) / Time left (seconds): 
20 USAID Tusome Pamoja - Preliminary Midline Report
Kazi 3b: Kutoa: Hatua ya 2 / Subtraction: Level 2 
 C3 
 Karatasi na kalamu /Paper and pencil 

 Hapa kuna maswali ya kutoa [Onesha kwa mkono kutoka juu 
hadi chini].
 Nitakwambia wakati wa kuanza na wa kumaliza. Toa majibu 
kwa kila swali na kama hujui jibu, endelea na swali linalofuata. 
Upo tayari? 
 Anzia hapa [Onesha swali la kwanza]. 
Here are some subtraction problems [Glide hand from top to 
bottom]. 
 I will tell you when to start and when to stop. Say the answer for 
each problem. If you don’t know an answer, move to the next 
problem. Are you ready? 
 Start here [Point to first problem].

• Endapo mtoto hajajibu 
swali lolote la hatua ya 
kwanza kwa usahihi. . 
If the child did not 
answer any Level 1 
question correctly.
• Endapo mwanafunzi 
amefanya makosa 
manne (4) mfululizo. . 
If the child makes 4 
consecutive errors. 

• Iwapo mwanafunzi 
atatumia njia 
isiyoridhisha (Mf; 
kuchora chora vimstari) 
muulize mwanafunzi 
kama anaweza kutumia 
njia nyingine? / If the 
pupil uses an inefficient 
strategy (e.g., tick 
marks), ask the child “Do 
you know another way 
to solve the problem?” 
• Endapo mwanafunzi 
ataendelea kutumia njia 
isiyosahihi au 
atasimama/ atakwama 
kwa sekunde 5 katika 
swali hilo. . 
If a pupil continues to 
use an inefficient 
strategy or stops on an 
item for 5 seconds.
 () 1 = Sahihi / Correct
 () 0 = Jibu sio sahihi au halijajibiwa / Incorrect or no response
18 – 4 = (14) *1* *0*
25 – 7 = (18) *1* *0*
36 – 12 = (24) *1* *0*
40 – 17 = (23) *1* *0*
46 – 29 = (17) *1* *0*
Mwanafunzi / The pupil:
** Ametumia vidole au kutali/ Used fingers or tallies 
 
** Ametumia karatasi na penseli / Used paper & pencil 
 
** Amejibu maswali kwa kichwa /Solved the problem(s) in his/her 
head
 
 
USAID Tusome Pamoja - Preliminary Midline Report 21 
Midline Reading Assessment – Tusome Pamoja Midline 2018 
Maelekezo kwa ujumla 
Fanya utangulizi/utambulisho rafiki na mwanafunzi kwa namna ya mchezo kwa 
mazungumzo mafupi kati yenu. (ona mfano hapo chini). Mwanafunzi anatakiwa ahisi 
maswali anayoulizwa ni kama vile anacheza badala ya kujihisi anatahiniwa. Tumia muda 
huu kumuuliza ni lugha ipi atakuwa huru kuitumia. Soma taratibu kwa sauti na kwa ufasaha 
maelezo yaliyomo NDANI ya kisanduku tu. 
Establish a playful and relaxed rapport with the child through a short conversation. The 
child should perceive the assessment almost as a game to be enjoyed rather than a test. 
Use this time to identify in what language the child is most comfortable communicating. Read 
aloud slowly and clearly ONLY the sections in boxes. 
Habari ya asubuhi. Jina langu ni __ naishi _ . Napenda nikueleze kidogo habari zangu. 
[Eleza idadi ya watoto ulionao, umri wao, mchezo unaoupenda, kipindi cha redio au luninga 
unachopendelea, n.k.] Good morning. My name is ____ and I live in _____. I’d like to tell you a 
little bit about myself. 
[Number and ages of children; favourite sport, radio or television program, etc.] 
1. Unapenda kufanya nini unapotoka shuleni? [ Subiri jibu lake; kama mwanafunzi hataki 
kusema lolote, uliza swali la pili, lakini kama anaonekana angependa kuendelea kuongea basi 
endelea kupata ridhaa yake kwa maneno] 
What do you like to do when you are not in school? 
[Wait for response; if student is reluctant, ask question 2, but if they seem comfortable continue 
to verbal consent]. 
2. Unapenda kucheza michezo ipi? What games do you like to play?
A.Tarehe ya 
tathmini
 Date of 
assessment :
F. Darasa la 
mwanafunzi: 
Student’s grade level : 
○ 1 = Darasa la 1
 Standard 1 
○ 2 = Darasa la 2 
Standard 2 
○ 3 = Darasa la 3
 Standard 3 
○ 4 = Darasa la 4
 Standard 4
B. Jina la mtafiti/ 
/namba: 
Assessor 
name/code : 
C. JINA la shule na 
mahali ilipo: 
 NAME and 
location of 
school : 
G. Umri la 
mwanafunzi: 
Student’s age: 
Umri/Age: ___________ 
D. Namba maalum 
ya Shule:
 Unique School 
code :
E. Namba maalum 
ya Mwanafunz:
 Unique student 
code : 
H. Jinsi ya 
Mwanafunzi: 
Student’s gender : 
○ msichana
/girl 
○ mvulana/
boy 
22 USAID Tusome Pamoja - Preliminary Midline Report
Subtask 1a. ORAL READING 
PASSAGE 

 A1 

Sekunde 
60 
seconds
Subtask 1b. READING 
COMPREHENSION 
1. Iwapo mwanafunzi atashindwa kusoma japo neno moja kwa usahihi kabla ya neno 
lililomo katika kisanduku sema “Asante!”, sitisha zoezi hili, kisha weka alama kwenye 
kisanduku kilichopo hapa chini kisha endelea na zoezi linalofuata. Usimuulize maswali 
ya ufahamu. If the child does not provide a single correct word before the word in a box, say 
“Thank you!”, discontinue this subtask and check the box at the bottom. Do not ask any 
comprehension questions. 2. Iwapo mwanafunzi atasema “Sijui”, chukulia kama ni kosa. If 
a child says, “I don’t know,” mark as incorrect.
Muoneshe mwanafunzi karatasi 
yenye hadithi wakati unasoma 
maelekezo. 
Show the child the sheet in the 
student stimulus booklet as you read 
the instructions.
 3. 
Endapo 
mtoto 
hajajibu 
swali 
baada ya 
sekunde 
tatu. 
If the 
child 
doesn’t 
respond to 
an item 
after 3 
seconds. 
Baada ya mwanafunzi kumaliza 
kusoma. ONDOA karatasi ya hadithi 
mbele yake. 
Muulize mwanafunzi maswali 
yanayohusiana na hadithi aliyosoma. 
Mwanafunzi lazima asome hadithi 
ambayo inahusiana na maswali 
atakayoulizwa. 
Iwapo mwanafunzi atashindwa 
kujibu swali baada ya sekunde kumi 
(10) weka alama ya ‘hakuna jibu’ na 
endelea kuuliza swali linalofuata. 
Usirudie kuuliza swali. 
After the child is finished reading, 
REMOVE the passage from in front of 
the child. 
Ask the child only the questions related 
to the text read. A child must read all 
the text that corresponds with a given 
question. If the child does not provide 
a response to a question after 10 
seconds, mark “no response” and 
continue to the next question. Do not 
repeat the question.

 Hii ni karatasi yenye hadithi 
fupi. Soma hadithi hii kwa sauti, 
haraka na kwa umakini: 
Ukimaliza kusoma nitakuuliza 
maswali kuhusu yale uliyosoma. 
Je umeelewa unachotakiwa 
kufanya. Nikisema “Anza” soma 
hadithi, haraka na kwa umakini 
kadri uwezavyo. Kama utaona 
neno usiloweza kusoma endelea 
kusoma neno linalofuata. Weka 
kidole chako kwenye neno la 
kwanza. Je, upo tayari? “Anza”. 
Here is a short story. I want you to 
read it aloud, quickly but carefully. 
When you finish, I will ask you 
some questions about what you have 
read. When I say “Begin”, read the 
story as best as you can. If you 
come to a word you do not know, go 
on to the next word. Put your finger 
on the first word. Ready? Begin. 
 Anza kupima muda mara 
mwanafunzi anapoanza kusoma 
neno la kwanza. 
Start the timer when the child reads 
the first word. 
 Endapo muda umekwisha 
(sekunde 60). If the time on the 
stopwatch runs out (60 seconds).

 Sasa nitaanza kuuliza maswali 
machache kuhusu hadithi uliyosoma. 
Jitahidi kujibu maswali vizuri kwa 
kadri uwezavyo. Now I will ask you a 
few questions about the story you just 
read. Try to answer the questions as 
well as you can. 
 () 1 = Sahihi / Correct 
 () 0 = Isiyosahihi / Incorrect 
 () . = Hakuna jibu / No response
 
USAID Tusome Pamoja - Preliminary Midline Report 23 
 ( / ) Weka alama ya mkwaju 
kwa maneno yote aliyokosea 
kuyasoma. Mark any incorrect 
words with a slash. 
 ( Ø ) Iwapo uliweka alama ya 
kosa aliposahihisha kusoma neno 
alilokosea, zungushia neno kisha 
endelea. Circle self-corrections if 
you already marked the word 
incorrect.
( ] ) Weka alama ya mabano 
katika neno la mwisho alilosoma.
Mark the final word read with a 
bracket.
One day Furaha went to the bush to pick fruits. On the way she saw 
an old woman sitting on the pathway. Near her was a basket of 
oranges. Some of them had spilled. The old woman was happy to 
see Furaha. She asked Furaha to remove a thorn that had pierced 
her leg. The old woman could not reach the thorn. Furaha took out 
the thorn slowly. The old woman thanked Furaha and gave her 
some oranges. 1. Where did Furaha go? [to the bush] 2. Who sat on 
the pathway? [an old woman] 3. What was in the basket? [oranges] 
4. What did the old woman ask Furaha? [to remove a thorn] 5. Why 
did the old woman give Furaha oranges? [because she helped; she 
removed the thorn; she removed the thorn slowly] 
Piloted January 2016; Used February 
2016 and August 2016 
Maswali [Majibu] 
Questions [Answers] 
Siku moja Furaha alikwenda porini kuchuma 
matunda. 7 Furaha alikwenda wapi?
[porini] 1 0 . 
Njiani alimuona bibi amekaa barabarani. 12 Nani aliketi barabarani?
[bibi] 1 0 . 
Karibu yake lilikuwepo kapu la machungwa. Baadhi 
ya machungwa yalikuwa yamemwagika. 23 Kwenye kapu kulikuwa na 
kitu gani? [machungwa] 1 0 . 
Bibi alifurahi kumuona Furaha. Bibi akamuomba 
Furaha amtoe mwiba mguuni. 33 
Bibi alimwomba Furaha 
afanye nini? [amtoe 
mwiba] 
1 0 . 
Bibi hakuweza kuufikia ule mwiba. Furaha alimtoa 
mwiba polepole. Bibi alimshukuru Furaha na kumpa 
machungwa machache. 
49 
Kwa nini Bibi alimpa 
Furaha machungwa? [kwa 
sababu alimsaidia; alimtoa 
mwiba mguuni; alimtoa 
polepole] 
1 0 . 
 Muda uliobaki (sekunde) 
Time remaining (seconds)
 
 Sitisha zoezi kwa sababu 
mwanafunzi hana majibu kwenye 
mstari wa kwanza.
Exercise discontinued because the 
child had no correct answers in 
the first line.
 
24 USAID Tusome Pamoja - Preliminary Midline Report
Annex B. Sample Methodology, Weighting, and 
Precision 
This annex discusses the details of the sample, the population that it is meant to 
represent, and how the sample is properly representative of that population. It also 
discusses the precision estimates for the major outcome variables from which the 
sample size was derived. 
Population of Interest and Sample Frame 
The population of interest includes all primary government schools located in the five 
USAID Tusome Pamoja regions. More specifically, the population of interest was all 
Standard 2 and Standard 4 pupils enrolled in these 2,883 schools. 
The sample frame used to draw the midline sample was created from two census 
lists of schools: one from the Mainland, the other from Zanzibar. The Mainland 
census data came from the 2016 Primary School Leaving Certificate Examination.1
. 
The census data from Zanzibar came from the 2016 census list of schools collected 
by the Zanzibar Ministry of Education and Vocational Training. Table B.1 provides 
the total number of schools in the list frame along with the total number of schools 
excluded for the given reason. Finally, it provides the total number of schools that 
make up the defined population. Table B.2 provides the population of schools by 
region as well as the Standard 2 and Standard 4 enrollment. These updated pupil 
enrollment population figures by region and standard were based on the most up-to￾date estimates from USAID Tusome Pamoja. 
Table B.1. Schools Excluded from the List Frame Prior to Sampling 
# Schools Percent 
Total number of schools in the sample frame 16,141 
Not excluded [defined population] 2,883 17.86 
Reason for exclusion 
Not a Tusome Pamoja Region 13,226 81.94 
School was marked as non-government 32 0.2 
Table B.2 Population Counts of Schools and Pupils by Tusome Pamoja Region 
Region Schools Standard 2 Pupils Standard 4 Pupils 
Iringa 453 38,622 29,652 
Morogoro 827 86,237 64,192 
Mtwara 607 51,331 38,426 
Ruvuma 721 58,057 44,466 
Zanzibar 275 42,855 38,119 
Total 2,883 277,102 214,855 
 
1
 2016 Primary School Leaving Certificate Examination : http://opendata.go.tz/dataset/primary-schools-leaving￾certificate-examination-performance-ranking-with-location/resource/d4703a57-3fa1-4464-a459-683e3dfc5e04 
 
USAID Tusome Pamoja - Preliminary Midline Report 25 
Sample Methodology 
The sample methodology was for a three-stage random sample of schools, 
classrooms, and pupils. Schools were separated by region, then sorted by district 
and number of pupils at the school. Schools were selected with probability 
proportional the number of pupils at the school. A total of 90 schools were selected 
(18 from each region), and for each selected school, 2 replacement schools were 
automatically selected. The replacement schools were the closest resemblance of 
the originally selected schools by district and enrollment. The replacement schools 
would replace the originally sampled school if the originally sampled school did not 
meet the requirements as defined by the population. Only 2 schools (both in 
Morogoro) needed to be replaced due to inaccessibility.
Once the assessment team arrived at the school, it randomly selected one Standard 
2 classroom and one Standard 4 classroom. Within each of the selected Standard 2 
and Standard 4 classrooms, 10 pupils were selected (5 boys and 5 girls). . Table B.3
displays the sample methodology. Table B.4 provides the final counts of completed 
school, classroom, and pupil assessments. 
Table B.3. Sample Methodology for the Tusome Pamoja Midline 2018 
Assessment. : Three-Stage Stratified Sample of Schools, Classrooms, and 
Pupils 
Stage Item Sampled 
(Expected Counts) 
Stratification 
(Number of Strata) 
Probability of Selection 
1 Schools 
(90) 
Region 
(5) 
Proportional to enrollment* 
2 Classroom 
(180) 
Standard 2 /Standard 4 (2) Equal 
3 Pupils 
(1,795) 
Boys/Girls 
(2) 
Equal 
*Proportional to Enrollment: For the Mainland data, enrollment was the total number of pupils who sat for the 
Primary School Leaving Certificate Examination in 2016. For the Zanzibar data, it was the Standard 2 
enrollment for the 2016 census data. 
26 USAID Tusome Pamoja - Preliminary Midline Report
Table B.4. Final Sample Counts of Assessments, Interviews, and Observations 
Collected for the Tusome Pamoja 2018 Midline Evaluation 
Regions Sampled 
Schools Standard Classrooms Pupil 
Gender Pupils Assessed 
5 Tusome 
Pamoja 
Regions 
90 
Standard 2 90 
Boys 448 
Girls 453 
Standard 4 90 
Boys 446 
Girls 448 
Total 90 180 1,795 
Sample Weights 
All sample weights were calculated as the inverse of the probability of selection at 
each stage of selection (school, classroom, and pupil). School weights were scaled 
to the known populations of schools by region. Classrooms were not scaled to the 
final population as that information was not available in the school census data. The 
Standard 2 and Standard 4 pupil weights were scaled to the population region and 
standard. We were unable to scale to the boy/girl pupil population because the 
Mainland sample frame did not provide enrollment by gender. Table B.5 provides the 
weighted totals and the sample totals by region and standard. 
Table B.5. Weighted Sample (Estimated Population) and Sample Counts by 
Each Stage Sampled (School, Classroom, Pupil) 
Tusome 
Pamoja 
Region 
Weighted Sample 
 (Estimated Population) Sample 
School 
Classroom Pupils 
Schools
Classroom Pupils 
S2 S4* S2 S4 S2 S4 S2 S4 
Iringa 453 688 688 38,622 29,652 18 18 18 180 180 
Morogoro 827 1580 1580 86,237 64,192 18 18 18 180 177 
Mtwara 607 1241 1241 51,331 38,426 18 18 18 180 177 
Ruvuma 721 882 882 58,057 44,466 18 18 18 180 180 
Zanzibar 275 655 655 42,855 38,119 18 18 18 181 180 
Total 2,883 5,046 5,046 277,102 214,855 90 90 90 901 894 
*S=standard; Standard 4 enrollment was not collected as part of the midline data collection. Therefore, Standard 
2 values were used as proxies (thus leading to identical estimates for Standard 2 and Standard 4 in the weighted 
sample). 
Precision Estimates 
The Standard 2 and Standard 4 sample size calculations were derived using the oral 
reading fluency score and having a desired 95% confidence band of ±3.0 words per 
minute (wpm) by standard. With a predicted standard deviation of 25.0, and 
intraclass correlation of 0.25, the school sample size was estimated to be 90 schools 
across all five regions if we were to sample 10 pupils in each Standard 2 and 
 
USAID Tusome Pamoja - Preliminary Midline Report 27 
Standard 4. In other words, the number of schools was based on the ability to 
confidently estimate the average oral reading fluency score to within 3 correct wpm 
above or below the estimate. Larger sample sizes would have reduced this band, but 
the additional costs were not justified by the marginally increased precision. 
Additionally, it turns out that both Standard 2 and Standard 4 actually showed greater 
precision than that used for the sample calculation. . The Standard 2 estimate is very 
precise with ±1.9 wpm, and the Standard 4 precision estimate was ±2.4. Table B.6
provides the estimates from the actual data. 
Table B.6. Reading Fluency Means and Precision Estimate by Standard 
Standard n 
Range 
(wpm) 
Mean 
(wpm) 
95% CI Band 
(wpm) 
SD 
(wpm) ICC DEF 
Std2 901 [0,76] 16.4 ±1.9 14.2 0.245 2.2 
Std4 893 [0,101] 36.7 ±2.4 19.6 0.219 1.7 
Overall 1794 [0,101] 25.3 ±1.8 19.5 0.123 1.9 
wpm: words per minute, CI: confidence interval, SD: Standard deviation, ICC: Intra class correlation, DEF: design 
effect 
It should be noted that the midline sample size was never meant to be disaggregated 
beyond the Standard 2 and Standard 4 level. The midline data were never meant to 
be disaggregated by region or any other school, classroom, or pupil characteristic. 
For this reason, the midline sample was significantly smaller than the baseline 
sample—which was based on the need for disaggregation at the regional level. 
28 USAID Tusome Pamoja - Preliminary Midline Report
Annex C. Inter-rater Reliability and Assessor 
Accuracy 
The following table provides an overview of the percent agreement across all 
subtasks during the second assessor accuracy measure exercise. 
Assessor Accuracy Measure 
Percent Agreement by Assessor 
Assessor Oral 
Reading
Reading 
Comprehension
Missing 
Number
Basic 
Addition
Addition 
Level 2 
Basic 
Subtraction
Subtraction 
Level 2 
Overall
1 100% 100% 100% 100% 100% 100% 100% 100% 
2 100% 100% 100% 100% 100% 100% 100% 100% 
3 100% 100% 100% 100% 100% 100% 100% 100% 
4 100% 100% 100% 100% 100% 100% 100% 100% 
5 100% 100% 100% 100% 100% 100% 100% 100% 
6 98% 100% 100% 100% 100% 100% 100% 100% 
7 98% 100% 100% 100% 100% 100% 100% 100% 
8 98% 100% 100% 100% 100% 100% 100% 100% 
9 98% 100% 100% 100% 100% 100% 100% 100% 
10 96% 100% 100% 100% 100% 100% 100% 99% 
11 96% 100% 100% 100% 100% 100% 100% 99% 
12 96% 100% 100% 100% 100% 100% 100% 99% 
13 96% 100% 100% 100% 100% 100% 100% 99% 
14 96% 100% 100% 100% 100% 100% 100% 99% 
15 96% 100% 100% 100% 100% 100% 100% 99% 
16 96% 100% 100% 100% 100% 100% 100% 99% 
17 96% 100% 100% 100% 100% 100% 100% 99% 
18 94% 100% 100% 100% 100% 100% 100% 99% 
19 94% 100% 100% 100% 100% 100% 100% 99% 
20 94% 100% 100% 100% 100% 100% 100% 99% 
21 94% 100% 100% 100% 100% 100% 100% 99% 
22 94% 100% 100% 100% 100% 100% 100% 99% 
23 92% 100% 100% 100% 100% 100% 100% 99% 
24 92% 100% 100% 100% 100% 100% 100% 99% 
25 96% 100% 100% 95% 100% 100% 100% 99% 
26 90% 100% 100% 100% 100% 100% 100% 99% 
27 90% 100% 100% 100% 100% 100% 100% 99% 
28 88% 100% 100% 100% 100% 100% 100% 98% 
29 96% 100% 90% 100% 100% 100% 100% 98% 
30 86% 100% 100% 100% 100% 100% 100% 98% 
31 94% 100% 90% 100% 100% 100% 100% 98% 
32 94% 100% 90% 100% 100% 100% 100% 98% 
 
USAID Tusome Pamoja - Preliminary Midline Report 29 
Assessor Accuracy Measure 
Percent Agreement by Assessor 
Assessor Oral 
Reading
Reading 
Comprehension
Missing 
Number
Basic 
Addition
Addition 
Level 2 
Basic 
Subtraction
Subtraction 
Level 2 
Overall
33 88% 100% 100% 95% 100% 100% 100% 98% 
34 92% 100% 90% 100% 100% 100% 100% 97% 
35 90% 100% 90% 100% 100% 100% 100% 97% 
36 80% 100% 100% 100% 100% 100% 100% 97% 
37 98% 100% 100% 100% 80% 100% 100% 97% 
38 96% 100% 100% 100% 80% 100% 100% 97% 
39 90% 100% 80% 100% 100% 100% 100% 96% 
40 96% 100% 90% 100% 80% 100% 100% 95% 
41 82% 100% 90% 95% 100% 95% 100% 95% 
42 98% 100% 80% 100% 100% 100% 80% 94% 
43 88% 80% 90% 100% 100% 100% 100% 94% 
44 96% 80% 80% 100% 80% 95% 100% 90% 
45 90% 60% 70% 100% 80% 100% 80% 83% 
The following table provides an overview of the percent agreement for oral reading 
and reading comprehension, obtained during the field inter-rater reliability activities 
(for each day of data collection). Each row in the table represents a pair of assessors 
who simultaneously scored a single student. These data were analyzed at the end of 
each day of data collection, and results were provided to the USAID Tusome Pamoja 
monitoring and evaluation team—who followed up with assessors in all cases where 
agreement was low. Overall, there was strong agreement throughout data collection, 
with an average of 97% agreement on oral reading and 98% agreement for reading 
comprehension. 
Field Interrater Reliability 
Percent Agreement by Assessor
ID Date Percent Agreement 
for Oral Reading 
Percent Agreement 
for Reading 
Comprehension
RVGNSB 8/23/18 100% 100% 
GGEQSG 8/23/18 86% 100% 
TTRVQB 8/23/18 98% 100% 
YQENNZ 8/23/18 100% 100% 
BUQEGV 8/23/18 100% 100% 
AZZXHW 8/23/18 94% 100% 
VFFAFR 8/23/18 98% 100% 
KEXNWE 8/23/18 100% 100% 
YNKEGC 8/23/18 100% 100% 
PVSQTQ 8/23/18 98% 100% 
BCBSUR 8/23/18 100% 100% 
CWQMQK 8/23/18 96% 100% 
EZCTXA 8/24/18 94% 100% 
30 USAID Tusome Pamoja - Preliminary Midline Report
Field Interrater Reliability 
Percent Agreement by Assessor
ID Date Percent Agreement 
for Oral Reading 
Percent Agreement 
for Reading 
Comprehension
EFFGXG 8/24/18 100% 0% 
SPZRTH 8/24/18 100% 100% 
ZBVBEG 8/24/18 96% 100% 
KRNSUU 8/24/18 100% 100% 
PUBSRM 8/24/18 94% 100% 
KBWWRW 8/24/18 100% 100% 
SZVBUE 8/24/18 94% 100% 
HXQUQH 8/24/18 98% 100% 
VKCBKV 8/24/18 98% 80% 
NNBAYH 8/24/18 100% 100% 
MHFWTP 8/24/18 100% 100% 
BAPQBW 8/27/18 100% 80% 
NQHNTH 8/27/18 96% 100% 
XCWQUE 8/27/18 98% 100% 
TXMSXG 8/27/18 84% 100% 
NRVFBC 8/27/18 98% 100% 
ENZSFB 8/27/18 98% 80% 
QBSTQX 8/27/18 98% 100% 
TEQFVR 8/27/18 100% 100% 
VBHHWZ 8/27/18 98% 100% 
RFZSRZ 8/27/18 100% 100% 
UEZGVH 8/27/18 86% 100% 
TKQWZG 8/27/18 86% 100% 
RWHYKY 8/27/18 96% 100% 
CGAMQR 8/27/18 100% 100% 
VHNNWU 8/27/18 98% 100% 
GTRFPU 8/27/18 86% 100% 
SUUGAG 8/28/18 100% 100% 
FVETTN 8/28/18 98% 100% 
TTMBMM 8/28/18 100% 100% 
NFVHPZ 8/28/18 100% 100% 
VYCXCU 8/28/18 100% 100% 
SUQRZK 8/28/18 92% 100% 
WSYUYF 8/28/18 98% 100% 
KQMEYS 8/28/18 98% 100% 
CASCEC 8/28/18 98% 100% 
EBFUGE 8/28/18 96% 100% 
EBFQVX 8/28/18 98% 100% 
KZFZUZ 8/28/18 96% 100% 
HUFSGR 8/28/18 100% 100% 
 
USAID Tusome Pamoja - Preliminary Midline Report 31 
Field Interrater Reliability 
Percent Agreement by Assessor
ID Date Percent Agreement 
for Oral Reading 
Percent Agreement 
for Reading 
Comprehension
AZAKQM 8/28/18 98% 100% 
CNMGZX 8/28/18 100% 100% 
MWWHKQ 8/28/18 100% 100% 
YFSQUN 8/29/18 100% 100% 
HSMAWZ 8/29/18 100% 100% 
ZHBNPT 8/29/18 100% 100% 
RBMEYY 8/29/18 96% 100% 
SWVQZF 8/29/18 100% 100% 
QARAWY 8/29/18 100% 100% 
HACHYT 8/29/18 94% 100% 
TRPYRR 8/29/18 98% 100% 
FBKFXF 8/29/18 100% 100% 
HWCKEY 8/29/18 100% 100% 
BWRPYQ 8/29/18 100% 100% 
NYQKYH 8/30/18 100% 100% 
FSKCUN 8/30/18 90% 100% 
ZEQSVR 8/30/18 100% 100% 
CRQGSQ 8/30/18 94% 100% 
EUVAXQ 8/30/18 100% 100% 
ECMXHM 8/30/18 100% 100% 
XZYVFE 8/30/18 100% 100% 
PRSTEY 8/30/18 94% 100% 
GMCXCE 8/30/18 100% 100% 
FCPKEH 8/30/18 98% 100% 
RFFMMB 8/30/18 100% 100% 
ZUBGMF 8/31/18 97% 98% 
SRXHCN 8/31/18 100% 100% 
NRRRWG 8/31/18 82% 100% 
CTPZNX 8/31/18 100% 100% 
YCGMPT 8/31/18 98% 100% 
ZTRPPR 8/31/18 100% 100% 
TMPFRK 8/31/18 98% 100% 
PBKYZR 8/31/18 96% 100% 
OVERALL 97% 98%