Musomesa Field School (MFS) Impact Evaluation — Project Overview

Project at a Glance

Client: Japan International Cooperation Agency Research Institute (JICA-RI)
Principal Investigator: Dr. Yoko Kijima (National Graduate Institute for Policy Studies – GRIPS)
Duration: May 2024 – February 2027 (Baseline, Follow-up 1, Follow-up 2)

Baseline Survey (May-July 2024):
Follow-up Survey 1 (January-March 2025)
Follow-up Survey 2 (December 2025-February 2027)

Geographic Scope: West Nile Region, Uganda (5 districts: Nebbi, Maracha, Koboko, Yumbe, Moyo)
Sample Size: 660 households across 55 villages
Study Design: Randomized Controlled Trial (RCT)
Methodology: Household surveys, Village-level Focus Group Discussions (FGDs), GPS plot measurements, intervention monitoring
Role: Sole consultant for all survey phases and RCT implementation support

Background & Research Context

The Musomesa Field School (MFS) approach was developed under the Promotion of Sustainable Rice Farming Development Project (Eco-PRiDe/PRiDe II, 2019-2024), implemented through collaborative work between JICA, Uganda’s National Agricultural Research Organization (NARO), and the National Crops Resources Research Institute (NaCRRI).

The MFS Model:

The Musomesa Field School represents an adaptation of Farmer Field School methodologies specifically designed for rice production systems in Uganda. The approach involves:

Training selected “lead farmers” in improved rice cultivation practices
Establishing demonstration plots managed by these trained farmers
Facilitating farmer-to-farmer knowledge transfer within communities
Promoting sustainable rice farming technologies developed through the Eco-PRiDe project

Policy Context:

Uganda faces significant constraints in agricultural extension services:

Low ratio of extension workers to farming households
Limited reach of government extension programs to remote areas
High costs of traditional extension delivery models

The MFS approach offers a potential solution by leveraging social networks and peer learning. However, empirical evidence on the effectiveness of farmer-to-farmer extension diffusion remained limited.

The Research Challenge

JICA-RI and Dr. Kijima designed a rigorous impact evaluation to examine barriers to farmer-to-farmer extension in the MFS approach. The research questions centered on:

Primary Research Objectives:

Does providing additional training to Musomesa graduates enhance their understanding of rice technologies?
Does improved training enhance their ability to teach neighbors effectively?
Does increasing visibility and accessibility of lead farmers (through field days) promote technology diffusion?
What are the barriers preventing effective farmer-to-farmer knowledge transfer?

Intervention Design:

The RCT involved randomizing 55 villages into treatment and control groups:

Treatment villages: MFS graduates receive additional training sessions and organize community field days
Control villages: Standard MFS implementation without additional interventions

Data Collection Requirements:

HDS was tasked with:

Conducting baseline survey before intervention deployment
Tracking and supervising RCT implementation
Conducting follow-up surveys to measure impact
Collecting both household-level and village-level data
Measuring rice plot characteristics with GPS precision
Monitoring attendance and participation in training sessions

HDS’s Role Across Study Phases

Phase 1: Baseline Survey (May-July 2024)

Preparation Activities:

HDS collaborated with Dr. Kijima to:

Review and refine survey instruments for local context
Track survey sites and identify target villages
Select farmer groups and individual households for the RCT sample
Design sampling framework ensuring randomization integrity

Sample Design:

660 households across 55 villages in 5 West Nile districts
Stratified sampling to ensure representation across districts
Inclusion of both MFS graduate farmers and their neighbors (potential adopters)
Village-level sampling for focus group discussions

CAPI Development:

The Data Development Consultant (Abdul Naafi Nsereko) digitalized survey instruments using SurveyCTO, programming:

Household demographic modules
Land access and tenure arrangements
Rice cultivation practices (varieties, inputs, techniques)
Access to finance and credit
Knowledge of improved practices
Social networks and information sources
Village-level characteristics and infrastructure

The CAPI system incorporated GPS data collection protocols for:

Rice plot boundaries and area measurement
Musomesa demonstration plot locations
Lead farmer homestead coordinates
Key community infrastructure points

Team Mobilization:

HDS assembled field teams specifically for West Nile deployment:

2 field supervisors with agricultural survey experience
12 enumerators (6 per team, plus 2 reserves)
Preference for enumerators speaking West Nile languages (Lugbara, Madi, Alur, Kakwa)
Project Manager for overall coordination

Training Program:

Supervisor Pre-Training (2 days):

Early training for supervisors to refine questionnaires
Identification of potential issues before general training
Preparation of supervisors to support enumerator training effectively

General Enumerator Training (9 days):

Days 1-4: Questionnaire Comprehension

Module-by-module review of hard copy questionnaires
Individual enumerators presenting their understanding of modules
Role-play exercises with enumerators as respondents and interviewers
Discussion of probing techniques for accurate responses

Days 5-7: CAPI and Technical Skills

Hands-on training with SurveyCTO on tablets
GPS unit operation and calibration procedures
GPS plot measurement techniques (walking plot boundaries, recording coordinates)
Practice capturing altitude, latitude, and longitude data
Data backup procedures (notebook recording, CAPI entry, GPS unit storage)

Days 8-9: Field Protocols and Ethics

Research ethics and data confidentiality
Appropriate field behavior and cultural sensitivity
Communication with village leaders and respondents
Obtaining informed consent
Problem-solving difficult interview situations

Dr. Kijima participated in training sessions (Day 1 introduction to study purpose and motivation, virtual attendance for later sessions as needed). All proposed questionnaire changes during training were documented and referred to Dr. Kijima for approval before incorporation into the CAPI.

GPS Measurement Training:

Special emphasis was placed on GPS plot measurement protocols:

Walking around rice plot perimeters as directed by farmers
Taking coordinates at main corners of irregularly shaped plots
Recording center-point coordinates
Documenting plot identifiers in notebooks for backup
Entering GPS data into designated CAPI fields
Saving data with unique household IDs in GPS units
Maintaining uniform GPS format calibration across all units

Enumerators were trained to never alter GPS unit format settings to ensure data consistency.

Pre-test/Pilot Study (3 days):

HDS conducted pre-testing in West Nile districts outside actual survey villages:

2 days travel to/from West Nile region
1 day field testing of instruments

The pilot served to:

Validate questionnaire clarity and flow
Determine time required per interview
Test CAPI functionality in field conditions
Assess GPS measurement procedures
Allow supervisors to observe enumerator performance for final selection

Supervisors and Project Manager randomly observed enumerator interviews during pre-test to assess:

Understanding of questions and proper administration
Interview technique and respondent rapport
Technical proficiency with tablets and GPS units
Suitability for final team selection

Pre-test data was reviewed for:

Missing fields requiring CAPI correction
Logic errors or skip pattern problems
Question clarity issues

All identified issues were resolved through CAPI revisions and additional training before actual survey launch.

Baseline Data Collection:

Field deployment was organized with 2 teams of 6 enumerators each, supervised by 1 supervisor per team:

Team Assignment:

Each team assigned 2.5 districts
Teams worked simultaneously across West Nile region
Each enumerator expected to complete 2 household interviews per day
Coverage of approximately 1 village per day per enumerator

Data Collection Timeline:

28 working days of interviews
4 Sundays (no interviews, used for error resolution, planning, rest)
2 days for inter-district travel and callbacks
2 days travel from Kampala to West Nile and return
Total: 36 days

Supervisor Responsibilities:

Daily tasks included:

Planning enumerator assignments and interview schedules
Making bookings and coordination with village leaders
Reviewing completed questionnaires for completeness and quality
Conducting village-level focus group discussions
Uploading approved data from enumerator tablets daily
Uploading GPS data from enumerator units daily to supervisor laptop
Backing up all data files
Communicating with Project Manager on progress and issues

Village-Level Data Collection:

Supervisors conducted Focus Group Discussions covering:

Village size and demographics
Access to markets and distances
Infrastructure availability (roads, storage facilities)
Population movements and migration patterns
Community organizations and leadership structures
Agricultural practices and common challenges

Quality Assurance During Baseline

Standardized Checklists:

HDS developed quality assurance checklists for supervisors covering:

Questionnaire completeness verification
Internal consistency checks
GPS data quality review
Photographic documentation verification

Field Supervision:

Project Manager made random field visits to observe teams
Direct observation of enumerators conducting interviews
Verification that training protocols maintained in field
Assessment of team adherence to ethical guidelines

Daily Data Review:

Supervisors checked all questionnaires before electronic submission
Project Manager and Data Manager conducted additional data quality checks
Error queries generated and communicated to field teams
Daily data submission enabled quick review and feedback

Random Back-Checks:

Sample of completed households re-contacted for verification
Cross-checking of key responses for accuracy
Documentation of any discrepancies for investigation

Data Cleaning and Submission:

The Data Manager ran tests to check structural integrity of submitted data. Upon completion:

Both RAW and cleaned datasets prepared
Formats: Stata, MS Excel, CSV (as preferred by Dr. Kijima)
GPS data files compiled and organized
All data submitted via secure connection by agreed deadline

Baseline Survey Report:

HDS prepared comprehensive technical reports including:

Methodology and sampling approach
Field observations and challenges encountered
Data quality assessment
Attrition documentation (households not located/interviewed)
Recommendations for follow-up waves

Phase 2: Follow-up Survey 1 (January-March 2025)

Objective:

The first follow-up survey revisited the same 660 baseline households to:

Measure early impacts of MFS interventions
Track changes in rice cultivation practices
Assess knowledge diffusion from lead farmers to neighbors
Document participation in training sessions and field days
Update household circumstances and rice production data

Preparation:

HDS worked with Dr. Kijima to:

Update survey instruments to capture intervention exposure
Add modules on training attendance and knowledge acquisition
Maintain comparability with baseline for impact assessment
Refine questions based on baseline experience

Tracking and Re-Contact:

The intention was to locate and revisit all 660 baseline households. HDS’s approach:

Use baseline GPS coordinates to locate households
Follow households that had shifted within same district
Employ tracking methods proven in previous longitudinal studies
Document reasons when households could not be located

Team Deployment:

Same team structure as baseline (2 teams, 2 supervisors, 12 enumerators) to maintain consistency and leverage familiarity with:

Survey areas and village structures
Baseline respondents
Local languages and cultural contexts
Survey protocols and instruments

Quality Assurance:

Identical quality control procedures as baseline:

Daily data submission and validation
Supervisor quality checklists
Random back-checks
Project Manager field visits
Real-time error resolution

Deliverables:

Cleaned dataset submitted by March 20, 2026
Technical and financial reports by April 15, 2026

Phase 3: Follow-up Survey 2 (December 2025-February 2027)

Two-Phase Implementation:

The second follow-up is structured in two phases:

Phase 1: December 1, 2025 – October 2026
Phase 2: November 20, 2026 – February 27, 2027

Activities of Phase 1 replicated in Phase 2, ensuring consistent measurement across time.

Survey Objectives:

Measure longer-term impacts including:

Sustained changes in rice production practices
Technology adoption patterns among non-trained farmers
Diffusion pathways through social networks
Economic outcomes (yields, incomes)
Barriers to adoption persistence

Preparation for Follow-up 2:

Contract Negotiation: December 2025
Inception Meeting: December 2025

Review of objectives and any instrument modifications
Planning for field deployment
Budget and timeline confirmation

Review and Refinement of Tools: December 2025

Dr. Kijima provides updated questionnaires
HDS reviews for consistency and local appropriateness
Comments and suggestions shared with Dr. Kijima
Final versions agreed upon

CAPI Construction: January 2026

Data Manager digitalizes updated questionnaires in SurveyCTO
Programming of any new modules or modifications
Testing of CAPI functionality
Submission of draft CAPI to Dr. Kijima for review and approval

Recruitment: January 2026

Priority given to enumerators from baseline and follow-up 1 (familiarity with survey and areas)
New recruitment if needed for local language coverage
Minimum qualification: Bachelor’s degree
Verification of integrity and reliability
7 enumerators recruited per team (6 plus 1 reserve)

Training: January 2026

2-day supervisor pre-training for questionnaire refinement
9-day general enumerator training (same structure as baseline)
Questionnaire comprehension, CAPI training, GPS protocols
Field protocols and research ethics
Dr. Kijima participation (in-person or virtual)

Pre-test: January 2026

3 days (travel, field testing, return)
Testing in West Nile districts outside survey villages
Validation of instruments and CAPI
Observation of enumerator performance
Interim report submitted
Final CAPI version submitted after incorporating pre-test lessons

Data Collection: February-March 2026 (Phase 1)

Same 36-day deployment structure
2 teams, 2 supervisors, 12 enumerators
Target: All 660 baseline households
Household surveys plus village FGDs
Daily data submission and quality checks

Data Cleaning and Submission: March 2026 (Phase 1)

Data cleaning and validation
RAW and cleaned datasets prepared
Submission by March 20, 2026 via secure connection

Technical and Financial Reports: April 2026 (Phase 1)

Comprehensive field report documenting methodology, observations, challenges
Financial accounting of project expenditures
Submission by April 15, 2026

Phase 2 Implementation: November 2026-February 2027

Replication of Phase 1 activities
Data submission by February 19, 2027
Final reports by February 26, 2027

Additional Component: Monitoring and Supervision of RCT Participants

Training Session Mobilization (May-October 2026):

After the first follow-up survey, HDS deployed officers to support RCT implementation:

Responsibilities:

Mobilize MFS graduate participants to attend up to 6 training sessions
Call participants to remind them of training schedules
Take attendance at training sessions
Monitor training implementation and participant engagement
Supervise adherence to RCT protocols (ensuring treatment/control group separation)
Document any implementation challenges or protocol deviations

Session Reporting:

Prepare reports after each training session
Document attendance, participation quality, topics covered
Note any issues affecting intervention fidelity
Share reports with Dr. Kijima for RCT monitoring

This monitoring component ensures:

Treatment is delivered as designed
Participation rates are tracked
Implementation challenges are documented
Data quality for impact evaluation is maintained

Survey Instruments & Data Collection Modules

Household-Level Survey Modules

Demographics:

Household composition and member characteristics
Age, gender, education levels
Migration and residence history

Land Access and Tenure:

Total land holdings by parcel
Tenure arrangements (owned, rented, borrowed)
Land use patterns
Tenure security perceptions

Rice Cultivation Practices:

Rice varieties planted
Plot-level production data
Input use (seeds, fertilizers, pesticides)
Labor allocation
Harvesting and post-harvest practices
Storage methods

Knowledge and Technology Adoption:

Awareness of improved rice practices
Sources of agricultural information
Training participation
Technology adoption decisions
Barriers to adoption

Access to Finance and Credit:

Credit sources and amounts
Savings patterns
Financial service access
Agricultural investment capacity

Social Networks:

Information sources
Relationships with lead farmers
Community participation
Extension contact

Economic Outcomes:

Rice yields and production
Sales and income
Household expenditures
Food security indicators

Village-Level Survey (FGD Topics)

Community Characteristics:

Village size and population
Demographic composition
Settlement patterns

Infrastructure:

Road access and quality
Market distances
Storage facility availability
Communication networks

Agricultural Context:

Common crops and farming systems
Land availability and pressure
Water resources and irrigation
Major agricultural challenges

Market Access:

Distances to input and output markets
Trader presence and frequency
Price information access
Market infrastructure quality

Social Organization:

Community leadership structures
Farmer groups and associations
Extension service presence
NGO/project activities

Population Dynamics:

In-migration and out-migration patterns
Seasonal labor movements
Land pressure trends

Data Quality Assurance Framework

Training Quality Controls

Enumerator Selection:

Competitive selection process with reserves
Performance evaluation during training
Pre-test observation for final selection
Language capability verification

Training Documentation:

Detailed training manuals prepared
Questionnaire clarifications documented
CAPI user guides developed
GPS protocols written

Field Quality Controls

Daily Oversight:

Supervisor review of all completed questionnaires
Quality checklist application before submission
Immediate feedback to enumerators on errors
Same-day resolution of data quality issues

Data Validation:

Automated consistency checks in CAPI
Real-time error flagging during data entry
Overnight validation by Data Manager
Morning error reports to field teams

Field Supervision:

Project Manager random field visits
Observation of interview protocols
Verification of GPS measurement procedures
Assessment of ethical compliance

Back-Checking:

Random re-contact of sample households
Verification of key responses
GPS coordinate validation
Investigation of discrepancies

Data Management Controls

Security Protocols:

Encrypted data transmission
Multiple backup systems (server, laptop, cloud)
No local data retention after upload
Access controls on data files

Processing Protocols:

Structural integrity testing
Outlier investigation and resolution
Consistent variable naming and labeling
Comprehensive documentation of cleaning decisions

Quality Outputs:

RAW datasets preserving original entries
Cleaned datasets with documented modifications
Data dictionaries explaining all variables
Codebooks detailing transformations
GPS coordinate files properly organized
Photographic documentation archived

Technical Innovations & Best Practices

GPS Plot Measurement

HDS developed robust protocols for accurate rice plot area measurement:

Equipment:

High-precision GPS units calibrated to uniform format
Backup recording in field notebooks
Multiple data storage points (notebook, CAPI, GPS unit)

Methodology:

Walking complete plot boundaries
Coordinates taken at all major corners
Center-point coordinates for verification
Unique household and plot identifiers
Multiple verification layers

Quality Assurance:

Cross-checking GPS data against farmer-reported areas
Visual inspection of plot shape reasonability
Coordinate verification against known district boundaries
Photographic documentation of plots

Longitudinal Tracking

HDS applied proven tracking methods from previous studies:

Pre-Field Preparation:

Review of baseline GPS coordinates
Mapping of household locations
Identification of village contact persons

Field Tracking Methods:

Use of baseline coordinates to locate households
Consultation with village leaders
Contact with relatives and neighbors
Follow-up of households that migrated within district
Documentation of attrition reasons

Success Factors:

Building rapport during baseline survey
Recording multiple contact points
Maintaining communication with village leaders
Flexibility in scheduling return visits

RCT Implementation Support

HDS’s role in maintaining RCT integrity:

Baseline:

Careful implementation of sampling frame
Verification of village assignments to treatment/control
Prevention of contamination between groups

Monitoring Phase:

Documentation of training attendance
Verification of treatment delivery
Monitoring for protocol deviations
Reporting implementation fidelity

Follow-up Surveys:

Blind interviewing (enumerators not informed of village assignments)
Consistent measurement across treatment and control
Documentation of any exposure to interventions

Project Significance

Research Contributions

This RCT provides rigorous evidence on:

Effectiveness of farmer-to-farmer extension models
Barriers to agricultural technology diffusion
Optimal design of farmer field school approaches
Cost-effective alternatives to traditional extension

Policy Relevance

Findings inform:

Uganda’s agricultural extension strategy
Design of farmer training programs
Resource allocation for extension services
Scaling decisions for MFS and similar models

Regional Impact

Evidence relevant for:

Sub-Saharan Africa’s extension challenges
Rice production intensification strategies
Sustainable agriculture promotion
Community-based extension models

The study contributes to JICA’s broader research agenda: “An Empirical Analysis on Expanding Rice Production in Sub-Sahara Africa.”

HDS’s Competitive Advantages for This Project

Prior Experience with Same Study

HDS conducted the 2017 follow-up survey for the earlier phase of this research (rice production in Uganda for JICA). This provided:

Familiarity with West Nile survey areas
Understanding of rice farming systems
Established relationships with district authorities
Knowledge of local languages and cultural contexts
Experience with Dr. Kijima’s research standards

Agricultural Survey Expertise

Relevant experience from related projects:

IFPRI rice value chain studies (Adjumani, multiple years)
University of Sydney Doho Rice Scheme research
JICA rice production surveys
Agricultural technology adoption studies

RCT Implementation Capability

Proven RCT expertise from:

Tsukuba land formalization RCT (pilot, baseline, intervention, endline)
IFPRI IPM technology adoption RCT
AfDB food loss and waste RCT
Multiple intervention studies requiring protocol fidelity

Longitudinal Study Track Record

Demonstrated household tracking capability:

4% attrition over 3 years (IFPRI COVID-19 follow-up)
96% re-contact rate (USAID agricultural inputs study)
Multiple survey waves on various projects
Creative tracking methods for mobile populations

West Nile Operational Experience

Extensive work in the region:

Multiple IFPRI studies in refugee settlements
Tsukuba land tenure research in refugee camps
Agricultural surveys across all 5 target districts
Teams fluent in Lugbara, Madi, Alur, Kakwa

JICA Partnership History

Established working relationship:

2017 rice production follow-up survey
Understanding of JICA reporting requirements
Familiarity with Japanese researcher collaboration protocols
Track record of successful project delivery

Project Team

Core HDS Staff:

George Sentumbwe — Project Manager

Overall coordination and management
JICA-RI and Dr. Kijima liaison
Fieldwork supervision and quality control
Team recruitment and training
Technical report preparation
Contract signatory

Abdul Naafi Nsereko — Data Development Consultant/Programmer

CAPI digitalization in SurveyCTO
GPS and data security protocols
CAPI training for enumerators
Daily data validation and error reporting
Environmental protection protocols for equipment

Jennifer Alinaitwe — Field Supervisor

Questionnaire review and training support
Enumeration team bookings and logistics
Data collection supervision
Focus group discussions
Routine data quality checks
Field notes and reporting

Denis Tishekwa — Field Supervisor

Questionnaire review and training support
Enumeration team bookings and logistics
Data collection supervision
Focus group discussions
Routine data quality checks
Field notes and reporting

Timothy Katungi — Finance and Administration

Financial management and reporting
Budget tracking and accountability
Administrative support

Data Collectors (12 main survey, varying for monitoring):

Participation in training
High-quality data collection per assignments
GPS plot measurement
Daily data upload
Response to data queries

Timeline Summary

2024 Baseline Survey:

May: Preparation, training, pre-test
June-July: Field data collection
Late July: Data cleaning and submission

2025 Follow-up Survey 1:

January: Preparation, training, pre-test
February-March: Field data collection
March 20: Data submission
April 15: Technical report submission

2025-2027 Follow-up Survey 2:

Phase 1:

December 2025: Contract negotiation, inception, preparation
January 2026: Training, pre-test
February-March 2026: Data collection
March 20, 2026: Data submission
April 15, 2026: Technical report

May-October 2026: Monitoring and supervision of RCT training sessions

Phase 2:

November 2026: Preparation
December 2026-February 2027: Data collection
February 19, 2027: Data submission
February 26, 2027: Final technical report

Contact

Homeland Data Services Ltd
Plot 1688, Kiwatule Along Kaduyu Road
Nakawa-Kampala, Uganda

Project Manager: George Sentumbwe
📧 sentumbwe@hds-survey.com
📞 +256 779 838 104

Principal Investigator: Dr. Yoko Kijima
National Graduate Institute for Policy Studies (GRIPS), Japan
📧 kijima@grips.ac.jp

JICA-RI Contact: Akiko Uenohara (Mrs)
📧 Uenohara.Akiko@jica.go.jp

Musomesa Field School Impact Evaluation

Abstract