sammyamanya/impact_evaluation_education_program_uganda

To estimate the impact of an education support program on pupil learning outcomes among Primary 5 and Primary 6 students in a government primary school in rural Uganda. The project demonstrates the application of causal inference methods commonly used in education, development, and social policy evaluations. ## Impact Evaluation of an Education Support Program in Rural Uganda ## Project Overview This project evaluates the impact of a hypothetical education support program on learning outcomes among Primary 5 and Primary 6 pupils in a government primary school in rural Uganda. Using synthetic data, the project demonstrates the application of causal inference methods commonly used in Monitoring and Evaluation (M&E), education research, social policy evaluation, and development economics. The analysis follows a structured impact evaluation workflow, beginning with exploratory analysis and progressing through increasingly rigorous estimation approaches. --- ## Evaluation Question **What is the effect of an education support program on pupil learning outcomes?** Specifically, the project estimates the impact of program participation on pupil test scores using quasi-experimental methods. --- ## Intervention The hypothetical education support program includes: - Remedial reading sessions - Mathematics tutoring - Teacher coaching - Learning materials - Attendance monitoring The program targets Primary 5 and Primary 6 pupils and aims to improve learning outcomes through enhanced instructional support and pupil engagement. --- ## Theory of Change ```text Education Support Program ↓ Improved Learning Environment ↓ Higher Attendance and Study Effort ↓ Improved Academic Performance ↓ Higher Test Scores ``` --- ## Data The dataset was synthetically generated for educational and portfolio purposes. ### Dataset Characteristics - 240 pupils - Primary 5 and Primary 6 learners - Treatment and comparison groups - Baseline and endline observations - Total observations: 480 ### Variables - Student ID - Class Level - Gender - Treatment Status - Attendance Rate - Weekly Study Hours - Test Score - Survey Round (Baseline/Endline) --- ## Methods The project applies four analytical approaches: ### 1. Exploratory Data Analysis (EDA) - Descriptive statistics - Distributio …