abessenba/somalia-education-analytics-code

Bar ama Baro - “Teach or Learn” Somalia’s Accelerated Quality Learning Program # Reproducible Evaluation Toolkit: Bar ama Baro (USAID/Somalia) This repository contains a curated sample of the R-based reproducible data pipeline developed for the Bar ama Baro (BAB) USAID/Somalia Accelerated Basic Education initiative. By transitioning from manual, spreadsheet-based analysis to a modular, metadata-driven RStudio workflow, our evaluation team achieved an 80–90% reduction in reporting time while establishing a fully transparent, FAIR-aligned research pipeline across 20+ complex, multilingual datasets. 🛠️ Repository Structure This repository illustrates the end-to-end data pipeline—from raw data intake and cleaning to automated analysis and report generation: - student_data_build.Rmd # Data build pipeline for student-level datasets - teacher_data_build.Rmd # Data build pipeline for teacher-level datasets - somalia.Rmd # Master analysis & report generation script - README.md # Project documentation and pipeline overview # Script Overview & Execution Order 1. student_data_build.Rmd - Processes, cleans, and standardizes raw student assessment and tracking data. - Handles multi-site data validation, anonymization, and variable labeling for longitudinal tracking across target districts, outputting structured datasets for analysis. 2. teacher_data_build.Rmd - Cleans, transforms, and structures raw educator datasets. - Standardizes teacher-level metrics, training records, and school-level administrative links, outputting structured datasets for analysis. 3. somalia.Rmd - The primary analysis and report generation engine. - Ingests the clean dataset outputs generated by student_data_build.Rmd and teacher_data_build.Rmd. - Executes statistical routines, applies standardized visualization themes (e.g., custom Likert charts, lollipop plots), and compiles the final evaluation reporting artifacts. 📊 Project Scope & Impact - Program: Bar ama Baro (USAID/Somalia Accelerated Basic Education) - Scale Evaluated: 39,930 learners …