Managing and Analysing Early Grade Reading Assessment Data in Stata

The African Foundational Learning (AFLEARN) Data Hub’s second one-week residential course focused on managing and analysing Early Grade Reading Assessment (EGRA) data in Stata. The five-day course brought together researchers, data managers, and analysts from African education ministries, NGOs, and research institutions. The course combined lectures, hands-on data analysis, small group sessions, and individual and group exercises.

Course Objectives

The course aimed to:

• Build capacity in EGRA-specific data cleaning, preparation, and analysis
• Promote high standards of transparency, replicability, and good coding practice
• Support effective data use for policy-relevant research and advocacy
• Foster a network of African practitioners working with foundational learning data

Key focus areas

Key focus areas included:

Practical Stata Skills: Participants engaged in hands-on sessions covering data importing from SurveyCTO/Tangerine; variable generation, labelling, cleaning, and transformation; merging and reshaping datasets; looping, macros, and working with strings, dates, and missing values.

EGRA Data Challenges: The course highlighted common data quality issues in field-based EGRA data collection, including enumerator and tool-related errors; missingness, outliers, and internal inconsistencies; and unique identifiers and data confidentiality.

Statistical Analysis and Interpretation: Participants learned to construct outcome measures (fluency, accuracy, benchmarks); understand distributions and standard errors; used survey weights and design features (e.g. svy: commands); and avoided common pitfalls in data interpretation.

Replicability and Do-File Hygiene: Strong emphasis was placed on writing clean, modular, and transparent Stata code for reproducible research. Participants developed well-documented do-files; modularized workflows using globals, locals, and master do-files; and applied best practices in error handling and code structure.

Application of skills: Participants had opportunities to apply skills learnt to their own EGRA data, reinforcing learning and encouraging ongoing application following the course.

Building a Community of Practice

One of our key goals is to strengthen the community of professionals working with foundational learning data across Africa. Networking opportunities were provided through group exercises, social events during the week, and ongoing support in the months following the course, with online sessions designed to help continue applying new skills.

Course Lecturers

Professor Cally Ardington

Dr Linda Zuze

Dr Jack Thunde

Ms Asanda Lobelo

Dr Emma Whitelaw

Participants