From Retrospective Reporting to Predictive Intelligence: AI-Augmented Monitoring and Evaluation Systems in African Development Contexts

This dataset supports the article: From Retrospective Reporting to Predictive Intelligence: AI-Augmented Monitoring and Evaluation Systems in African Development Contexts (Sisimayi & Makoni).

The repository contains four files required to reproduce all reported results:

1. anonymised_survey_data.csv: Anonymised expert validation survey data from 56 M&E practitioners, academics, and development professionals across Africa (March–April 2026). All personal identifiers have been removed. Contains 44 analytical variables including five AA-MES layer rating scales, demographic variables, and open-text responses.
2. python_analysis_script.py: Full Python 3 analysis script (pandas, numpy, scipy, matplotlib, seaborn) reproducing all descriptive statistics, Delphi consensus analysis, layer composite scores, Cronbach alpha, inter-layer Pearson correlations, hypothesis tests (H1–H3), and all eight figures reported in the article.
3. qualitative_codebook.csv: Codebook for thematic analysis of open-text survey responses, containing 12 codes mapped to AA-MES framework layers with definitions, inclusion and exclusion criteria, and anchoring examples.
4. README.txt: Full variable dictionary, layer-to-column mapping, step-by-step reproduction instructions, and expected output values for verification.