CausalMedia-GH Role 2: Synthetic Ghanaian Educational Context Generator

This deposit contains the synthetic data generation and validation pipelinefor Role 2 (Synthetic Context Generation & Validation) of the CausalMedia-GHproject. The pipeline generates Ghana-specific educational contextualvariables (school location, tablet access, bandwidth, school resource level,teacher qualification) using two synthesizers — CTGAN and Gaussian Copula —trained on a 10,000-record seed dataset (SDV library). Gaussian Copula was selected as the final synthesiser based on SDMetricsquality evaluation (overall quality score: 0.9948, vs. CTGAN's 0.8846).Statistical validation via Chi-square goodness-of-fit tests confirmed allfive contextual variables pass at p > 0.05. This synthetic contextual layer is designed to supplement a real Ghanaianfield validation subset (Role 2b, forthcoming) and serves as atraining-phase context layer only. It does not represent real Ghanaianschools; distributions [are derived from published Ghanaian educationalstatistics / reflect stated modeling assumptions where no confirmed sourcewas available — delete whichever doesn't apply]. Contents: SDV pipeline code, seed dataset schema (metadata.json), andSDMetrics evaluation report. This work operated under Path D ethics governance and is consistent withthe Ghana Data Protection Act, 2012 (Act 843).