An Audit-Before-Modeling Framework for Reliability, Fairness, and Interpretable IRT Analysis of Mental-Health Self-Report Scales

This project documents a research-grade psychometric audit pipeline implementing an explicit audit-before-modeling workflow for mental-health self-report scales. The framework conducts sequential diagnostics of data integrity, internal consistency, and measurement fairness prior to any latent or predictive modeling, with optional interpretable latent analysis using Item Response Theory. The project is intended for methodological research and education rather than clinical or diagnostic use. The canonical software release is archived on Zenodo with a persistent DOI, and the full implementation is available via a public source-code repository.