Supplementary Materials for "Beyond Grades: Multi-Target Predictive Modeling of Formative Academic Indicators in a Latin American Public University"

Le résumé fourni par la source

This repository contains supplementary materials associated with the article: “Beyond Grades: Multi-Target Predictive Modeling of Formative Academic Indicators in a Latin American Public University.” The materials support transparent and auditable reporting of the study’s identity-audited multi-target modeling workflow. They document the feature and target definitions, the algebraic target-reconstruction audit, contextual literature comparison, and exploratory supplementary analyses. They do not provide the complete institutional event-level extraction pipeline or raw academic records. Contents: Supplementary Appendix ASystematic-review coding matrix summarizing variable categories across 55 empirical studies of academic-performance prediction in higher education. This appendix supports Figure 2 in the manuscript. Supplementary Appendix A1Dataset variable dictionary, including formulas, units, source platforms, target and predictor roles, and identity-audit classification for the originally engineered variables and the final 15-predictor benchmark. Supplementary Appendix BContextual literature benchmark describing reported models, target types, data sources, validation designs, and metrics from prior studies. It is provided for methodological context rather than direct numerical performance ranking. Supplementary Appendix CExtended model-attribution and descriptive analyses, including SHAP visualizations, exploratory clustering outputs, and per-cutoff SHAP summaries. These analyses are reported as exploratory inspection tools, not as independent validation or causal evidence. Together, these materials complement the manuscript by documenting the target-reconstruction audit, feature-construction decisions, model-evaluation context, and the scope and limitations of the supplementary analyses.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.