Predição e monitoramento do desempenho acadêmico em uma IFES: uma análise de índices

Abstract

The relationship between school dropout and academic performance has been the subject of academic work in several higher education institutions, motivating studies that seek to identify and mitigate factors associated with academic failure. The objective of this work is to improve the process of identifying students with unsatisfactory academic performance at a federal higher education institution (IFES), focusing on scholarship holders from the National Student Assistance Program (PNAES) of the Dean of Student Affairs (PRAE). This work used Educational Data Mining (EDM) techniques, analyzed the approval rate and the CI index, and observed the correlation of these indicators with academic performance. The research used anonymized data from students entering between 2018 and 2023, and the analysis revealed that a minimum pass rate of 63.6% can be a relevant parameter to evaluate student performance. However, the lack of a universal predictive solution indicates the need for broader analysis and a more robust historical database to validate conclusions.Among the limitations of the study, the difficulty in obtaining complete data and the lack of some important indicators stand out. Based on the results, future studies are suggested that integrate a more extensive historical series, allowing the creation of a more accurate predictive model to evaluate students' academic performance and reduce school dropout rates.