Kaplan-Meier Method. Partial Modeling Survival Analysis: A Methodological Alternative for Studying Student Trajectories in Higher Education

Abstract

This article explores the application of advanced statistical methods, specifically survival analysis, to examine academic trajectories in higher education, underscoring its advantages over traditional models that often struggle with incomplete data or events of uncertain duration. In particular, the Kaplan-Meier method, Cox regression, and the log-rank test are highlighted for their ability to handle censored and uneven data, offering a more accurate perspective on the timing of key events such as dropout or graduation. This methodological approach not only enables comparisons of academic pathways across different groups but also provides a deeper understanding of the time spent in the academic process, going beyond traditional success-or-failure analyses. By focusing on time as the main variable, the article illustrates how these statistical tools can enhance educational research and support decision-making in the university setting. The goal is to understand when and how critical educational events occur and how they depend on various factors, emphasizing the need to rigorously adapt methods to the specific study objectives. This approach strengthens the interpretation of results and aids in understanding how students make decisions about continuing or discontinuing their studies.