Analysis of generic competencies using a Linear Regression Model: A case study in short-cycle programs

Abstract

Introduction. The Saber tests, administered by the Colombian Institute for the Evaluation of Education (ICFES), assess students’ competency performance across educational levels. This study evaluates the progress in competencies among students enrolled in three technological programs (online learning, in-person learning, and dual-online learning) at a university in eastern Colombia by comparing their Saber 11 and Saber TyT (refers to the Technical and Technological Knowledge Test) test results. Materials and Methods. This descriptive, cross-sectional study analyzed data from students who completed the Saber 11 test beginning in 2014 and subsequently the Saber TyT test starting in 2016. A multiple linear regression model and association tests were applied. Results. Among the 3.095 graduates, only 14.7% had complete scores for both tests, limiting the scope of the analysis. A moderate correlation was observed between Saber 11 and Saber TyT competency scores, with stronger relationships in English (r(409) = 0.65; p < 0.01) and quantitative reasoning (r(409) = 0.60; p < 0.01). Accompanying programs positively impacted these areas, whereas citizenship competencies and written communication demonstrated less consistent performance. Study modality and cumulative GPA significantly influenced Saber TyT results. These findings provide actionable insights for refining educational strategies to enhance retention and advance critical competencies. Conclusions. This study highlights the importance of adapting pedagogical strategies to integrate essential competencies into academic programs, supporting institutional self-evaluation and accreditation efforts