Detección de estudiantes en riesgo académico en el Instituto Tecnológico de Costa Rica

Abstract

This study presents a statistical analysis conducted to determine a model which detects students at risk of academic failure at the Costa Rica Institute of Technology. The population is composed of students who were admitted to undergraduate programs in 1995. The dependent variable was the students´ performance for seven years: drop-outs in the first semester, drop-outs in the second semester, drop-outs in any other semester, students currently registered and graduate students. The independent variables were: the Institute’s entrance examination score and its components (High School, Math and Oral Components of the Academic Aptitude Test), type of high school (private or public, academic, technical or agricultural) geographic area, academic achievement indicators during the first semester, gender and major. A Logistic Regression Analysis model was used for an accurate classification of the 79.9% of the students in the categories of undergraduate and graduate. The model includes indicators of both academic performance and the number credits obtained in the students´ majors.