Mi SciELO
Servicios Personalizados
Articulo
Indicadores
- Citado por SciELO
Links relacionados
- Similares en SciELO
Compartir
EDUMECENTRO
versión On-line ISSN 2077-2874
Resumen
MARTINEZ PEREZ, José Ramón; PEREZ LEYVA, Elmer Héctor; FERRAS FERNANDEZ, Yenny y BERMUDEZ CORDOVI, Lourdes Leonor. Predictive analysis on student dropout in the Medicine degree. EDUMECENTRO [online]. 2021, vol.13, n.3, pp. 217-236. Epub 30-Sep-2021. ISSN 2077-2874.
Background:
school dropout should be analyzed in a multivariate context to identify its causes and effects; in no way, it should be attributed to a single cause.
Objective:
to determine the predictive capacity of some factors on the school dropout of medical students, through a multiple logistic regression model.
Methods:
an analytical, predictive study was carried out in 87 medical students enrolled in the 2015-2016 academic year. Theoretical and empirical methods were applied and it was carried out in two stages: in the first, the variables most associated with school dropout were identified through a bivariate analysis; and in the second, the ability of these variables to predict dropout was analyzed through logistic regression (multivariate analysis).
Results:
in the bivariate analysis, nine variables showed a significant relationship with school dropout; when subjected to multivariate analysis (correlation and logistic regression), only four maintained statistical significance, that´s why they were finally chosen as predictor variables.
Conclusions:
school dropout in Medicine students can be predicted by the synergistic combination of the four predictors: dedicating less than 15 hours per week to study, female sex, school repetition and low academic performance in Morphology-physiology.
Palabras clave : forecasting; logistic models; student dropouts; underachievement; education, medical..