Abstract

DIAZ MACIAS, Efraín  and  GUANIN FAJARDO, Jorge. Generalized mixed linear model for the detection of successful first-year system engineering students. Conrado [online]. 2020, vol.16, n.73, pp. 138-142.  Epub Apr 02, 2020. ISSN 1990-8644.

The importance of anticipating educational outcomes is essential to intervene and control the factors that lead to academic failure. It is important for higher education institutions to have tools that serve for the in-depth analysis of student data. In order to do so, the influence of variables linked to the passing of the first year of the systems engineering degree was evaluated. Thus, the data set of students from 4 study periods with 5325 observations and 8 categorical variables was used. Hence, the generalized linear mixed model with binominal response variable (success/failure) with random effects was used. The results of the variables linked to academic success are related to the model obtained, through the contrast of fixed and random models that may affect successful students. The validation of the models was carried out by means of the Anova type III test. In the light of the results, the findings suggest a significant model, related to the tertile of grades of the first partial, quartile of attendance to classes, subject and teachers body. In addition, they provide timely insight into decision making to educational administrators.

Keywords : Educational analysis; higher education; generalized mixed linear model; academic performance.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )