Prediction model of first-year student desertion at Universidad Bernardo O´Higgins (UBO)^*

Abstract

The objective of this study is to model a retention predictive system for first-year students at Universidad Bernardo O’Higgins (Santiago de Chile), by determining which of the variables of entry into higher education, whether these are academic, social or relatives, are revealed significant for this analysis. The construction of the research model was based on a thorough bibliographic review which made possible to identify explanatory variables of university dropout in the national context. Afterwards, from the systematization of socio-educational backgrounds of the students from the 2014 and 2015 cohorts available in the university’s computer systems, a tripartite matrix was consolidated with the data associated with the variables that emerged from the analysis of the target group consulted. Consequently, we analyzed the relationship of each of the explanatory variables of the study with the variable control student desertion. The bivariate analysis allowed us to identify seventeen variables, significantly associated with student desertion and to specify dependency relations with the abandonment of studies. The multivariate model predicted abandonment behavior in 86.4%, indicating seven independent categorical variables that, finally, are revealed as relevant factors of the prediction model. The varied and sustained interpretations delivered in the results of the model, as well as the proposed suggestions to improve the university retention index, provide a direct value to the study aimed at optimizing one of the most important indicators linked to the quality management in universities, as is the student retention.

Students; Student retention; Entrance behavior variables; Higher education