Persistence in Distance Education: A Study Case Using Bayesian Network to Understand Retention

Abstract

This article presents a study on the variables promoting student retention in distance undergraduate courses at Federal University of Pará, aiming to help school managers minimize student attrition and maximize retention until graduation. The theoretical background is based on Rovai's Composite Model and the methodological approach is conditional probability analysis using the Bayesian Networks graphical model. Network modeling has shown that among internal factors after admission to the course (as defined in the Composite Model) face-to-face tutorial sessions need to be better planned and executed, learning materials are still not adequate to online course specificities and the support structure needs to be remodeled.