Identification of Early Predictors of Adult Learners’ Academic Performance in Higher Education

Abstract

Universities are inundated with detailed applicant and enrolment data from a variety of sources. However, for these data to be useful there is a need to convert them into strategic knowledge and information for decision-making processes. This study uses predictive modelling to identify at-risk adult learners in their first semester at SIM University, a Singapore University that caters mainly to adult learners. Fourteen variables from the enrolment database were considered as possible factors for the predictive model. To classify the at-risk students, various algorithms were used such as a neural network and classification tree. The performances of the different models were compared for sensitivity, specificity and accuracy indices. The model chosen is a classification tree model that may be used to inform policy. The implications of these results for identification of individuals in need of early intervention are discussed. [ABSTRACT FROM AUTHOR]