PREDICTING STUDENT COMPLETION STATUS USING LOGISTIC REGRESSION ANALYSIS

Abstract

To fulfil the mandate of the Polytechnics, which is to fill the gap in the manpower supply needs of the country, in an appropriate and timely manner, students must perform well and successfully complete their course of study in a timely manner. A failure of a student in a subject could among other things, increase the number of credit hours a student takes in a particular semester, affect the academic performance of the student and delay the time of certification of that student. Using longitudinal data of students’ academic performance in each semester, this study used the logistic regression technique to study the influence of Gender, mode of entry, study intensity and the Age of the student on the likelihood of a student completing without ever failing in any subject in the Bolgatanga Polytechnic. The results showed that Gender, Mode of entry, Study intensity and the Age of the student, all significantly influenced the completion status of a student. Among other things, Female students were shown to be twice as likely to complete their study without ever failing in any subject as compared to their male counterparts. The results suggest that some targeted interventions, for particular student segments, could minimize the possibility of student failing.