A Logistic Regression Model to Identify Factors Influencing Secondary School Students’ University Attendance Decision in the Southern Part of Sierra Leone

International Journal of Research and Innovation in Social Science (IJRISS) | Volume V, Issue VI, June 2021 | ISSN 2454–6186

A Logistic Regression Model to Identify Factors Influencing Secondary School Students’ University Attendance Decision in the Southern Part of Sierra Leone

Regina Baby Sesay
Department of Mathematics and Statistics, School of Technology, Njala University, Njala, Sierra Leone

Abstract University education does not only increase earnings by providing skills that increase the chance of employment opportunities, but also promotes the economic growth of the country concern. For a developing country like Sierra Leone, a step to attain university education is a step towards moving away from poverty. It is therefore the desire of each and every parent to see their children through university level education. However, there are some unavoidable factors that may influence the ability and desire of provincial secondary (high) school students to further their education to university level. In Sierra Leone today, despite the increasing number of school dropouts, no clear-cut research has been carried out to address this great issue of concern. This research work, therefore, used a binary logistic regression modelling technique to identify the main factors influencing students’ decision to attend a university of their choice after secondary (or high) school education. For this purpose, a stratified random sampling method was employed to select 363 respondents proportionately from each of the secondary schools in the Mokonde community, Korie chiefdom, Moyamba District, southern part of Sierra Leone. Data were collected from the selected respondents using structured questionnaires. To further ascertain the appropriateness of the chosen binary logistic regression model, two additional regression models, the proportional odd ordinal logistic regression model and the unconstrained partial proportional odd ordinal logistic regression model were also used in the analysis. However, statistical tests showed that the chosen binary logistic regression model outperformed the two ordinal regression models. Based on the result of the empirical analysis, the gender of the secondary (high) school student; the father’s income level; the mother’s income level; the annual average score and the number of study hours are the main factors influencing student’s university attendance decision in the study area. Male secondary (or high) school students are more likely to attend university than their female counterparts. High school students whose fathers are on a high level income scale are more likely to attend university than those whose fathers are on the low level income scale. Also, the higher the average score of the secondary school student the greater the possibility of the student to enter university and above all, the more hours the student spends on studying his or her academic work, the greater the possibility for the student to enter university.

Keywords: Logistic Regression; Predictors, Sierra Leone; Variance inflation Factor; Binomial Distribution, University Attendance

І. INTRODUCTION

University is a place where people prepare for life as it impacts knowledge, and inject career values that contribute to long-term knowledge that may leads to a stable economic growth. In a developing country like Sierra Leone, where most of the inhabitants living in the provinces are subsistence farmers, university education can be the only means of escape from the poverty that is already evident in the lives of most of the provincial inhabitants. University education does not only increase earnings by providing skills that increase the chance

Read More