INVESTIGATING FACTORS AFFECTING STUDENTS’ PERFORMANCE IN MATHEMATICS USING GENERALIZED ESTIMATING EQUATION

Abstract

The method of Generalized Estimating Equation has been of tremendous use in the research world of analysis of correlated data. Data collected for a single subject over several time frame exhibit this property which makes the use of certain statistical methods non-suitable for analysis. One of the major problems encountered in such data is the modelling of the within- subject correlation that accounts for the structure of correlation that exists between observation belonging to a particular subject, when such information is ignored, it leads to gross invalidity of inferences. This research work showed the study of continuous response variable, using GEE to estimate the parameters in a fitted model. It is an application to a real life education related data. The dataset was obtained from the Department of Mathematics, Federal University of Technology Akure, Nigeria. It comprises of students’ scores in Mathematics for three consecutive semesters and the corresponding load units in these semesters for two groups (adult and teenagers). 60% of the sample size is made up of Adult (age ≥20), while 40% are teenagers (age ≤19). The four mostly used “working correlation structure” (Independent, Autoregressive (AR1), Exchangeable and Unstructured) are used to fit four different models using R programming Language, where the model fitted using Exchangeable “working Correlation matrix” is adjudged the best based. The effect of Time, load unit and the interactive effect of load unit over time were found to be highly significant on the students’ scores. The box-plot revealed that the Adult performances for the first two semesters are above the average. However, category has no significant effect on scores.