Multiple linear regression of Asia Pacific University Malaysia students' performance in statistics and mathematics course using R software

Abstract

The study aims to investigate the factors associated with statistics and mathematics performance among university students by using multiple linear regression, a popular method of predicting students' academic performance, to establish a prediction model. The demographic characteristics were explored as well as the differences in overall score between local and international students were identified using independent t—test. The findings concerned on the relationship between the performances in statistics and mathematics with six significant factors; (1) math diagnostic test score, (2) in—course score, (3) examination score, (4) programme, (5) nationality and (6) students' perception on the course content and organization. The Adjusted R—squared, R-Adj = 0-9599 indicates that 95.99% of total variability in students' performance can be explained by set of all six predictor variables.