Application of Multiple Linear Regression in Understanding the Factors that Impact Academic Performance in Mathematics

Abstract

This article is an excerpt from the course completion work (TCC) carried out as part of the Degree in Mathematics Teaching course with specialization in Statistics at Pungue University, Tete Extension, in the academic year 2021. The study aims to analyze the factors that influence the academic performance in Mathematics of 10th grade students at Escola Secundária Heróis Moçambicanos de Moatize. An applied quantitative approach was used, with explanatory objectives and a statistical survey. Due to the large target population, convenience sampling was adopted, resulting in a sample of 421 students. Data were collected through a questionnaire with closed questions to support the research. Data analysis was carried out mainly using the multiple linear regression technique, assisted by descriptive and frequency analyses. The class analyzed is made up of six classes, divided into three sub-classes per class, totaling 452 students. The results revealed the construction of a multiple linear regression model, indicating that the variable "number of meals" (NRef) has the greatest impact on academic performance in Mathematics. The model suggests that students tend to have greater success in the subject when they eat, on average, three meals a day, as suggested by the distribution of the NRef variable.