Data Mining of TIMSS 2023 to Identify the Factors Influencing Mathematics Achievement of Fourth-Grade Students in the Kingdom of Saudi Arabia

Abstract

This study aimed to identify the key factors influencing mathematics achievement of fourth-grade students in the Kingdom of Saudi Arabia by applying data mining techniques to TIMSS 2023 data using the Decision Tree Algorithm. A total of 22 categorical measures extracted from the student, school, home, and teacher questionnaires were included as independent variables, while mathematics achievement was treated as a continuous dependent variable, represented separately through the five plausible values. The results of the five models revealed three recurring behavioral patterns among students as follows: (1) students with low confidence in mathematics, who are exposed to bullying on a weekly or monthly basis and have low digital self-efficacy, with expected scores ranging between 376–382; (2) students with similarly low confidence in mathematics and frequent exposure to bullying, but with high digital self-efficacy, with expected scores ranging between 414–418; and (3) students with high confidence in mathematics, whose expected scores ranged between 459–464. Analysis of variable importance indicated that students’ confidence in mathematics had the highest influence (mean importance = 0.514), followed by exposure to bullying (0.265), and digital self-efficacy (0.134). The findings also pointed to a relative stability in the predictive structure of the decision tree across