Mathematics anxiety and problem-solving ability in numerical methods courses: A repeated cross-sectional study across two Southeast Asian universities

Abstract

Mathematics anxiety (MA) impedes mathematical learning by taxing cognitive resources, negatively affecting problem-solving ability (PSA). Some evidence suggests that the strength and nature of this negative association may vary across educational and cultural contexts. However, cross-cultural studies on MA and PSA, especially in Southeast Asia, remain limited. This study investigated cohort-level patterns of MA and PSA in a transnational Numerical Methods course at Universiti Tun Hussein Onn Malaysia (UTHM) and Universitas Pahlawan (UP), Indonesia. The study spanned four independent cohorts from 2021 to 2024 and focused on institutional differences and temporal variation during a period of instructional transition. Using a repeated cross-sectional design, a total of 139 undergraduate students across four independent cohorts (UTHM: n=23; UP: n=116) were assessed during the 2021–2024 academic years using the Mathematics Anxiety Rating Scale and standardized problem-solving tests. Two-way factorial ANOVA examined the effects of university and cohort year on MA and PSA, and Pearson correlation analysis examined the association between the two constructs. A significant main effect of university indicated that UP students reported higher MA and lower PSA scores than UTHM students (p < 0.001). Significant year effects were observed for MA and PSA scores across successive cohorts (p ≤ 0.01), during a period characterized by a transition in instructional modalities from online to hybrid learning. Significant University×Year interactions indicated differing trends by institution. A moderate negative correlation between MA and PSA was observed (r = –0.49, p < 0.001).