Comparative Effectiveness of ICT-Based APOS and M-APOS Learning Models on Students' Self-Efficacy Viewed from Initial Mathematical Ability

Handayani, Ita
Husnul, Nisak Ruwah Ibnatur

Prisma Sains: Jurnal Pengkajian Ilmu dan Pembelajaran Matematika dan IPA IKIP Mataram
Vol. 13 No. 3: July 2025

Statistics is an essential course in the curriculum of the Informatics Engineering Study Program, as it plays a crucial role in equipping students with skills in data analysis, information processing, and data-driven decision-making. However, various studies have reported that students often experience low self-efficacy in statistics, which can hinder their academic performance and engagement. This study aims to compare the effectiveness of the ICT-based APOS model and the M-APOS model on students’ self-efficacy, taking into account their Initial Mathematical Ability (IMA). The APOS theory provides the foundation for the instructional models used, while Bandura’s theory of self-efficacy underpins the construct being measured. A quasi-experimental method was employed using a 2 × 2 factorial design (treatment by level). The data analyzed were students’ scores on a validated and reliable self-efficacy questionnaire, and a two-way ANOVA was used as the analytical technique. Prerequisite tests included the Lilliefors test for normality and Levene’s test for homogeneity. The rationale for selecting the ICT-based APOS and M-APOS models lies in their potential to address cognitive and metacognitive learning challenges in mathematical thinking. The results showed that, overall, students taught using the ICT-based APOS model demonstrated higher self-efficacy than those taught with the M-APOS model. A significant interaction was found between the learning model and IMA on students’ self-efficacy. Students with high IMA who received instruction through the ICT-based APOS model exhibited greater self-efficacy than those who experienced the M-APOS model. Among students with low IMA, there was no significant difference. These findings suggest that integrating technology into constructivist learning models such as APOS can enhance students’ confidence in learning statistics, particularly for those with a strong mathematical foundation, and inform future innovations in adaptive, ability-sensitive instructional design.