Deep Learning Approach and Students’ Numeracy Skills: The Roles of Mathematical Disposition and Gender

Abstract

International and national assessments, including PISA 2022, indicate that Indonesian students’ numeracy performance remains low. This situation highlights the need for instructional approaches that strengthen conceptual understanding and higher-order thinking. This study examined the effect of a deep-learning instructional approach on eighth-grade students’ numeracy skills, while considering mathematical disposition and gender. We employed a non-equivalent control group quasi-experimental design with 62 Grade 8 students from a public junior high school in West Lombok, Indonesia. The experimental group (n = 31) received deep-learning instruction through concept exploration, context-based discussion, collaborative problem solving, and guided reflection, while the control group (n = 31) received regular instruction. Instruments included a numeracy test and a mathematical disposition questionnaire that underwent expert validation and met reliability criteria. We analyzed the data using descriptive statistics and ANCOVA, controlling for baseline ability (pretest). The results showed higher numeracy scores in the experimental group across disposition categories. ANCOVA indicated significant main effects of instructional approach, gender, and mathematical disposition, whereas several interaction effects were not significant. These findings support the use of deep-learning instruction as a promising strategy to strengthen numeracy within the Merdeka Curriculum implementation.