Analysis of Students’ Numeracy Literacy in Terms of Self-Efficacy through Problem-Based Learning (PBL) Supported by Flipbooks with an Ethnomathematics Approach

Students’ numeracy skills in Indonesia remain relatively low due to a lack of attention to the aspect of self-efficacy. In addition, there is a need for instructional models and media capable of improving numeracy skills. This study aims to analyze students’ numeracy skills in terms of self-efficacy following the implementation of an ethnomathematics-based Problem-Based Learning (PBL) model aided by a flipbook. This study employed a mixed-methods approach with an explanatory sequential design, involving 32 eighth-grade students at SMP Negeri 13 Semarang during the 2025/2026 academic year. Instruction focused on the topic of function relationships, integrating the ethnomathematics of Semarang City. The research instrument consisted of a numeracy literacy test that met validity criteria (a significance value of  < 0,05 for each item) and was deemed reliable based on a Cronbach’s Alpha coefficient of 0,640. The data were analyzed using descriptive and inferential statistics through normality tests, individual achievement tests, class achievement tests, simple linear regression tests in the experimental class, and descriptive analysis based on self-efficacy categories. The results of the study indicate that there was an increase in numeracy literacy skills in the class taught using Problem-Based Learning with an ethnomathematics approach aided by a flipbook, and that self-efficacy had a positive and significant effect on numeracy literacy skills with a coefficient of (R^2) of 40,5%  determination  of  We also found that students with high self-efficacy met all numeracy literacy indicators, whereas students with moderate and low self-efficacy met only some of the indicators. These findings indicate that the implementation of ethnomathematics infused Problem Based Learning supported by flipbooks effectively improves students’ numeracy literacy skills, as evidenced by a higher N-Gain of 0,74 in the experimental class which falls into the high category compared to the control class’s N-Gain of 0,59, which falls into the moderate category.