Integration of Paired Sample T-Test and Cohen's d in the Evaluation of a Palembang Ethnoscience-Based Physics Module

Abstract

This study aimed to analyze the effectiveness of a Palembang ethnoscience-based physics module on student learning outcomes in potential and kinetic energy materials using a methodologically rigorous approach. Employing a quasi-experimental design with a one-group pretest-posttest pattern, the study involved 37 grade X students from SMAN 22 Palembang. Data were collected through learning achievement tests and analyzed using SPSS version 26, incorporating Shapiro-Wilk normality tests, Paired Sample T-Tests, Normalized Gain (N-Gain), Power Analysis, and outlier examinations. The results indicated mean pre-test scores of 61.22 (SD = 13.41) and post-test scores of 82.97 (SD = 9.96). The Paired Sample T-Test yielded a t-value of -11.217 with a significance of 0.000 (p < 0.05), indicating a significant difference and a strong association between module implementation and learning outcome improvements. The N-Gain value of 0.56 was categorized as medium, while the Effect Size (Cohen's d) of 1.84 demonstrated a very large impact, although this requires careful interpretation due to the within-subject design. Cohen's d_z, calculated based on the standard deviation of difference scores, also produced a value of 1.84, confirming the consistency of the effect estimation. The study concludes that the Palembang ethnoscience-based physics module is positively associated with enhancing student learning outcomes. Integrating local wisdom bridges abstract concepts with students' real-life contexts, thereby rendering learning more meaningful.