The effect of the game-based learning model based on bar diagram board medium on students' mathematical representation ability

Authors

  • Yovita Puspasari STKIP PGRI Trenggalek Author
  • Wahyu Nurmalasari STKIP PGRI TRENGGALEK Author
  • Hidayatus Samawiya STKIP PGRI TRENGGALEK Author

DOI:

https://doi.org/10.26877/70dtxv70

Keywords:

Bar Diagram Board Medium, Game Based Learning, Mathematical Representation Ability

Abstract

Students' ability to represent mathematical concepts verbally, visually, and symbolically is a crucial skill in learning mathematics, as it supports understanding and the meaningful expression of mathematical ideas. Observations and interviews at SDN 2 Ngantru showed that, out of 38 students, only 6 (15.79%) met the minimum competency achievement criteria (KKTP), indicating the need for a learning model capable of fostering student engagement, one of which is Game-Based Learning supported by a bar-diagram board medium. This study aims to determine the effect of the Game-Based Learning model, using a bar-diagram board medium, on the mathematical representation ability of fourth-grade elementary school students. Using a quantitative approach with a quasi-experimental Non-Equivalent Control Group Design, the study involved 50 fourth-grade students at SDN 2 Ngantru, selected through saturated sampling, comprising 25 students from class IV-A (control group) and 25 from class IV-B (experimental group). Data were collected through tests, observation, and questionnaires. The data analysis technique used prerequisite tests (normality and homogeneity), an Analysis of Covariance (ANCOVA) with the pretest score as a covariate to control for initial ability differences between groups, an Independent Samples t-test as a complementary analysis, and an N-Gain effectiveness test. The ANCOVA results showed a statistically significant difference in posttest scores between groups after controlling for the pretest score (F = 48.156; p < 0.001; partial eta squared = 0.506), consistent with the hypothesis test results (significance value of 0.001 < 0.05) indicating that the null hypothesis was rejected. The experimental class's average posttest score of 83.36 was higher than the control class's 76.48, with an N-Gain of 60.18% (fairly effective) for the experimental class and 41.83% (less effective) for the control class.

Downloads

Published

2026-09-29