This study aimed to determine the effect of Project-Based Learning (PjBL) based on deep learning on fifth-grade students' learning outcomes in the ecosystem topic at SD Negeri 190 Pekanbaru. This study employed a quantitative approach with a quasi-experimental method and a Nonequivalent Control Group Design. The sample consisted of 58 students, with 29 students in class V A as the experimental group and 29 students in class V B as the control group. The experimental group received PjBL based on deep learning with a concrete ecosystem aquarium, while the control group received the learning commonly used by the teacher. Data were collected through multiple-choice learning outcome tests administered as pretests and posttests. Data were analyzed using normality, homogeneity, N-Gain, Paired Samples T-Test, and Independent Samples T-Test. The results showed that the experimental group increased from a mean pretest score of 67.62 to a posttest score of 89.28, with a mean N-Gain of 0.65. The control group increased from 64.41 to 79.55, with a mean N-Gain of 0.41. The Independent Samples T-Test yielded t = 5.896, df = 56, and Sig. (2-tailed) = 0.000 < 0.05. Cohen's d was 1.55, indicating a large effect. Therefore, PjBL based on deep learning had a significant and large effect on fifth-grade students' learning outcomes in the ecosystem topic.
Copyrights © 2026