The paper aims to compare the effectiveness of two learning models, namely Discovery Learning (DL) and Problem Based Learning (PBL), in improving student learning outcomes in science subjects. This type of research is experimental research. The population in this study were all class IV students at SDN 002 SK (Samarinda City), totaling three classes. The calculation results show that the average pretest score in experimental class I is 45.12, which is higher compared to 35.60 in experimental class II. After treatment, the average posttest score in experimental class I increased to 85.80, while in experimental class II it reached 82.11. The average N-Gain for both classes was 0.74, indicating a significant increase in learning outcomes in both classes, although experimental class I showed slightly better results. More in-depth analysis based on high and low learning outcome categories shows consistent differences in the effectiveness of learning models. Hypothesis testing using the T Test, which is applied because the data distribution is normal and the variance is homogeneous, shows a calculated t-value of 8.60 which far exceeds the t_table of 2.070. This causes rejection of H_0 and acceptance of H_1, indicating that the Discovery Learning (DL) model in experimental class I provides better learning results than the Problem Based Learning (PBL) model in experimental class II.
Read full abstract