Deep Learning-Based Instruction on Ecosystem Topics to Enhance Students’ Critical Thinking Skills

Abstract

This study aimed to analyze the improvement and differences in seventh-grade students’ critical thinking skills in ecosystem learning through deep learning-based instruction. A quantitative approach with a quasi-experimental method and non-equivalent control group design was employed. The sample consisted of 56 seventh-grade students at Junior High School, comprising 28 students in the experimental class and 28 students in the control class. The experimental class received deep learning-based instruction, while the control class received conventional instruction. Data were collected using a tiered multiple-choice critical thinking test administered before and after instruction. The data were analyzed using descriptive statistics, N-Gain, the Shapiro-Wilk test, and an independent-samples t-test. The mean score of the experimental class increased from 66.21 to 84.54, while the control class increased from 70.14 to 82.07. The N-Gain scores were 0.54 for the experimental class and 0.39 for the control class, both categorized as moderate. The independent-samples t-test indicated significant differences between the classes in the pretest (p<0.001) and posttest (p=0.035). The highest N-Gain in the experimental class was found in the inference indicator (0.67), whereas the lowest was found in the focus indicator (0.42). These findings indicate that deep learning-based instruction supported the development of students’ critical thinking skills and resulted in descriptively higher improvement than conventional instruction. However, differences in initial ability between the groups should be considered when interpreting the findings.