Critical thinking is one of the essential 21st-century competencies that should be fostered through biology learning. However, classroom instruction remains predominantly teacher-centered, limiting students' opportunities to develop higher-order thinking skills. This study aimed to analyze the effectiveness of ChatGPT-integrated Science, Technology, Engineering, and Mathematics (STEM) learning in improving students' critical thinking skills in biology. A quantitative approach with a quasi-experimental design employing a pretest–posttest control group design was used. The study involved 66 tenth-grade students from SMA Pergis Maros, divided equally into an experimental group and a control group. The experimental group received biology instruction through ChatGPT-integrated STEM learning, while the control group was taught using conventional instruction. Data were collected using a critical thinking test administered before and after the intervention and analyzed through descriptive statistics, normality and homogeneity tests, an independent samples t-test, gain score analysis, and simple linear regression. The results revealed that the experimental group achieved a higher posttest mean score (80.70) than the control group (70.28), with gain scores of 19.58 and 9.69, respectively. The independent samples t-test indicated a statistically significant difference between the two groups (t = 10.33, p < 0.05), demonstrating that ChatGPT-integrated STEM learning effectively enhanced students' critical thinking skills. Therefore, integrating ChatGPT as an artificial intelligence learning assistant within STEM-based biology instruction represents an innovative and effective instructional strategy for promoting critical thinking and improving the quality of biology learning.
Copyrights © 2026