The characteristics of environmental change material are dynamic, so instruction is required to produce innovative solutions through Critical Thinking Skills (CTS) as a key to responding to the complexity of the problem. However, its development in science education is still limited, due to the lack of learning designed to train CTS through scientific processes in exploring phenomena. This limitation impacts the low CTS of students in solving problems systematically and complexly in real life. STEAM (Science, Technology, Engineering, Arts, and Mathematics) is an innovative pedagogical framework to address this gap. The study aimed to determine the effectiveness of STEAM implementation in improving student CTS in science learning. This study used a quasi-experimental method with a non-equivalent control group pretest-posttest design because of the principle of randomization in educational research. 126 students were selected as a sample in a cluster relevant to the research design, which was non-random, so the results reflect contextual realism. The data collection instruments used were test and non-test instruments. Data analysis techniques used descriptive and inferential analysis through ANOVA and n-gain. The results showed a significant difference in CTS between students who studied STEAM and those who did not (0.001 < 0.050), with a moderate category (0.3≤g≤0.7). The implication is that STEAM creates a cognitive dissonance that encourages students' CTS habits by questioning assumptions, connecting concepts, and evaluating solutions from multiple perspectives. Supporting data for improving CTS through STEAM is proven by producing products with an average value of 89.33, and the collaboration has an average value of 96.10.
Sari et al. (2025) studied this question.