Meta-Analysis of Systems Thinking Studies in Science Education

This study undertakes a meta-analysis of the systems thinking approach, which has become increasingly significant in science education. A detailed literature review identified preeminent authors and synthesized the key features of systems thinking into a table. Scholarly articles for the research were carefully chosen according to well-defined criteria from established databases, including Web of Science, SCOPUS, ERIC, and Science Direct. Of 31 articles that employed quantitative or mixed methods and included the keyword "Systems Thinking," 12 were retained after duplicate removal and criterion application. To enhance analytical reliability, the Bias Tool and PRISMA-P protocols were employed. This study primarily explored how systems thinking influences academic outcomes in science education, assessing its utility across different sub-disciplines and the effectiveness of various educational strategies and methods. Key statistical metrics such as sample sizes, standard deviations, and means were collected. Additionally, Cohen's effect sizes and 95% confidence intervals for 26 variables drawn from the studies were calculated and illustrated in a forest plot. A heterogeneity analysis was conducted using Cochran’s Q test and the I² statistic, yielding a non-significant Q value (Q = 1.784, p = 1.000) and an I² value of 0%. These results confirmed the absence of substantial heterogeneity among the studies, supporting using a fixed-effects model to compute the overall effect size, estimated at ".47". This outcome reveals a moderate level of efficacy for the systems thinking approach within science education, indicating its advantageous use across various sub-disciplines.