Meta-Analysis of the Impact of the STEAM Approach on Science Learning

The STEAM approach (Science, Technology, Engineering, Arts, and Mathematics) has emerged as a widely adopted instructional strategy aimed at developing essential 21st-century competencies such as critical thinking, creativity, collaboration, and scientific literacy. Numerous studies have explored the implementation of STEAM in science education, yet the results remain inconsistent due to variations in context, research variables, and educational levels. This inconsistency highlights the need for a comprehensive review that consolidates previous findings. Meta-analysis is a suitable method for this purpose, as it enables the calculation of the overall magnitude of STEAM’s impact. This study aims to examine the overall effect of the STEAM approach in science education. The analysis is categorized by educational level, research variables, and instructional models integrated with STEAM. A quantitative meta-analysis method was employed. Data were collected using Publish or Perish software with sources from Google Scholar and Scopus. The selection process followed PRISMA guidelines. From an initial pool of 1,197 articles, 12 studies met the inclusion criteria and were further analyzed using JASP software version 0.19.3.0.The results indicate that the STEAM approach has a positive influence on science learning. The overall effect size was 0.213, which falls within the moderate category. When analyzed by educational level, the effect size was 0.200 (low category); by research variable, 0.221 (moderate category); and by instructional model, 0.234 (moderate category). These findings suggest that STEAM is an effective approach for enhancing the quality of science education and holds considerable potential as a pedagogical strategy across diverse learning contexts.