<![CDATA[The development of generic science skills is a crucial objective in contemporary science education to address the demands of 21st-century competencies. This study aims to explore the potential of a Hybrid STEM–Project-Based Learning (PjBL) model supported by web-based interactive learning environments and deep learning approaches in enhancing generic science skills within physics learning materials. A Hybrid Review methodology was employed, integrating a Systematic Literature Review (SLR) and a Bibliometric Review. A total of 30 peer-reviewed articles indexed in Scopus and SINTA, published between 2020 and 2026, were systematically analyzed. Bibliometric mapping was used to identify research trends, thematic clusters, and emerging research gaps, while the SLR examined instructional designs, learning outcomes, and pedagogical effectiveness. The results indicate that STEM–PjBL consistently improves students’ generic science skills, including scientific reasoning, problem-solving, data interpretation, and conceptual modeling, particularly when supported by interactive web-based platforms. These platforms facilitate visualization, collaboration, and iterative inquiry processes that are essential in learning abstract physics concepts. Furthermore, the findings highlight that deep learning approaches offer strong potential to provide adaptive scaffolding, personalized feedback, and learning analytics to support students’ inquiry processes, although their implementation in STEM–PjBL contexts remains limited. This study concludes that the integration of STEM–PjBL, web-based interactivity, and deep learning constitutes a promising and scalable framework for advancing generic science skills and provides important implications for future research and instructional design in science education.]]>
Copyrights © 2026
