Evidence-Based Education and Beyond: The Critical Role of Theory in Science Education Research and Practice

Evidence-based education has become a central concept in science education, with meta-analyses often regarded as the gold standard for informing practice. This emphasis raises critical questions concerning the applicability, generalizability and transferability of research findings into classroom practice. It remains unclear both what kind of evidence education should be based on and whether science education research can provide the type of evidence required to guide decisions at different levels. This paper argues that theories play a crucial role in building bridges between research and practice. Drawing on literature from science education and the philosophy of science, we contrast the explanatory scope of meta-analyses with the predictive and integrative potential of theories, understood in a structuralist sense as systems of models with defined domains of applicability. We propose that science education research requires both fundamental and applied research, each contributing to theory development at different levels, ranging from local and context-specific models to more fundamental theoretical frameworks. Importantly, we argue that theories in science education should not be viewed merely as applications of psychological or pedagogical theories, but as fundamental theories in their own right. We conclude that the future development of science education research may benefit more from the systematic refinement and integration of theories than from the continued accumulation of isolated local findings, and we propose ways to support the development of such theories. A theory-guided understanding of evidence-based education can strengthen the scientific foundations of the field while simultaneously enhancing its practical relevance, thereby helping to narrow the long-standing theory-practice gap.

💡 Research Summary

The paper offers a critical examination of the dominant “evidence‑based education” paradigm in science education, arguing that the current reliance on meta‑analyses as the gold standard for informing practice is insufficient. While meta‑analyses can synthesize effect sizes across studies, they often obscure contextual differences, yield modest and sometimes contradictory effects, and cannot provide universally applicable instructional guidance. To address these shortcomings, the authors foreground the role of theory, adopting a structuralist conception in which a theory is a system of models with clearly defined domains of applicability. Such a theory not only organizes phenomena but also generates general laws and predictions, much like Newtonian mechanics predicts the motion of diverse objects.

The authors contend that science‑education research must pursue both fundamental and applied investigations. Fundamental research uncovers universal mechanisms of learning and supplies the conceptual scaffolding of a theory, whereas applied research tests and refines that scaffolding in specific classroom settings. Both quantitative and qualitative data are essential, and teachers’ professional experience should be treated as a crucial “adjustment” component when applying a theory to a particular context.

A key problem identified is the superficial use of the term “theory” in the field—often reduced to a collection of empirical findings, pedagogical guidelines, or conceptual frameworks without a coherent explanatory structure. The paper calls for explicit definition of theories, specification of their scope, and clear articulation of their limits. In this view, meta‑analyses retain an important role, but not as the primary source of actionable knowledge. Instead, they serve to test existing theories, highlight gaps, and suggest directions for further theory development.

From a practical standpoint, the authors propose a “theory‑experience hybrid” approach to instructional decision‑making. Teachers should combine the general principles derived from a well‑developed theory with their contextual knowledge and classroom observations, mirroring the evidence‑based medicine model that integrates clinical expertise with research evidence. This hybrid model is presented as a way to rebuild trust among teachers who often distrust meta‑analytic recommendations that appear to ignore local realities.

To foster systematic theory building, the paper outlines several concrete strategies: (1) funding and institutional support for long‑term, theory‑driven research programs; (2) interdisciplinary collaboration among education, psychology, and the natural sciences to construct comprehensive models; (3) the creation of a meta‑theoretical framework that links meta‑analytic findings directly to theory testing and refinement; and (4) the establishment of continuous feedback loops between researchers and practitioners so that classroom experience informs theory revision.

In conclusion, the authors argue that re‑orienting evidence‑based education toward a theory‑guided paradigm will strengthen the scientific foundations of science education, enhance its relevance for teachers, and narrow the persistent theory‑practice gap. By moving beyond the mere accumulation of isolated empirical results toward the systematic refinement and integration of robust theories, the field can provide more reliable, generalizable, and context‑sensitive guidance for educational policy, research, and classroom practice.