What Makes Learner-Centered Teaching Effective? Protocol for a Systematic Review and Meta-Analysis of Pedagogical Approaches, Design Characteristics, and Contextual Moderators

Abstract

Learner-centered pedagogies are generally associated with favorable educational outcomes, but their effects vary substantially across instructional approaches, educational settings, outcomes, and implementation conditions. Existing meta-analyses have typically examined individual pedagogical traditions  such as cooperative, problem-based, inquiry-based, project-based, or flipped learning  using different eligibility criteria, outcome classifications, effect-size metrics, and analytical strategies. Consequently, the literature provides stronger evidence about whether particular approaches are effective on average than about the design characteristics and contextual conditions that may explain variation in their effectiveness. This protocol describes a systematic review and meta-analysis designed to address that gap. For the purposes of the review, learner-centered pedagogy is operationalized as instruction in which learners make substantive contributions to the learning process through active cognitive engagement, interaction, inquiry, application, reflection, or adaptation to learner needs, while teachers retain an intentional role in structuring and supporting learning. Eight pedagogical families will be examined: collaborative/cooperative learning, problem-based learning, inquiry-based learning, project-based learning, flipped learning, dialogic teaching, experiential/service learning, and personalized/adaptive learning. Rather than treating these labels as homogeneous interventions, the review will additionally code theoretically relevant design characteristics, including instructional guidance and scaffolding, interaction and group structure, intervention duration, teacher preparation, technological support, and implementation fidelity. The review will follow PRISMA-compatible procedures for systematic searching, independent screening, data extraction, risk-of-bias assessment, effect-size estimation, random-effects meta-analysis, treatment of statistically dependent effect sizes, moderator analysis, sensitivity analysis, and assessment of small-study and reporting biases. Academic achievement and learning performance will constitute the primary outcome domain, while cognitive, motivational, self-regulatory, social, and attitudinal outcomes will be analyzed as distinct secondary domains rather than combined into a single overall effect. Comparisons with conventional/business-as-usual instruction and comparisons with alternative learner-centered approaches will also be analyzed separately. No empirical results are reported because the systematic search and meta-analysis have not yet been completed. The intended contribution is to move beyond family-level judgments about whether learner-centered teaching “works” toward a theory-informed account of which design characteristics and contextual conditions are associated with stronger or weaker educational effects.