Follow transparent, replicable meta-analysis practices tailored to PBL: multiple databases, gray literature, full heterogeneity reporting, funnel plots with Egger's test, and RVE for dependent outcomes¶

Strategy · All strategies
Evidence · no claims cited

Description¶

This strategy distills the article's recommendations for conducting a PBL meta-analysis. Search should be "transparent, replicable, and fully cover relevant research," using multiple databases (ERIC, PsychInfo, Digital Dissertations, ScienceDirect, Scopus, Web of Science, plus health databases for medical education) and including gray literature to minimize the file drawer problem. Heterogeneity should be reported with Q, Tau2, and I2; publication bias should be assessed with a funnel plot and Egger's test rather than vote counting; and robust variance estimation should address multiple outcomes per study. R is recommended as free and up to date.

Design Implications¶

Context¶

Requirements¶

• Search should be transparent, replicable, and fully cover relevant research, using multiple databases and, when possible, trained librarians.
• Inclusion and exclusion criteria should be clearly reported with detailed justification, and all activities documented.
• A strong meta-analysis would report all three heterogeneity measures (Q, Tau2, I2).

Constraints¶

• Vote counting is not recommended as it will generally lack sensitivity.
• HLM meta-analysis is appropriate only in a shared measurement context, such as a standardized test.
• Comprehensive Meta-Analysis often presents results without walking through the need for intentional decision making.

Target Learners¶

• Researchers conducting meta-analyses of PBL literature

Target Learning Goals¶

• Rigorous quantitative synthesis of problem-based learning intervention effects

Related Strategies¶

Examples¶

-

Key Sources¶

• Walker, A., & Leary, H. (2023). Conducting Problem-Based Learning Meta-Analysis: Complexities, Implications, and Best Practices. The Interdisciplinary Journal of Problem-based Learning. https://doi.org/10.14434/ijpbl.v17i2.35844