Model choice in meta‐analysis should not be driven by Cochran's Q test

Key Points

• Evaluate the appropriateness of using Cochran's Q test for model selection in meta-analysis.
• Analyzed existing literature on meta-analysis methods and model selection criteria.
• Critically assessed Cochran's Q test's effectiveness in determining model choice.
• Argued that reliance on Cochran's Q test can mislead model selection.
• Proposed alternative criteria that may provide more consistent results.