Full Text
meta-analysis
CRAIG JOHNSON and BRIAN MULLEN
Extract
The statistical integration of the results of independent studies in a specific area of research. In a sense, meta-analysis involves methods of statistical analysis where the units of analysis are the results of independent studies rather than the responses of individual subjects. Like traditional narrative reviews of research, meta-analysis is based upon the assumption that the summary and integration of the results of previous studies can further our understanding of a phenomenon beyond the level of understanding achieved by any single study. However, meta-analysis differs from traditional narrative review in terms of its concern with precision. The specific strategies and techniques for taking the statistical results of independent studies, converting these results to common metrics, and then integrating these statistical results are designed to be rigorously precise. In addition, meta-analysis aspires to be more objective than traditional narrative review. The rules and standards for including studies in the review process, abstracting results from them, and weighing them in the final integration are never made explicit in the traditional narrative review, whereas these rules and standards must be made explicit in any meta-analytic review. Finally, meta-analyses are characterized by a higher degree of replicability than traditional narrative reviews. Unlike the divergence of ... log in or subscribe to read full text
Log In
You are not currently logged-in to Blackwell Reference Online
If your institution has a subscription, you can log in here: