NCJ Number
158709
Journal
Evaluation and The Health Professions Volume: 18 Issue: 3 Dated: (September 1995) Pages: 315-335
Date Published
1995
Length
21 pages
Annotation
This essay provides some perspectives on the possible uses of meta-analyses and standards that permit the evaluation of the worth of any particular meta-analysis.
Abstract
"Meta-analysis" is the term used to describe a variety of research procedures that aggregate and quantitatively summarize literature findings. The technique summarizes a set of empirical findings and tests that distribution of findings for sampling error as an explanation for inconsistency. This essay argues for an understanding of meta-analysis as a procedure that resolves apparent discrepancies in the research literature. Sometimes the resolution compares the efficacy of procedures for handling a particular problem. Sometimes the meta-analysis addresses underlying theoretical models and assumptions that are made in outlining the basis of procedures and approaches. As with primary research, the reviewer must make a decision, and the user must evaluate those choices. This essay explores these choices and some of the potential consequences of the choices. In so doing, it discusses four possible approaches to meta-analysis: as a comprehensive descriptive method, as a narrow descriptive method, as a narrow inferential method, and as a comprehensive inferential method. Other issues discussed are meta-analysis as theory driven, explicit literature search methods, the explicit reporting of statistical procedures, and interpretation guidelines. 55 references