Specifying Inclusion Criteria for a Meta-Analysis: Lessons and Illustrations From a Quantitative Synthesis of Crime Reduction Experiments

• NCJ Number: 155948
• Author(s): A J Petrosino
• Date Published: January 1995
• Length: 20 pages
• Agencies: NIJ-Sponsored

Annotation

This article presents problems and illustrations from the first phase of an ongoing meta-analysis of crime-reduction programs.

Abstract

Meta-analysis, a research method referred to as the quantitative analysis of studies, has been used extensively in social science. Along with the proliferation of meta-analysis as a research methodology, there is some evidence that it may be influential with policy and program decisionmakers. Practitioners look for more conclusive evidence before taking action and find a meta-analysis of many studies bearing on the issue more persuasive than a single one. Following a discussion of the importance of defining the inclusion criteria for the studies included in the meta-analysis, the eight study inclusions criteria for this study are listed. These research criteria distinguish this study from nearly all prior meta-analyses of intervention efficacy in criminal justice. Recommendations are that the specification of inclusion criteria be given careful consideration before the document-retrieval phase; inclusion criteria may need to be redefined when "gray-area" studies are found; and before specifying inclusion criteria, decide whether it may be better to include "gray-area" studies in a future meta- analysis. 5 notes and 64 references