Premium
Best‐practice recommendations for estimating interaction effects using meta‐analysis
Author(s) -
Aguinis Herman,
Gottfredson Ryan K.,
Wright Thomas A.
Publication year - 2011
Publication title -
journal of organizational behavior
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.938
H-Index - 177
eISSN - 1099-1379
pISSN - 0894-3796
DOI - 10.1002/job.719
Subject(s) - moderation , meta analysis , psychology , narrative , audience measurement , interaction , meta regression , value (mathematics) , management science , social psychology , computer science , political science , economics , medicine , linguistics , philosophy , machine learning , law
One of the key advantages of meta‐analysis (i.e., a quantitative literature review) over a narrative literature review is that it allows for formal tests of interaction effects—namely, whether the relationship between two variables is contingent upon the value of another (moderator) variable. Interaction effects play a central role in organizational science research because they highlight boundary conditions of a theory: Conditions under which relationships change in strength and/or direction. This article describes procedures for estimating interaction effects using meta‐analysis, distills the technical literature for a general readership of organizational science researchers, and includes specific best‐practice recommendations regarding actions researchers can take before and after data collection to improve the accuracy of substantive conclusions regarding interaction effects investigated meta‐analytically. Copyright © 2010 John Wiley & Sons, Ltd.