Testing theory predictions and intervention mechanisms of action in sport and exercise psychology using meta-analysis.

Abstract

Testing theoretical predictions in sport and exercise psychology typically involves simultaneous tests of networks of proposed effects among psychological constructs and outcomes. One approach that holds considerable promise in advancing theory in this and other fields is the application of analytic techniques combining meta-analysis and multivariate analyses such as structural equation modeling to test theoretical predictions in synthesized data from multiple studies. In this review, we provide an overview of this approach to theory testing in research in sport and exercise psychology. We provide a nontechnical overview of its basic principles and outline the advantages it affords, including: leveraging of the large sample size from multiple data sources to provide robust averaged point and variability estimates of theory effects with high statistical power; capacity to provide overall evaluations of theoretical models across available research and guideline ranges for the expected size of theory effects in subsequent tests in primary studies; ability to test mechanistic effects, namely, mediation effects and effects of study-level and, in some cases, sample-level moderator variables on theory effects; and capacity to test predictions of both current and novel theories. We outline forms of meta-analytic structural equation modeling procedures and methods available to test theory predictions and provide a series of examples that illustrate their application to test theories in sport and exercise psychology. We conclude by summarizing some of the limitations of this approach and recommend some future innovations to, and applications of, the approach that are expected advance cumulative knowledge of theory in the field. (PsycInfo Database Record (c) 2025 APA, all rights reserved)