Reliability Generalization Meta-Analysis

Abstract

A reliability generalization meta-analysis (RG MA) is a suitable method to statistically integrate the reliability estimates obtained in different applications of a test. RG MA allows researchers to characterize the average reliability of scores obtained by a test across multiple studies and situations and estimate the degree of variability in reliability coefficients across different types of measures, samples, and contexts. Therefore, its results enable us to offer guidelines to applied researchers and practitioners about which scales are more reliable for assessing a constructe and in what circumstances. Thus, applied researchers and practitioners need to know what RG MAs are, how they are done and, most importantly, how we can carry out critical appraisal of RG MAs. The purpose of this article is to present RG MAs, together with some guidelines to warrant a critical reading of them. For this, a recent reliability generalization metaanalysis is used to illustrate the guidelines proposed. Finally, some concluding remarks are presented.