Estimating the Expected Replicability of a Pattern of Correlations and Other Measures of Association

Replication is at the heart of all empirical sciences. However, there are no standard procedures for establishing the replicability of a pattern of correlations found linking a particular variable to an inventory or battery of other measures. This article introduces 2 statistics for quantifying the expected replicability of a pattern of associations (i.e., correlations, slope coefficients) between a variable of interest and a SET of other variables, items, measures, and so on. Using simulations and real data, we illustrate that these statistics are highly accurate estimates of the expected replicability of an observed pattern of correlations. These statistics can readily be used to indicate the replicability of patterns of association indexed by other statistics (e.g., regression slopes or covariances) and can be applied to other contexts, such as estimating the reliability of profile correlations. It is recommended that these statistics are regularly reported in such studies.