STATISTICAL METHODS FOR SUMMARIZING INDEPENDENT CORRELATIONAL RESULTS

This study discusses the description and application of certain combined tests and estimates based on k independent sets of bivariate normal correlational data. We consider the case in which only the sample correlation coefficients are available and the case in which the original paired data are available. A combined test for the correlation parameter based on Fisher's z-transformation is considered, as well as confidence intervals and a combined estimate for the correlation parameter. A test for the homogeneity assumptions is discussed and a table for the power of the additive test based on Fisher's z-transform is presented. Bayesian single and multiparameter estimation are also considered