Development and Monte Carlo evaluation of meta-analytic estimators for correlated data

Techniques for estimating the mean and variance of true rho from correlations between one common measure and several other measures in the same data set (i.e., correlated data) are derived and tested using Monte Carlo simulations. Large sample simulations revealed that mean rho and rho variance estimates are essentially unbiased. Small sample size simulations revealed that the estimators are useful for many theoretically and practically important problems.