Meta-analysis is a technique for combining the summary statistics from previously conducted research studies to indicate the direction of results and provide an index of the magnitude of effect size. This paper focuses on the effect of the violation of the assumption of independence (that the value of any included statistic is in no way predictable from the value of any other included statistic) when combining correlation coefficients to determine effect size. A Monte Carlo simulation used the following four parameters and specified values: (1) the sample size within a study (20, 50, 100); (2) the number of predictors (1, 2, 3, 5); (3) the population intercorrelated among predictors (0, .3, .7); and (4) the population correlations between predictors and criterion (0, .3, .7). Path diagrams are given for each predictor case. The means, medians, and standard deviations of the correlation coefficients and the Fisher's Z transformation of the correlation coefficients for all population correlations and population intercorrelation values for each sample size were calculated and the data is presented in table format. As all the standard deviations were very close to their expected values, it is concluded that nonindependence does not affect the estimation of either the measures of central tendency or the standard deviations when the same population parameter is being estimated. (BS)