The Effect of Skew on the Magnitude of Product-Moment Correlations

The effect of transforming actual skewed variables toward symmetry on the intercorrelations among the transformed variables was examined. The results showed that average correlations increased in all 20 cases studied, even though the transformations to correct skew produced slightly greater nonlinear relationships in some cases. Because the F and t tests for the significance of r are directly proportional to the size of r, correcting skew increases power and at the same time better meets underlying assumptions of those tests. The implications of the findings as well as potential difficulties associated with correction of skew in correlational research are discussed.