What does research indicate about the K correction?
The K correction does not improve individual scale validity in any setting. It significantly attenuates individual scale validity in non-clinical settings; the detrimental effects are most pronounced in settings in which defensive responding is frequent. K-corrected code types are no more valid, and are often less valid, as predictors of correlates in clinical settings. Non-K-corrected correlates (i.e., correlations with extra-test criteria) are similar in composition to K-corrected correlates, but are often stronger in magnitude. Therefore, non-K-corrected code types can be interpreted with greater confidence based on the existing literature.
