GREENE'S VALIDITY TECHNIQUES AND THE K-CORRECTION IN THE MMPI (CARELESSNESS SCALE, TEST-RETEST INDEX, SUBTLE AND OBVIOUS ITEMS)
Abstract
Two hundred males were divided into four equal groups. The groups consisted of college students who had sought counseling, college students, ROTC students and clients at a rehabilitation center. Each individual was administered the Group Booklet Form of the MMPI. The first purpose of this study was to determine to what extent, if any, the use of Greene's validity techniques would further eliminate profiles of questionable validity when added to the use of the traditional validity techniques across the four groups. A Chi-Square revealed Greene's validity techniques to detect more invalid profiles (p < .01) than when just the Cannot Say (?) scale was utilized. No significant differences were found between Greene's techniques and the L Scale. However, the latter results were inconclusive due to 33.3% of the cells having fewer than five frequencies. The second purpose was to determine to what extent, if any, a significant difference could be found on the five scales (1,4,7,8,9) which are externally corrected for K and Scale 3 which is internally corrected for K when scored both with and without the K-correction. Data were analyzed using a 4 x 2 ANOVA for each scale. For each scale, significant within (scoring procedure) effects were found. Significant between groups effect were found for each scale with the exception of Scale 9. A post hoc analysis utilizing the Student-Newman-Keuls procedure showed significant differences on educational level between the rehabilitation group and the other groups. Significant differences were also found between groups with age as the dependent variable. An analysis of covariance with age and educational level as separate covariates also yielded significant between group effects. It would appear that in addition to the Lie (L) Scale, the use of Greene's validity techniques are useful in the detection of profiles which are invalid (answered both inconsistently and inaccurately). Since significant differences were found for each scale corrected for K, either externally or internally, the routine use of the K-correction also comes into question.
This paper has been withdrawn.