RESPONSE SETS AND PERSONALITY MEASURES: THE K SCALE OF THE MMPI (ASSESSMENT, FAKING, TESTS)

摘要

Researchers in the area of personality assessment have long recognized the problem of response distortion and its effects on scores obtained. The K scale of the Minnesota Multiphasic Personality Inventory (MMPI) was designed to correct clinical scale scores for the level of distortion due to response sets.;The purpose of this study was to examine the effects of correction for response set on the MMPI in five sample groups: (1) normal individuals; (2) chronic pain patients; (3) psychology clinic outpatients; (4) psychiatric inpatients; and (5) hospitalized individuals carrying diagnoses of schizphrenia. Issues that were unresolved based on previous research included: (1) whether K-corrections lead to interpretive change in MMPI scores or profiles; (2) whether population samples differ in terms of the degree of interpretive change that may occur; (3) whether K-corrections increase differentiation of normal and criterion maladjusted groups; and (4) whether K-corrections lead to a reduction in false-negative judgements (i.e., maladjusted individuals who appear normal on the basis of their MMPI profiles).;Results indicated that a significant proportion of profiles in all samples evidenced interpretive change due to K-corrections. The degree of interpretive change was related to population membership when the normal group was compared with any maladjusted group or combination of maladjusted groups; however, the maladjusted groups did not vary among themselves on the degree of change variable. K-corrections increased the differentiation of normal individuals and individuals with schizophrenia on the Schizophrenia scale; however, K-corrections did not significantly increase the differentiation of normal individuals and chronic pain patients on the Hypochondriasis scale. Finally, K-corrections did not significantly affect the proportion of false-negative profiles in the maladjusted samples, although a slight decrease in frequency of false-negative profiles was found.;These findings suggest that K-corrections have a greater effect than might be expected on the basis of previous studies. One reason may have been that an emphasis of this study was to examine the effects of K-corrections on individual profiles within samples rather than simply to compare group mean scores. The clinical utility of K-corrections needs further examination.