MMPI clinical scale correlates.

Abstract

This article presents and compares the results of seven independent studies that have attempted to empirically identify the behavioral or symptom correlates of individual Minnesota Multiphasic Personality Inventory clinical scales for psychiatric patients. Symptom correlates, in general, tend to provide construct validity for the "traditional" interpretation of individual scales. The correlations are uniformly quite low, however, and individual scales evidence relatively few instances of unique discrimination. Major symptom correlates for individual scales also tend to be significantly related to two-point profile codes that include the individual scales, although there are obvious differences that may be associated with scale profiles. Some possible interactions of scale correlates with race, sex, and clinical setting are noted.

Empirical validation of the Marital Satisfaction Inventory in a nonclinical sample.

“…Hase & Goldberg, 1967) and result in part from relating heterogeneous scales to highly specific criteria as well as the limited variability in criterion ratings along a 3-point scale. In a review of 7 independent studies investigating the external correlates of individual Minnesota Multiphasic Personality Inventory (MMPI) scales, Hedlund (1977) noted that observed relationships seldom accounted for as much as 10% of the variance in criterion measures. In addition to their relative increase in absolute magnitude, results of the present study reflect a generalizability across sex for 203 of 269 correlates (75%), in contrast with 73 of 122 correlates (60%) identified in the clinical validation study.…”

Establishing Behavioral Correlates: The MMPI as a Case Study

“…Based on Hedlund's (1977) findings, it is hypothesized that the relationship between a codetype and a behavior is weak, e.g., ~ _ .20. In order that the two techniques control equally for Type I errors, a is set at .01 for the whole-sample technique and at .10 for each of the subsamples for the split-sample approach, assuming independence between the subsamples.…”

Racial bias in the MMPI: A methodological review.

“…We found six published studies that purported to test the accuracy of inferences drawn from the MMPI in racial subgroups. Gynther, Altman, and Sletten (1973) and Hedlund (1977) asserted that high scores on the MMPI scales mean something different for blacks and whites, but they offered insufficient evidence for that claim. Both of these reports failed to test the significance of the difference between black scale correlates and white scale correlates (American Psychological Association et al, 1974;Humphreys, 1973) and consequently cannot conclude that the racial subgroups differed significantly in their respective scale correlates.…”

“…In summary, one study (Davis & Jones, 1974) provided indirect evidence that the MMPI is not biased when used to differentiate between schizophrenics and those with other psychiatric disorders; one study (Persons & Marks, 1971) provided direct evidence that the 4-3 code type is not biased when used to predict violent behavior ex post facto; and one study (see reanalysis by Shore, 1976, of Strauss et al, 1974 found no evidence for racial bias in distinguishing between psychotics and those with behavior disorders. The remaining investigators (Cowan et al, 1975;Gynther et al, 1973;Hedlund, 1977) claimed to have found evidence for racial bias in the MMPI, but in fact misanalyzed their data.…”