Equivalence of the MMPI-2 with the MMPI in psychiatric patients.

Abstract

Psychiatric patients were administered the Minnesota Multiphasic Personality Inventory (MMPI), its revision (MMPI-2), or both, in a counterbalanced repeated-measures design. MMPI-2 T scores were found to be significantly lower than MMPI T scores on several of the clinical scales. Subject rank order on T scores and dispersion of the basic clinical scales did not differ between the tests, and measures of profile similarity indicated congruence between the two instruments. Among subjects who completed both the MMPI and the MMPI-2, code-type concordance was not significantly lower than stability rates of the tests. The results support the assignment of 65T as the lower boundary of clinical elevation on the MMPI-2 and the psychometric equivalence of the MMPI-2 and the MMPI with respect to mean r scores, score rankings, and measures of score distribution.

MMPI-2 Profile Code Types and Measurement Error

“…Rosen (1953) used the card form of the MMPI. Three samples did not include retest data on Scale 5 (Mf; Harrell et al, 1992), and one sample did not contain retest data on Scale 0 (Si) (Lichtenstein & Bryan, 1966).…”

“…These calculations were based on nine samples of retest coefficients from seven studies: Eichman (1973, as cited in Dahlstrom, Welsh, & Dahlstrom, 1975, p. 258; 1-to 2-week interval), Jurjevich (1966a, as cited in Dahlstrom et al, 1975, p. 258; 1-to 2-week interval), Lichtenstein and Bryan (1966; 2.2-day interval), Newmark (1971; within 24 hr), Newmark (1973, as cited in Dahlstrom et al, 1975; 1-day interval), Rosen (1953;4.1-day interval) and Harrell et al (1992; 7.9 days). We note a few cautions concerning the correlations used in developing the weighted estimates for this study.…”

“…Of the 42 comparisons of correlations made across the basic validity, clinical, and eight supplemental scales retained on the MMPI-2, only two significantly different correlations between groups were obtained on the F scale and Es scale for women. Of the 42 comparisons of correlations made in a similarly designed study with college students, Chojnacki and Walsh (1992) found two significantly different correlations for men on Scale 8 (Sc) and Scale R. Harrell et al (1992) administered the MMPI and MMPI-2 in a counterbalanced design to psychiatric patients at a 7.9-day retest interval. Harrell et al interpreted their results as indicating the psychometric equivalence of the MMPI-2 and the MMPI with respect to score distribution, rank order of individuals, and within-subjects measures of profile shape and scatter.…”

“…In the second simulation, the clinical means and standard deviations of the Caldwell clinical data set were used in the calculations along with estimates of short-term clinical scale retest stability for psychiatric patients retested with the MMPI or MMPI-2 at short-term intervals. Because only one investigation could be located that included the study of the short-term retest stability of the MMPI-2 with a small sample of psychiatric patients (n = 31; Harrell et al, 1992), estimates of clinical scale stability for psychiatric patients retested at very short intervals with the MMPI and/or MMPI-2 were developed for this study. These estimates were based on six studies that had considered the stability of the MMPI clinical scales for psychiatric patients retested over very short intervals with the MMPI and one investigation (Harrell et al, 1992) of the repeat administration of MMPI, MMPI-2, and the cross-administration of both tests with psychiatric patients.…”

“…For mental health patients tested on two separate occasions with an average interval between testing of 688 days, Munley (2002) found scale high-point agreement of 38.60%, high 2-point agreement of 16.67%, and high 3-point agreement of 19.30%, with high-point agreement of subsets of participants with well-defined high points, 2 points, and 3 points of 41.07%, 27.50%, and 25.93%, respectively. Harrell, Honaker, and Parnell (1992) conducted a study of the equivalence of the MMPI-2 with the MMPI in a sample of 131 psychiatric patients with an average 7.9-day test-retest interval. Among patients who took the MMPI twice (n = 31), unrestricted high point and high 2-point agreement was 67.7% and 41.9%, respectively.…”