A cross-validation of the keane and penk MMPI scales as measures of post-traumatic stress disorder

Abstract

The scores of DSM‐III‐diagnosed post‐traumatic stress disorder (PTSD) patients (N = 116), psychiatric patients who did not meet the criteria, and normals (N = 19) on the Keane, Malloy, and Fairbank (1984) PTSD and Penk Combat scales for the MMPI were compared. The Keane scale discriminated PTSD‐positive patients from normals at a substantial level of accuracy (ω2 = .23; mean hit rate = 80.50%) and PTSD‐positive from PTSD‐negative patients at a more modest level (ω2 = .09; mean hit rate = 64%). The scores of the PTSD‐positive and PTSD‐negative patients were considerably lower than those of Keane et al.'s (1984) samples, which sugggests that local norms may be needed to facilitate their interpretation. The Penk Combat Intensity scale, which was correlated highly with the Keane scale, differentiated the PTSD‐positive patients from both the normals and the PTSD‐negative patients, but with less differentiating power (ω2's = .07 and .08). The Penk Combat Exposure scale did not separate the groups.

MMPI profiles of acute and chronic PTSD in a civilian sample

“…With the optimal cutoff score set at 30, they correctly classified 82% of the total sample using this scale. In reviewing six replication studies (Cannon, Bell, Andrews, & Finkelstein, 1987;Gayton, Burchstead, & Matthews, 1986;Hyer et al, 1986; O r r , Vanderploeg, Sison, & Hickling, 1987Watson, Kucala, & Manifold, 1986), Watson (1990) found an acceptable median hit rate of .70 for the PK scale used with Vietnam veterans, using cutoff scores varying from 8 5 to 25. Some of these hit rates were inflated, however, by high specificity in contrast to low sensitivity.…”

The validation of a self-report measure of posttraumatic stress disorder: The Posttraumatic Diagnostic Scale.

“…First, because they were empirically rather than conceptually derived, their reliability and validity are especially vulnerable to the specific characteristics of any particular sample, and thus these scales require repeated examination of their psychometric properties. For example, Keane et al (1984) originally proposed a cutoff score of 30 for the presence of PTSD diagnosis; however, other researchers have found much lower cutoff scores corresponding to the diagnosis (e.g.. Query et al, 1986;Watson et al, 1986). Second, the items of the scales do not correspond to the DSM-III-R or DSM-IV symptoms, and their conceptual relationship to PTSD is ambiguous.…”

The Utility of the PAI and the MMPI-2 for Discriminating PTSD, Depression, and Social Phobia in Trauma-Exposed College Students

“…Investigators have consistently found significant differences between PTSD and non-PTSD groups on several MMPI/MMPI-2 scales, but specific conclusions have varied across studies (Elhai, Frueh, & Gold, 2000;Gayton, Burchstead, & Matthews, 1986;Hyer et al, 1986;Keane et al, 1984;Morrell & Rubin, 2001;Orr et al, 1990;Vanderploeg, Sison, & Hickling, 1987;C. G. Watson, Juba, Anderson, & Manifold, 1990; C. G. Watson, Kucala, & Manifold, 1986). One consistent and salient finding is the multidimensional presentation of PTSD on the MMPI/MMPI-2: The mean PTSD profile consistently includes clinically significant elevations on multiple scales-typically scales F, 2, and 8.…”