The Effect of Defensive Responding by Substance-Abusing Patients on the Millon Clinical Multiaxial Inventory

Abstract

This study explored the effect of defensive (i. e., fake-good) responding by substance-abusing patients on the scale scores of the Millon Clinical Multiaxial Inventory (MCMI-II; Millon, 1987). Patients asked to respond honestly (N = 62) had significantly higher scores on most of the scales than patients who were instructed to respond defensively (N = 62) and forensic subjects suspected of abusing psychoactive substances (N = 54). Significantly fewer subjects in the defensive responding simulation and the forensic group had elevated the Drug Dependence and Alcohol Dependence scales compared to the honestly responding patients. These results indicate that most drug-abusing individuals can conceal the presence of a substance-related disorder, as reflected by scale scores on the MCMI-II, if motivated to do so.

The detection of fake-bad and fake-good responding on the Millon Clinical Multiaxial Inventory III.

“…Despite these significant group differences, 52% of substance abusers produced valid Modifying Indices and successfully avoided detection of drug or alcohol problems by Scales B (Alcohol Dependence) and T (Drug Dependence) when instructed to fake good. Fals-Stewart (1995) also found significant group differences, but the Modifying Indices mean scores were not clinically elevated. Using a cutoff base-rate (BR) score of ≥75 on Scale Y, only 19.4% of defensive responders and 20.4% of a substance abuse forensic group were identified as potentially faking good.…”

“…The strongest research findings of faking on the MCMI–II are those derived from studies of specific populations. Faking good (defensive responding) has been researched among substance abusers (Bishop, 1993; Craig, Kuncel, & Olson, 1994; Fals-Stewart, 1995). Faking bad (malingering) has been researched by one study evaluating the MCMI–II with a sample of identified malingerers seeking compensation benefits (Lees-Haley, 1992).…”

“…In summary, relevant MCMI–II dissimulation research is limited by student samples (Bagby et al, 1990, 1991; Millon, 1987; Retzlaff et al, 1991) and the lack of usable cutoffs or decision rules. Although research that uses substance abuse samples (Craig et al, 1994; Fals-Stewart, 1995) and a malingering pseudo-PTSD sample (Lees-Haley, 1992) has been completed, further clinical samples are necessary because this is the intended population of the MCMI–III. Although regression equations resulting from discriminant analyses are valuable in establishing classification rates, the most practical result of such research would be the establishment of proposed single or multiple cutoff points on the Modifying Indices that could be used by clinicians for the classification of fake-bad and fake-good response sets.…”

Distinguishing between Neuropsychological Malingering and Exaggerated Psychiatric Symptoms in a Neuropsychological Setting

“…The items that make up this index generally reflect neutral traits, and scores on this scale are an indication of the degree to which the examinee is open and revealing in their responses (Choca, 2004). High scores on Scale X reflect greater endorsement of items and potential exaggeration of symptomatology (Retzlaff, Sheehan, & Fiel, 1991;van Gorp & Meyer, 1986), while low scores reflect underreporting of symptomatology and possibly defensive responding (Bagby, Gillis, Toner, & Goldberg, 1991;Daubert & Metzler, 2000;Fals-Stewart, 1995). The utility of this scale for the detection of ''fake-good'' response sets, however, is questionable (Retzlaff et al, 1991;van Gorp & Meyer, 1986).…”

The ability of individuals with psychoactive substance use disorders to escape detection by the Personality Assessment Inventory.

“…It is thus assumed that these subscales are less susceptible to positive dissimulation than more direct and face valid measures. Nonetheless, substance-abusing patients who respond defensively to these inventories frequently obtain significantly lower scores on the drug and alcohol abuse scales than patients who respond under standard instructions (Fals-Stewart, 1995;Otto, Lang, Megargee, & Rosenblatt, 1988;Svanum, McGrew, & Ehrmann, 1994). Although direct questioning of individuals about drug and alcohol misuse appears to solicit accurate information under many circumstances (Hesselbrock, Hesselbrock, Tennen, Meyer, & Workman, 1983;Sobell & Sobell, 1975), Morey (1991) conceded that marked denial or dishonesty is likely to reduce significantly the ALC and DRG scale scores of the PAI.…”