Detection of random responding on the MMPI-A

We examined random responding on the MMPI-A in 106 adolescents from the general population. Participants were asked to report on the frequency, location, and reasons for any random responses occurring during a standard administration of the MMPI-A. Relationships between self-reported random responding and validity indices (F1, F2, F, and Variable Response Inconsistency [VRIN] scale) were examined. In addition, each participant was randomly assigned to 1 of 5 groups, with each group completing an assigned portion (0, 25%, 50%, 75%, or 100%) of an MMPI-A answer sheet without access to the test booklet, and the utility of the validity scales in discriminating standard protocols from all or partially random protocols was investigated. Most adolescents acknowledged one or more random responses correlated significantly with F but not VRIN. Validity scales were sensitive to all or partially random protocols, and produced high classification rates when discriminating among groups.     

Detecting and Discriminating Between Random Responding and Overreporting on the MMPI-A

The utility of traditional validity scales in detecting random responding and overreporting on the MMPI-A and in discriminating between these two response sets was investigated in a sample of 89 clinical and nonclinical adolescents aged 14 to 17 years. The overreporting group consisted of 24 nonclinical adolescents instructed to complete the MMPI-A as if they were trying to create an impression of psychological disturbance. The random group included 20 nonclinical adolescents instructed to complete the MMPI-A answer sheet without access to the test booklet. Clinical (n = 20) and nonclinical (n = 25) standard groups completed the MMPI-A under the standard instructions. Results showed that validity scales designed to detect random responding and overreporting on the MMPI-A were quite effective in doing so. Significant differences between groups and high classification rates were noted. In addition, the VRIN scale was found to be sensitive only to random responding, whereas Scale F was sensitive to both random responding and overreporting. Scale L also was found to be sensitive to random responding, and support for the utility of the F - K index in identifying overreported adolescent profiles was obtained.     

Detecting random, partially random, and nonrandom Minnesota Multiphasic Personality Inventory-Adolescent protocols

The ability of the Minnesota Multiphasic Personality Inventory-Adolescent (MMPI-A; J. N. Butcher et al., 1992) validity scales to detect random, partially random, and nonrandom MMPI-A protocols was investigated. Investigations included the Variable Response Inconsistency scale (VRIN), F, several potentially useful new F and VRIN subscales, and formulas F-sub-2 - F-sub-1 and F + F-sub-2 + |F - F-sub-2|. Protocols completed by 150 adolescents at a juvenile court setting, screened for randomness with a matched-pair Millon Adolescent Clinical Inventory (MACI) or Jesness Inventory, were compared with 100 computer-generated, all-random protocols, and with 5 levels of partially random protocols. VRIN was the most effective scale in detecting all-random protocols; however, the optimum cutoff of >or= 75 failed to identify 1/3 of them. Using the new scales, a decision algorithm was described that correctly classified 94%-95% of protocols as interpretable, partially interpretable, or uninterpretable.     

Identification of random responding on the MMPI-A

Although substantial research literature on the effects of random responding on the MMPI-2 exists, there is very limited data available on this issue with the MMPI-A. The purpose of this study was to evaluate the utility of selected MMPI-A validity scales in detecting differences in response patterns between protocols produced by 354 adolescents assessed in clinical settings and a group of 354 randomly produced MMPI-A protocols. Results indicate that MMPI-A validity and basic clinical scales differ significantly between random and clinical groups and that MMPI-A validity Scales F, F1, F2, and VRIN appear to be most useful in correctly identifying protocols from actual clinical participants versus randomly generated response patterns. Findings are discussed in terms of the dramatic effects of the sample base rate for random responding on overall classification accuracy results. Furthermore, it was noted that the optimal cutting scores for MMPI-A Scales F, F1, F2, and VRIN were largely consistent with interpretive recommendations found in the test manual (Butcher et al., 1992) when the relative frequency of random response protocols to clinical protocols was evaluated at a ratio of 1:10. Finally, future recommendations for evaluation of the F1-F2 difference score and the TRIN scale are offered in terms of the most relevant research designs to evaluate these measures.     

MMPI-A validity scale uses and limitations in detecting varying levels of random responding

Although there is a substantial research literature on the effects of random responding on the MMPI-2 (Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989), there are very few studies available on this topic with the MMPI-A (Butcher et al., 1992). Archer and Elkins (1999) found that MMPI-A validity scales F and VRIN were particularly useful in detecting entirely random profiles from those derived standardly in clinical settings but noted that "all random" protocols could not be used to evaluate the usefulness of the T-score difference between the first half (F1) and the second half (F2) of the MMPI-A test booklet. Following up on this issue, this study extended the methodology of previous research by examining the hit rate, positive predictive power, negative predictive power, sensitivity, and specificity of VRIN, F, F1, F2 and the absolute value of the T-score difference between F1 and F2 (denoted as IF1-F21) in 5 samples varying in the degree of protocol randomness. One of the samples consisted of 100 adolescent inpatients administered the MMPI-A under standard instructions, and another sample consisted of 100 protocols randomly generated by computer. The additional 3 samples of 100 protocols each contained varying degrees of computer-generated randomness introduced in the latter half of the MMPI-A item pool. Over- all, the results generally indicate that several MMPI-A validity scales are useful in detecting protocols that are largely random, but all of these validity scales are more limited in detecting partially random responding that involves less than half the total item pool located in the second half of the test booklet. Clinicians should be particularly cautious concerning validity inferences based on the observed T-score difference that occurs for the F1 and F2 subscales and current findings do not support the clinical usefulness of this index.     

Detecting random, partially random, and nonrandom Minnesota Multiphasic Personality Inventory-2 protocols

The ability of the Minnesota Multiphasic Personality Inventory-2 (MMPI-2; J. N. Butcher et al., 2001) validity scales to detect random, partially random, and nonrandom MMPI-2 protocols was investigated. Investigations included the Variable Response Inconsistency scale (VRIN), F, several potentially useful new F and VRIN subscales, and Fb- F>or=30 and Fb>or=90. Protocols completed by 150 adults participating in custody evaluations at a juvenile court setting, screened for randomness with a matched-pair Millon Clinical Multiaxial Inventory-III (T. Millon, R. Davis, & C. Millon, 1997), were compared with 500 computer-generated all-random protocols and with three levels of partially random protocols. VRIN was the most effective scale in detecting uninterpretable random protocols; however, VRIN >or= 80 failed to identify 37% of them. Fb- F>or=30 and Fb>or=90 misidentified 41% of the 50%-65% random protocols as partially interpretable. Using the new scales, a decision algorithm was described that correctly classified 97%-100% of the protocols as interpretable, partially interpretable, or uninterpretable.     

Effective cutoffs for detecting random, partially random, and nonrandom 370-item MMPI-2 short form protocols

The ability of the 370-item short form MMPI-2 (Butcher, Graham, Ben-Porath, Tellegen, Dahlstrom, & Kaemmer, 2001) validity scales to detect random protocols was investigated using samples of 500 nonrandom protocols, 250 half-random protocols, and 250 all-random protocols. The long-form cutoff of VRIN ≥ 80 was unable to detect protocols with either level of randomness. The long-form cutoffs of Fp ≥ 100 or F ≥ 100 were able to detect all-random but not half-random protocols. Alternative cutoffs for VRIN, Fp, and F were investigated and short-form subscales of those scales were developed to improve detection of partially random protocols. An algorithm using alternative cutoffs for the scales and the new subscales was highly effective, detecting almost all of the random protocols as well as the nonrandom protocols. A follow-up cross-validation study was conducted that confirmed the effectiveness of the algorithm.     

Normal range MMPI-A profiles among psychiatric inpatients

The present study examined the base rates of normal range Minnesota Multiphasic Personality Inventory-Adolescent (MMPI-A) profiles in an inpatient sample and examined the differences between adolescents with apparently valid normal range profiles (all clinical scale T-scores < 60) and those with elevated profiles on prior interventions, reported internalizing and externalizing symptoms, and MMPI-A validity scale scores and other indexes of underreporting. Normal range profiles cannot be adequately explained by a less pathological history prior to hospitalization or by defensiveness. Thirty percent of male and 25% of female adolescents produced valid MMPI-A profiles in which none of the clinical scales were elevated. Both male and female adolescents with normal range profiles were generally less likely to report internalizing symptoms than those with elevated profiles, but both groups report externalizing symptoms. Neither the standard MMPI-A validity scales nor additional validity scales discriminated between profile groups. Clinicians should not assume that normal range profiles indicate an absence of problems.     

An evaluation of the MMPI-2 and MMPI-a true response inconsistency (TRIN) scales

The Minnesota Multiphasic Personality Inventory-Adolescent (MMPI-A) and Minnesota Multiphasic Personality Inventory-2 (MMPI-2) True Response Inconsistency (TRIN) scales are measures of acquiescence and non-acquiescence included among the standard validity scales on these instruments. The goals of this study were to evaluate the effectiveness of these scales in detecting varying degrees of acquiescence and non-acquiescence and to evaluate cutoff scores for clinical use. After the removal of invalid protocols from the MMPI-2 and MMPI-A normative samples, each normative sample was randomly divided in half. For each measure, one half of the normative sample served as a comparison group and the other half was modified with increasing degrees (10%, 20%, 30%, 40%, and 50%) of randomly inserted true or false responses. The results for a 9.1% base rate of acquiescence or non-acquiescence provide support for TRIN cutoff scores at or near those presented in the MMPI-A and MMPI-2 manuals.     

MMPI-A performance of African and Mexican American adolescent first-time offenders

54 African and Mexican American adolescent first-time offenders were examined with the MMPI-A to evaluate ethnic differences. Multivariate analyses by ethnicity and MMPI-A scales (validity, clinical, content, and supplementary scales) were not significant; however, there was a significant univariate difference where African American adolescents scored significantly higher on the Repression scale than the Mexican American group. A greater percentage of within-normal-limits profiles were African American (50%) than Mexican American (25%). Research and clinical implications for using the MMPI-A with these groups are discussed.     

MMPI-A characteristics of male adolescents in juvenile justice and clinical treatment settings

The current study investigates the extent to which the Minnesota Multiphasic Personality Inventory-Adolescent (MMPI-A) profiles of 196 male adolescents evaluated in a South Carolina detention center could be successfully discriminated from the protocols of 200 male adolescent psychiatric inpatients in three states and 151 dually diagnosed male adolescents. Results showed significant differences in mean T-score values among these three groups of adolescents across a variety of MMPI-A scales and subscales. Results from discriminant function analyses indicate that treatment setting can be predicted effectively from MMPI-A profiles. Beyond the MMPI-A profile differences established for adolescents in these groups, the similarities of adolescents were also noted, particularly in terms of the frequent occurrence of Within-Normal-Limits profiles for adolescents in all groups. Potential directions for future research with the MMPI-A are presented.     

Incremental validity of the MMPI-A content scales in the prediction of self-reported symptoms

The incremental contribution of the MMPI-A (Butcher et al., 1992) content scales to the prediction of scores on self-report measures of psychopathology was examined in a sample of 62 adolescents in inpatient treatment and 59 adolescents from the community. All participants completed the MMPI-A and a battery of criterion measures. A series of hierarchical regression analyses was conducted in which the MMPI-A clinical and content scales served as the independent variables and the criterion measures as the dependent variables. The content scales were found to have incremental validity beyond the clinical scales in predicting variance in the criterion measures. Similarly, the clinical scales also demonstrated incremental validity over the content scales in making these predictions. Both sets of scales made independent contributions to the prediction of sample membership (clinical vs. nonclinical). Findings suggest that both the clinical and content scales of the MMPI-A make significant contributions to the assessment of adolescents' psychological functioning.     

Adolescent personality and two measures of response inconsistency

This study examined the relationship between the Response Inconsistency Scale (RINC) of the Adolescent Personality Questionnaire and the Variable Response Inconsistency Scale (VRIN) of the Minnesota Multiphasic Personality Inventory for Adolescents (MMPI-A) in a group of 62 participants. RINC and VRIN were positively correlated .30 (p < .01). Additionally, both scales were correlated with five broad personality scales available from the Adolescent Personality Questionnaire. Scores on both inconsistency scales correlated negatively with those on Extraversion, Independence, and Self-control (but the latter two not significantly).     

Detecting fake-bad MMPI-A profiles

This study examined the extent to which the validity scales of the Minnesota Multiphasic Personality Inventory-Adolescent (MMPI-A; Butcher et al., 1992) could identify adolescents who were faking-bad. This study also examined the extent to which the validity scales could differentiate between nonclinical adolescents instructed to fake-bad and both clinical and nonclinical adolescents who received standard instructions. Participants were 58 male and 80 female high school students and 58 male and 80 female adolescents from clinical settings. Results indicated that the mean profiles resembled those previously reported for adults on the MMPI (Graham, 1987) and MMPI-2 (Graham, Watts, & Timbrook, 1991) and for adolescents on the original MMPI (Archer, Gordon, & Kirchner, 1987). Accurate identification of students who were faking-bad was achieved. It was possible to differentiate between the clinical adolescents and nonclinical adolescents who were faking, and it was possible to differentiate between nonclinical adolescents who were faking and nonclinical adolescents with standard instructions; however, different cutoff scores were needed for these two discriminations. Optimal cutoff scores were presented.     

Comparing three new MMPI-2 randomness indices in a novel procedure for random profile derivation

This study compared three new validity indices; /F-Fb/, VRIN+/F-Fb/, and F+Fb+/F-Fb/; to the F, back-F, and VRIN according to their effectiveness in detecting degrees of profile randomness. Participants completed the Minnesota Multiphasic Personality Inventory-2 (MMPI-2; Butcher, Dahlstrom, Graham, Tellegen, and Kaemmer, 1989) according to standardized instructions. All profiles starting at Items 142, 284, or 426 were substituted with computer-generated random responses to produce 75%, 50%, and 25% random profiles, respectively. Twenty-five profiles were derived using 100% computer-generated items. Finally, one unaltered set of profiles was designated as 0% random. This generated five groups of MMPI-2 profiles, differing according to the degree of profile randomness (i.e., 0%, 25%, 50%, 75%, and 100%). Results showed that the F and /F-Fb/ indices were unable to distinguish authentic from all degrees of randomness, whereas the F+Fb+F-Fb scale could reliably distinguish all levels of randomness. The actuarial analysis found back-F and F+Fb+/F-Fb/ misclassified the least number of profiles, suggesting the latter index should be included in the evaluation of MMPI-2 profile randomness.     

Ability of substance abusers to escape detection on the Minnesota Multiphasic Personality Inventory-Adolescent (MMPI-A) in a juvenile correctional facility

The ability of respondents to underreport successfully on substance abuse and validity scales of the Minnesota Multiphasic Personality Inventory-Adolescent (MMPI-A; Butcher et al., 1992) was evaluated. Incarcerated teens (67 substance abusing, 59 non-substance abusing) completed the MMPI-A twice: once under standard instructions (SI) and once under instructions to fake good (FG). Under SI, substance scales correctly classified about 60% to 85% of adolescents. Under FG, substance- and non-substance-abusing juveniles produced lower scores on substance scales. However, the Lie Scale (L) was able to detect more than 75% of deceptive profiles and about 77% of honest profiles. When scale L and the best substance scale were used in combination, only about 18% of faking substance abusers were not identified as either substance abusers or as underreporting. For feigning substance abusers, only about 10% of substance abusers were detected, with about 72% being categorized as faking and needing further assessment.     

Comparing Three New MMPI-2 Randomness Indices in a Novel Procedure for Random Profile Derivation

This study compared three new validity indices; |F-Fb|, VRIN+|F-Fb|, and F+Fb+|F-Fb|; to the F, back-F, and VRIN according to their effectiveness in detecting degrees of profile randomness. Participants completed the Minnesota Multiphasic Personality Inventory-2 (MMPI-2; Butcher, Dahlstrom, Graham, Tellegen, and Kaemmer, 1989) according to standardized instructions. All profiles starting at Items 142, 284, or 426 were substituted with computer-generated random responses to produce 75%, 50%, and 25% random profiles, respectively. Twenty-five profiles were derived using 100% computer-generated items. Finally, one unaltered set of profiles was designated as 0% random. This generated five groups of MMPI-2 profiles, differing according to the degree of profile randomness (i.e., 0%, 25%, 50%, 75%, and 100%). Results showed that the F and |F-Fb| indices were unable to distinguish authentic from all degrees of randomness, whereas the F+Fb+F-Fb scale could reliably distinguish all levels of randomness. The actuarial analysis found back-F and F+Fb+|F-Fb| misclassified the least number of profiles, suggesting the latter index should be included in the evaluation of MMPI-2 profile randomness.     

Incremental validity of selected MMPI-A content scales in an inpatient setting

To date, relatively few studies have been published evaluating the validity or incremental validity of the content scales from the adolescent version of the Minnesota Multiphasic Personality Inventory (MMPI-A; J. N. Butcher et al., 1992). A sample of 629 psychiatric inpatient adolescents who had completed the MMPI-A was used to evaluate the ability of selected clinical and content scales to predict conceptually related clinical variables. Criteria were based on clinician ratings, admission and discharge diagnoses, and chart reviews. Results from hierarchical multiple and logistic regression analyses indicated the content scales offered incremental validity over the clinical scales and supported the use of the content scales as an adjunct to the traditional clinical scales.     

Evaluation of an MMPI--a short form: implications for adaptive testing

Efforts to develop a viable short form of the MMPI (Hathaway & McKinley, 1943) span more than 50 years, with more recent attempts to significantly shorten the item pool focused on the use of adaptive computerized test administration. In this article, we report some psychometric properties of an MMPI-Adolescent version (MMPI-A; Butcher et al., 1992) short form based on administration of the first 150 items of this test instrument. We report results for both the MMPI-A normative sample of 1,620 adolescents and a clinical sample of 565 adolescents in a variety of treatment settings. We summarize results for the MMPI-A basic scales in terms of Pearson product-moment correlations generated between full administration and short-form administration formats and mean Tscore elevations for the basic scales generated by each approach. In this investigation, we also examined single-scale and 2-point congruences found for the MMPI-A basic clinical scales as derived from standard and short-form administrations. We present the relative strengths and weaknesses of the MMPI-A short form and discuss the findings in terms of implications for attempts to shorten the item pool through the use of computerized adaptive assessment approaches.