Stability of the Million Adolescent Personality Inventory in an Incarcerated Delinquent Population

Clinical use of the Millon Adolescent Personality Inventory (MAPI) depends on computerized interpretation of the pattern of scale score elevations in the profile beyond certain cutoff scores rather than the elevations of single scales. There has been no reported work on the stability of the scale scores or the stability of the 2-point codes in a delinquent population. The MAPI was administered to 46 incarcerated male juvenile delinquents and was readministered after a period of 2 to 4 weeks. The test-retest correlations of the base-rate scale scores ranged from .33 to .89 with a mean of .74, which compare favorably to Millon's (1982) standardization sample. However, only 41% of the 2-point codes were judged to be congruent between administrations. The poor congruence of the 2-point codes across administrations raises doubts about interpretive statements based on these codes.     

MAPI code types in an inpatient crisis-unit sample

Limited data is available concerning the use of the Millon Adolescent Personality Inventory (MAPI; Millon, Green, & Meagher, 1982) and high-point code combinations of its personality style scales. Our study reports on the frequency of high-point codes in an inpatient crisis-unit sample. Data analyses comparing personality style codes using MAPI expressed concerns scale elevations and Rorschach data are reported. Results comparing the personality style codes and expressed concerns scales were consistent with the findings of Millon et al. (1982). Analyses comparing Rorschach and MAPI data revealed only the presence or absence of vista responses as having any appreciable relationship to MAPI codes or personality scales.     

MAPI Code Types in an Inpatient Crisis-Unit Sample

Limited data is available concerning the use of the Millon Adolescent Personality Inventory (MAPI; Millon, Green, & Meagher, 1982) and high-point code combinations of its personality style scales. Our study reports on the frequency of high-point codes in an inpatient crisis-unit sample. Data analyses comparing personality style codes using MAPI expressed concerns scale elevations and Rorschach data are reported. Results comparing the personality style codes and expressed concerns scales were consistent with the findings of Millon et al. (1982). Analyses comparing Rorschach and MAPI data revealed only the presence or absence of vista responses as having any appreciable relationship to MAPI codes or personality scales.     

The Utility of the MAPI in the Assessment of Depression

The relationships between the Millon Adolescent Personality inventory (MAPI; Millon, Green, & Meagher, 1982) and depression, as assessed by the Multiscore Depression Inventory (MDI; Berndt, 1968) were examined. Elevations on the MDI subscales were positively related to elevations on MAPI personality style Scale 2 (Inhibited) and Scale 8 (Sensitive) and on six of the eight MAPI expressed concerns scales. MAPI personality style Scale 4 (Sociable) and Scale 5 (Confident) were negatively correlated with the MDI. MAPI code types containing either Scale 2 or 8 were associated with a high MDI full-scale score. The subjects who received a MAPI computer-generated diagnosis of borderline personality disorder also had elevated MDI full-scale scores. The findings of this study appear consistent with the existing body of MAPI research, and the data suggest that the MAPI personality style scales may have both state and trait qualities.     

The utility of the MAPI in the assessment of depression

The relationships between the Millon Adolescent Personality Inventory (MAPI; Millon, Green, & Meagher, 1982) and depression, as assessed by the Multiscore Depression Inventory (MDI; Berndt, 1968) were examined. Elevations on the MDI subscales were positively related to elevations on MAPI personality style Scale 2 (Inhibited) and Scale 8 (Sensitive) and on six of the eight MAPI expressed concerns scales. MAPI personality style Scale 4 (Sociable) and Scale 5 (Confident) were negatively correlated with the MDI. MAPI code types containing either Scale 2 or 8 were associated with a high MDI full-scale score. The subjects who received a MAPI computer-generated diagnosis of borderline personality disorder also had elevated MDI full-scale scores. The findings of this study appear consistent with the existing body of MAPI research, and the data suggest that the MAPI personality style scales may have both state and trait qualities.     

MMPI--2 Code-Type Congruence of Injured Workers

In this study, the authors examined the stability of Minnesota Multiphasic Personality Inventory--2 (J. N. Butcher, W. G. Dahlstrom, J. R. Graham, A. Tellegen, & B. Kaemmer, 1989) code types in a sample of 94 injured workers with a mean test-retest interval of 21.3 months (SD = 14.1). Congruence rates for undefined code types were 34% for high-point codes, 22% for 2-point codes, and 22% for 3-point codes. The data provide tentative evidence suggesting that defined code types are more stable than undefined code types. Cohen's kappa, a statistic that controls for chance agreement, was calculated for each clinical scale for both 2-point and 3-point code types. Only 2 of the 20 kappa coefficients were not significant at the p = .05 level.     

MMPI-2 profile code types and measurement error

We studied simulated MMPI-2 (Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989) code type stability and change expected with measurement error for 12 MMPI-2 well-defined mean code type profiles. Profile scores for the 2 scales defining the code type were systematically varied to represent target code type profiles at 9 different levels of T-score profile definition. We randomly generated samples of 50 simulated, estimated true score profiles at each level of profile definition for each code type around the estimated true scores for each scale at each level of profile definition. Two sets of simulated profiles were developed. The first simulation was based on the reported means, test-retest reliabilities, and the standard errors of measurement for the MMPI-2 normative group. The second simulation was based on the means, standard deviations, and estimated retest stability for a clinical group of psychiatric patients. We calculated frequencies and percentages of simulated profiles with the highest estimated true scores on the same 2 scales as the original code type profile. Percentages of simulated profiles with the same 2 highest scales as the original code type profiles increased from 27% to 37% for the 3-point level of definition, 37% to 49% for the 5-point definition, 46% to 61% for 7-point definition, 63% to 78% for 10-point definition, 78% to 89% for 13-point definition, 83% to 93% for 15-point definition, and greater than 90% for profile definition greater than 15 points.     

Comparability of MMPI-2 scales and profiles over time

This study investigated Minnesota Multiphasic Personality Inventory-2 (MMPI-2; Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989) scale and profile comparablilty for MMPI-2 profiles completed on 2 separate occasions by mental health patients receiving treatment at a Veterans Affairs Medical Center (n = 114). Patients were predominantly men (96.5%), with an average age of 44.08 and an average of 12.39 years of education at the time of initial testing. MMPI-2 tests were completed on 2 separate occasions as a routine part of treatment with a mean interval between test administrations of 688 days. Findings were analyzed for the complete sample and for 3 subsamples with different test-retest intervals. MMPI-2 scale test-retest correlation coefficients for the entire sample ranged from .48 to .69 for the Basic scales, .49 to .80 for the Supplementary scales, and .56 to .78 for the Content scales with scale high-point agreement = 38.60%, high 2-point agreement = 16.67%, and high 3-point agreement = 19.30%. High-point agreement for subsets of participants with well-defined high points, 2-points and 3-points was 41.07%, 27.50%, and 25.93% respectively. Pearson r correlation coefficients for T scores across the Basic scales for pairs of profiles averaged .78, suggesting similarity of profile shape across testing occasions. MMPI-2 profiles were also examined in relation to Skinner and Jackson's 3 modal MMPI profile types.     

Measures of consistency for Holland-type codes

“Consistency” in Holland's theory refers to the extent to which more closely related scale types are found together in codes of the Self-Directed Search sort. This paper describes two new measures of consistency. One is based on the hexagonal model and is for use with 3-point codes. The other is based on conditional probabilities and is for use with 2-point codes.     

MMPI-2 Profiles of Male Pre-Trial Defendants

Since its introduction, the MMPI-2 has been used as an equivalent to the MMPI. However, the clinical correlates of MMPI-2 profiles are still being investigated. This study examined the MMPI-2 profiles of 217 male pre-trial defendants. Data indicated mean scale elevations over 65 on F, 8, 6, 7, 4, 2, and 1 (in descending order); and mean content scale elevations on DEP, BIZ, HEA, and TRT. The most frequently elevated scales were 6, F, and 2. The most frequently elevated content scales were CYN, DEP, and ANX. The most common 2-point code was a 6–8/8–6 for subjects charged with both violent and nonviolent offenses. The F–K Index was greater than 11 in 22% of the subjects. Generalizability of these data and suggestions for future research are discussed.     

The Self-Perception Profile for Children (SPPC): a multiple-indicator multiple-wave analysis using LISREL

The longitudinal psychometric properties of SPPC were assessed using a sample of 157 schoolchildren across four time periods using structural equation modeling. The theoretically derived multidimensional structure of SPPC (Harter, 1985) and its associated reliabilities were found to be fairly consistent overtime. The stability of the factor structure, however, did vary across repeated administrations: the stability of the second-order self-perception factor was relatively low (0.629) between first and second administrations, increasing between times two and three (0.812), while reaching the highest point (0.941) across times three and four. This result could be attributable to either a growing familiarization with scale items, the response format, or natural respondent maturation. While the reason underlying temporal instability is uncertain, the results do suggest that caution should be exercised when interpreting SPPC scores particularly in intervention designs with comparatively long intervening periods.     

Confidence intervals for the MMPI-2

The confidence intervals for the Minnesota Multiphasic Personality Inventory (MMPI-2) clinical scales were investigated. Based on the clinical scale reliabilities published in the MMPI-2 manual, estimated true scores, standard errors of measurement for estimated true scores, and 95% confidence intervals centered around estimated true scores were calculated at 5-point MMPI-2 T-score intervals. The relationships between obtained T-scores, estimated true T-scores, scale reliabilities, and confidence intervals are discussed. The possible role of error measurement in defining scale high point and code types is noted.     

The Kuwait University Anxiety Scale: psychometric properties

This study was undertaken to construct the Kuwait University Anxiety Scale in two comparable Arabic and English versions. The intent was to provide a research tool as well as a measure of predisposition to develop anxiety in general among adults and adolescents. The scale is comprised by 20 brief statements answered on a 4-point intensity scale, anchored by 1: Rarely and 4: Always. Factor analysis yielded three high-loaded factors of Cognitive/Affective, Behavioral/Subjective, and Somatic Anxiety, with moderate interfactor correlations. Item-remainder correlations ranged from .27 to .74. Reliabilities ranged from .88 to .92 (alpha) and between .70 and .93 (test-retest), denoting good internal consistency and stability. Criterion-related validity of the scale ranged between .70 and .88 (5 criteria), while the loadings of the scale on a general factor of anxiety were .93 and .95 in two factor analyses, demonstrating the scale's criterion-related and factorial validity. Discriminant validity of the scale was demonstrated. The scale correlated .65 with scores on the Beck Depression scale. Male adolescents attained a higher mean score than male undergraduates. Females have significantly higher mean scores than their male counterparts. Kuwaiti norms (N = 4,660) were reported. An English version of the scale is available.     

Relationships between MMPI codetypes and MAC scale elevations in adolescent psychiatric samples.

This study investigated the relationship between subjects' MacAndrew Alcoholism Scale (MAC) elevations and their placement in various codetype classifications. The sample consisted of 1,347 adolescents receiving psychiatric services in Missouri, Virginia, or Texas. The frequency of MAC scores greater than or equal to 24 and greater than or equal to 28 was evaluated, separately by gender, for single-scale and two-point codetypes, containing a minimum of at least 10 subjects. Marked differences occurred in the frequency of MAC scale elevations depending on the adolescent's codetype classification. MAC scores for both genders, for example, were most frequently elevated for the 4-9/9-4 codetype and more rarely elevated for the 2-3/3-2 codetype. Results are discussed in terms of potential measurement overlap which may occur between the MAC scale and several MMPI standard clinical scales, including 4, 6, and 9. It was also noted that the sensitivity and specificity of the MAC scale may vary in relation to subjects' clinical scale profile characteristics.     

Three methods of developing MMPI taxonomies of sexual offenders

In a sample of 261 state hospital sexual offenders, Minnesota Multiphasic Personality Inventory (MMPI) profiles did not differ for offenders with adult victims versus offenders with child victims when offender age was controlled. MMPI 2-point analyses for the whole sample revealed five common codes that were independent of victim maturity. The sample was randomly divided in half and subjected to a cluster-analytic procedure which revealed two MMPI clusters. The first cluster was unelevated, with Scale 4 as its high point. The second cluster had multiple elevations, with Scales 8, 4, 2, and 7 as the highest scales. These clusters were replicated in a cluster analysis of the second half of the sample. However, when the sample was recombined, the two clusters were not externally validated basis on demographic and criminological variables. The results suggest that common psychological variables among sexual offenders may have more discriminative value than victim maturity in developing sexual offender taxonomies.     

Test-Retest Properties of the Least Preferred Co-worker (LPC) Score

The test-retest properties of the Least Preferred Co-Worker (LPC) scale and two subscales were examined under two conditions: (a) a naive condition, in which respondents were uninformed as to the meaning of the scale and (b) an informed condition, in which the scale was explained to respondents between administrations. The results support the hypothesis that the test-retest stability of the total LPC score would decrease significantly when respondents understood the scale. The stability of the interpersonal subscale exhibited a decrease similar to that of the total scale score, whereas the stability of the task subscale remained constant even when the scale was explained.     

Validity and clinical scale comparisons between the MMPI and MMPI-2 with psychiatric inpatients.

This study was designed to investigate the comparability of the original MMPI (1950) and the MMPI-2 (1989) with a psychiatric patient population. 34 male and 3 female patients, shortly after admission to one of two acute psychiatry units, completed the old and revised versions of the MMPI. Paired t tests indicated but scant differences for raw scores, while many more differences were found among T scores for validity, clinical, and supplemental scales. Analyses, however, showed all scales on the two forms to be highly correlated. Analysis of the high-point and two-point codes across the two administrations also showed relative stability, although the proportion of Scales 2 (Depression) and 8 (Schizophrenia) decreased, while those for Scales 6 (Paranoia) and 7 (Psychasthenia) increased markedly in the MMPI-2 protocols. Examination of each version's discriminability among mood- and thought-disordered subsamples suggested that the MMPI provides slightly better delineation between diagnostic classes. Discriminant function analyses showed that there were essentially no differences between the two forms in the accurate classification of clinical and nonclinical groups. The findings reported here provide support for the MMPI-2; despite modification, the newer form retains the advantages of the original MMPI. Differences found here may be unique to psychiatric patients and their patterns of MMPI/MMPI-2 equivalence and may not generalize to other special populations.     

Wine judging,context and New Zealand Sauvignon Blanc

Wine show competitions require judges to evaluate a large number of wines, typically within a time constraint. Under such circumstances, some form of quantification of wine quality is essential to achieve the aim of the task, namely allocation of a numerical score, or an award (e.g. a gold medal) that symbolises a numerical range of scores, to each wine. In this paper, we consider the relation between method of scoring, the scores awarded Sauvignon Blanc wines in a simulated wine show competition, and several aspects of wine-judging behaviour. Twenty experienced wine industry professionals judged 15 New Zealand Sauvignon Blanc wines via a 20-point scoring system, the system currently used in New Zealand wine shows, and via a 100-point scoring system in a context that simulated a wine competition. We were interested in two quantitative issues. The first related to the wines, where we investigated whether the 100-point judging system led to greater differentiation of the wines than the 20-point system. The second issue concerned wine-judging behaviour. We were interested in whether scoring method would influence between- and/or within-judge variability, with greater consistency resulting from use of the 20-point system. Results showed that there was no greater differentiation of the Sauvignon Blanc wines when they were judged by the 100-point scale than when judged out of 20 points. Variation in scores given to each wine on each scale was also generally consistent. With respect to whether method of scoring influenced variability of wine judges, we employed a model described by Schlich (1994) to consider measures of wine-evaluation behaviour. The major result was that consistency, both within judges and between judges, was independent of type of scoring method. Implications of the findings are discussed.     

Total number of MMPI-2 clinical scale elevations predicts premature termination after controlling for intake symptom severity and personality disorder diagnosis

Premature termination of therapy by patients is a common phenomenon that can be deleterious to treatment outcome for patients and also negatively affect therapists, treatment centers, and research programs. Therefore, a method of identifying patients at risk for premature termination could have widespread benefits. This study investigated whether patients’ MMPI-2 profiles, including clinical scale elevations and Negative Treatment Indicator (TRT) scores, could predict premature termination in an outpatient sample, controlling for personality disorder diagnosis and symptom severity at intake. Results indicated that while TRT scores were not incrementally predictive of premature termination, the total number of clinical scale elevations was significantly incrementally predictive of dropout. Clinical implications are discussed.     

A method for measuring "primitive" and "advanced" elements in pleasures and aversions

Subjects were asked to list pleasures and aversions actually experienced, and pleasures and aversions based on their "innermost fantasies." A 7-point rating scale was developed to rate "primitive" and "advanced" elements in these pleasure-aversion responses. The "primitive" pole of the scale was keyed with the phrase "gut-level experiencing" and the "advanced" pole with "detached objective thinking." Reality Pleasure, Fantasy Pleasure, Reality Aversive, Fantasy Aversive, and Total scores were derived by having trained raters rate each subject's response with the 7-point rating scale. In Study 1, 25 females and 25 males provided pleasure-aversion responses and the data analyzed in a mixed 2 (male-female) x 2 (reality-fantasy) x 2 (pleasure-aversive) ANOVA. The most important finding was a highly significant reality-fantasy x pleasure-aversive interaction: subjects tended to "progress" (give more advanced responses) going from reality to fantasy in their pleasure responses, while exactly the opposite occurred with the aversive responses. Also, as predicted, reality aversive responses tended to be more "advanced" than reality pleasure responses. Study 2 was similar in design, but several additional personality measures were administered to assess the construct validity of the pleasure-aversion scores. As in Study l, the reality-fantasy x pleasure-aversive interaction was highly significant and again reflected a pattern of "progression" going from reality to fantasy for pleasure responses and "regression" for aversive responses. The correlation analysis also produced several interesting findings with sex serving as a major moderator variable.