The predictive validity of subtle and obvious empirically derived psychological test items under faking conditions

The relative contributions of subtle and obvious item endorsements to the prediction of a relevant criterion were assessed under faking and control ("honest") conditions. The MMPI and a nonconformity questionnaire were first administered to 100 male college students. Items on the Pd scale and 101 additional MMPI items that correlated significantly with the nonconformity questionnaire were then rated by 38 other male college students for apparent relationship to psychopathology. From these ratings, a scale (designated PdX) was constructed, which consisted of 21 subtle and 21 obvious items. After a third group of 98 male college students completed the nonconformity questionnaire, they were asked to respond to the items of the Pd and PdX subscales under control, fake-good, and fake-bad instructions. Significant correlations between the nonconformity scale and certain PdX and Pd subscales were found only for the control group. Implications for test construction and for clinical interpretation under faking conditions are discussed.     

A DIMENSIONAL ANALYSIS OF MMPI ITEMS

A method of analysis specifically designed for binary data was applied to 100 MMPI items. Sixty, items were carefully chosen to represent the nine major clinical scales with respect to direction of keying, social desirability scale value and endorsement frequency. The remaining 40 items were randomly chosen from items not appearing on any of these scales. Although a complete solution was obtained in five dimensions, only three were retained. The three dimensions were related to scale membership, gender of the respondent and various item characteristics. The results clearly support the two major MMPI factors obtained on a scale level and additionally show a strong gender dimension.     

Construction and cross-validation of an MMPI Black-White scale

A robust set of MMPI items, sensitive to the effect of race (Blacks versus Whites), was constructed by compiling the results of six prior studies. Thirty-two items comprised the MMPI B-W scale. The scale was cross-validated on psychiatric patients and police cadets. Statistically significant mean differences were found between races in both groups, with 83% and 60% correct racial classification, respectively. These findings suggested that when used as a clinical instrument, the MMPI might pick up nonpathology-related racial variance contributing to scale devotion biased against the black respondent, but the degree of nonpathology-related racial scale score covariance has yet to be determined. For future research, the B-W scale was proposed as a set of standard queries in studies assessing connotative meaning for Black-White contrast groups, and to index "functional" as opposed to "objective" race membership.     

Aggressive response sets and subtle-obvious MMPI scale distinctions in male offenders

The effects of aggressive and nonaggressive response sets on the MMPI subtle and obvious clinical subscales were investigated. Fifty-eight male prison inmates answered the MMPI as if they were either highly aggressive or highly nonaggressive. The clinical scales with sufficient items in each category were scored for subtle, neutral, and obvious subscales. Inmates successfully feigned aggressiveness on several of the obvious subscales (p = .0056) and one neutral scale; the subtle subscales were not significantly different across groups, consistent with previous research on this population in terms of the resistance of subtle items to these response sets.     

The Validation of Projective Tests

There are 16 items in the standard MMPI group forms which are included twice. It was found that a number of computerized scoring services use only the first occurrence of repeated items in scoring the MMPI scales, whereas the handscoring templates use an arbitrary combination of the first and second occurrence of these items. Comparison of these conventions suggested a potential for significant differences in score, particularly on the Sc scale. Scoring a set of 126 MMPIs of chronic pain patients by both of these scoring conventions revealed differences of up to 10 T points on the Sc scale. It is recommended that a single scoring convention for the MMPI be adopted by psychologists. For several reasons we suggest that only the first occurrence of repeated items should be used for scoring purposes. In the absence of a single convention, comparisons between Sc scores on different protocols should not be made without first insuring that the protocols were scored in the same manner.     

Alternative scoring system of repeated items on the MMPI: caveat emptor

There are 16 items in the standard MMPI group forms which are included twice. It was found that a number of computerized scoring services use only the first occurrence of repeated items in scoring the MMPI scales, whereas the handscoring templates use an arbitrary combination of the first and second occurrence of these items. Comparison of these conventions suggested a potential for significant differences in score, particularly on the Sc scale. Scoring a set of 126 MMPIs of chronic pain patients by both of these scoring conventions revealed differences of up to 10 T points on the Sc scale. It is recommended that a single scoring convention for the MMPI be adopted by psychologists. For several reasons we suggest that only the first occurrence of repeated items should be used for scoring purposes. In the absence of a single convention, comparisons between Sc scores on different protocols should not be made without first insuring that the protocols were scored in the same manner.     

The MMPI–2–RF Personality Psychopathology Five (PSY–5–RF) Scales: Development and Validity Research

This article describes the development, internal psychometric, and external validation studies on scales designed to measure the Personality Psychopathology Five (PSY–5) from MMPI–2 Restructured Form (MMPI–2–RF) items. Diverse and comprehensive data sets, representing various clinical and nonclinical populations, were classified into development and validation research samples. Item selection, retention, and exclusion procedures are detailed. The final set of PSY–5–RF scales contain 104 items, with no item overlap between scales (same as the original MMPI–2 PSY–5 scales), and no item overlap with the Demoralization scale. Internal consistency estimates are comparable to the longer MMPI–2 PSY–5 scales. Appropriate convergent and discriminant validity findings utilizing various self-report, collateral rating, and record review data are reported and discussed. A particular emphasis is offered for the unique aspects of the PSY–5 model: psychoticism and disconstraint. The findings are connected to the broader PSY–5 literature and the recommended review of systems (Harkness, Reynolds, & Lilienfeld, this issue) presented in this series of articles.     

Conversion equations for modified scoring of the MMPI-168

For a sample of 300 patients who had been administered the Minnesota Multiphasic Personality Inventory (MMPI), the MMPI-168 was extracted from the full MMPI and scored to incorporate those items normally excluded by Form R keys. MMPI-168 correlations with the full MMPI ranged from .80 to .97 with a mean of .90, indicating satisfactory statistical validity, and modified scoring was shown to improve predictability for Pa and Sc. Using these data, substitution equations for transforming MMPI-168 raw scores to estimates of full-scale scores were calculated. These transformations did not differ greatly from those reported in previous research except on Pa and Sc, where additional items increase scale length substantially.     

Relations between the family environment scale (FES) and the MMPI

This study investigated the relations between FES scales and MMPI scales and the relation of social desirability and endorsement of FES items in a sample of 185 college students. FES scales were found generally not to be highly redundant with MMPI variables although MMPI Scale K (defensiveness) was significantly related to several FES scales. The social desirability and the endorsement rates of FES items were found to correlate approximately .80. Implications of these findings for the research and clinical use of the FES are discussed.     

Interpretation of a Full-Information Item-Level Factor Analysis of the MMPI-2: Normative Sampling and Nonpathognomonic Descriptors

An exploratory item-level full-information factor analysis was performed on the normative sample for the MMPI-2 (Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989). This method of factor analysis, developed by Schilling and Bock (Bock & Schilling, 1997) and based on item response theory, works directly with the response patterns and avoids the artifacts associated with phi coefficients and tetrachoric coefficients. Promax rotation of the factor solution organizes the clinical scale items into 10 factors that we labeled Distrust, Self-Doubt, Fitness, Serenity, Rebelliousness, Instrumentality, Irritability, Artistry, Sociability, and Self-Reliance. A comparison was made to the results of Johnson, Butcher, Null, and Johnson (1984), who performed a principal-component analysis on an item set of 550 items from the previous version of the MMPI (Hathaway & McKinley, 1943). Along with version changes and sampling differences, the essential differences between Johnson et al.'s results and ours may be attributed to differences between the Schilling and Bock method, which uses all information in the item responses, and the principal-component analysis, which uses the partial information contained in pairwise correlation coefficients. This study included 518 of the complete 567 items of the MMPI-2, versus Johnson et al.'s retention of 309 of the initially included 550 items of the previous MMPI. The full-information analysis retained all 518 initially included items and more evenly distributed the items over the 10 resulting factors, all sharply defined by their highest loading items and easy to interpret. Sampling effects and factor label considerations are discussed, along with recommendations for research that would validate the clinical utility of the implied scales for describing normal personality profiles. The full-information procedure provides for Bayes estimation of scores on these scales.     

Dimensions of stuttering and relationship to psychopathology

This article reports two studies. The first attempted to define stable dimensions within the Stuttering Severity (SS) scale. A factor analysis of correlations among its 64 items defined four factor-based subscales, and three additional subscales were developed rationally. Relationships among these seven subscales suggested the existence of two major dimensions. The second study investigated the relationship of the seven subscales to psychopathology. Correlations were obtained among the subscales and MMPI scales for 69 subjects, and were subjected to a second factor analysis. The MMPI scales and the SS subscales clearly loaded on separate factors, indicating little relationship between dimensions of stuttering and psychopathology as defined by MMPI scores. The second analysis also supported the previous identification of two general dimensions for stuttering. Items were selected for two final scales to represent these dimensions, labeled behavior (22 items) and sensitivity (20 items). Norms were developed for them and also for the full 64-item SS scale.     

Racial identity and the MMPI in African American male college students

The relation of the Minnesota Multiphasic Personality Inventory (MMPI) to the Racial Identity Attitude Scale--Black, Short Form (RIAS-B) was examined among 50 African American male college students in a reanalysis of unpublished MMPI data described in R. H. Dana (1993). This permitted study of relationships between MMPI scores and specific psychological variables hypothesized to produce cultural differences among African Americans. Results indicated RIAS-B scale scores functioned as predictors of MMPI scale scores. Similar research with the MMPI-2 incorporating larger samples that more adequately represent African American heterogeneity is needed. Methodological implications of these findings for MMPI-2 research with ethnic and racial groups are discussed, emphasizing the importance of research to depathologize African American racial and cultural identity.     

Interpretation of a full-information item-level factor analysis of the MMPI-2: normative sampling and nonpathognomonic descriptors

An exploratory item-level full-information factor analysis was performed on the normative sample for the MMPI-2 (Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989). This method of factor analysis, developed by Schilling and Bock (Bock & Schilling, 1997) and based on item response theory, works directly with the response patterns and avoids the artifacts associated with phi coefficients and tetrachoric coefficients. Promax rotation of the factor solution organizes the clinical scale items into 10 factors that we labeled Distrust, Self-Doubt, Fitness, Serenity, Rebelliousness, Instrumentality, Irritability, Artistry, Sociability, and Self-Reliance. A comparison was made to the results of Johnson, Butcher, Null, and Johnson (1984), who performed a principal-component analysis on an item set of 550 items from the previous version of the MMPI (Hathaway & McKinley, 1943). Along with version changes and sampling differences, the essential differences between Johnson et al.'s results and ours may be attributed to differences between the Schilling and Bock method, which uses all information in the item responses, and the principal-component analysis, which uses the partial information contained in pairwise correlation coefficients. This study included 518 of the complete 567 items of the MMPI-2, versus Johnson et al.'s retention of 309 of the initially included 550 items of the previous MMPI. The full-information analysis retained all 518 initially included items and more evenly distributed the items over the 10 resulting factors, all sharply defined by their highest loading items and easy to interpret. Sampling effects and factor label considerations are discussed, along with recommendations for research that would validate the clinical utility of the implied scales for describing normal personality profiles. The full-information procedure provides for Bayes estimation of scores on these scales.     

A Memory-Systems Approach to the Classification of Personality Tests: Comment on Meyer and Kurtz (2006)

The purpose of this study was to examine whether gender differences on the Minnesota Multiphasic Personality Inventory–2 (MMPI–2; Butcher, Dahlstrom, Graham, Tellegen, &; Kaemmer, 1989) and Minnesota Multiphasic Personality Inventory–Adolescent (MMPI–A; Butcher et al., 1992) items are comparable across 2 distinctive cultural samples: Americans and Koreans. Using large, representative adult and adolescent samples from both cultures, we found that the American samples were associated with a higher proportion of items with gender differences than the Korean samples. The American adult sample produced gender differences on a higher proportion of items than did the American adolescent sample, but no such age difference was found between the Korean samples. Despite these differences between cultures and between age groups, content dimensions underlying items with gender differences were very similar across cultures and age groups, centering on stereotypical gender interests, behaviors, and emotions.     

Retention and treatment failure in Classwide Peer Tutoring: Implications for further research

Knowledge of basic arithmetic facts aids in the acquisition and speed of performance of arithmetic operations. A peer mediated instructional procedure, Classwide Peer Tutoring (CWPT), has been shown to increase the rates and accuracy of students responses to academic tasks and to improve student performance on a variety of academic tasks. The present investigation evaluated the retention of 100 subtraction items by primary grade-age children with mild disabilities using CWPT for 10 weeks. Short- and long-term retention of items, and rate of correct responding were assessed. Results indicated that the students' average pretest score was 58.0% correct, and their average weekly posttest score was 87.12 correct, a 27% improvement in accuracy. Short- and longer-term retention measured on posttests was 88.7 % and 85.0% correct, respectively. Improvement in students rate of accurate responding to subtraction items practiced during CWPT was obtained. Student failure to learn particular items was not attributable to fewer opportunities to practice these items, less accurate practice, or item difficulty. Students reported positive evaluation of CWPT and perceived positive social and self esteem outcomes. Directions for future research focusing on retention and analysis of treatment failure are suggested.     

Failed back surgeries and minnesota multiphasic personality inventory (MMPI) profiles

MMPI profiles were evaluated for 105 prospective surgical patients who had previously undergone surgery or other procedures for treatment of back pain. Patients were classified into groups having undergone zero, one, two, three, or four or more previous surgeries. While all groups demonstrated a characteristicsomatogenic profile, none of the MMPI validity or clinical scales significantly differentiated the groups and there was no relationship between increased number of surgeries and MMPI scale characteristics. These results support the nonoptimistic prognostication of thesomatogenic MMPI profile for surgical intervention for back pain but show no clear relationship of MMPI profile characteristics to degree of experience of previously failed surgery.     

MMPI scale development methodology

A multitude of special MMPI scales have been developed without sufficient consideration for the methodology of empirical scale construction. Suggestions are presented in this article regarding the selection of items for a new MMPI scale and the tasks needed for the evaluation of a new scale. An investigator who develops a new MMPI scale should consider (a) the need for K corrections of scale scores, and (b) the possibility that the information provided by the special scale can also be obtained from the standard MMPI scales.     

Some properties of the ego-resiliency subtle scale of the MMPI.

The Ego Resiliency-Subtle (ER-S) scale was developed by Block (1965) as a social desirability free measure of the first factor of the MMPI. The purpose of the present experiments was to determine whether or not the 40-items comprising the ER-S scale remain neutral regarding social desirability and if the subject's score on the scale remains constant when taken out of the context of the other MMPI items; if so, the ER-S scale could be used as a screening device. Experiment I demonstrated that the items were not neutral regarding social desirability when taken out of context, but Experiment II indicated that this did not influence the responses subjects gave to the items. Because a memory factor may have been operating in Experiment II, a third experiment was conducted to minimize this factor resulting in a male-female vs. embedded-alone interaction and an order of presentation vs. embedded-alone interaction.