Substance or style? An investigation of the NEO-PI-R validity scales

The Revised NEO Personality Inventory (NEO-PI-R; Costa & McCrae, 1992b) has been criticized for the absence of validity scales designed to detect response distortion. Recently, validity scales were developed from the items of the NEO-PI-R (Schinka, Kinder, & Kremer, 1997) and several studies have used a variety of methods to test their use. However, it is controversial whether these scales are measuring something that is substantive (such as psychopathology or its absence) or stylistic (which might be effortful distortion or less conscious processes such as lack of insight). In this study, we used a multimethod-multitrait approach to examine the validity of these scales in a clinical sample of 668 participants diagnosed with personality disorders or major depression. Using various indicators of both stylistic and substantive variance, confirmatory factor analyses (CFA) suggested that these validity scales measure something that may be conceptually distinct from, yet highly related to, substantive variance in responding.     

An openness scale for the California Psychological Inventory

We developed a 36-item scale to measure Openness, using items on the California Psychological Inventory (CPI; Gough, 1957, 1987, 1996), Form 434. Items were initially chosen on the basis of content validity. Five samples (N = 2,375) were used to establish reliability, validity, and norms; 4 samples consisted of university undergraduate students, and 1 comprised applicants for nonmanagement call centerjobs. Internal consistency estimates obtained in each sample averaged approximately .75, and test-retest stability, assessed in 1 sample, was estimated at .84. Cross-correlations with related scales, for example, the NEO Personality Inventory-Revised Openness scale (Costa & McCrae, 1992) and other CPI-based scales, provided evidence of construct validity. Statistically significant predictive validities were obtained in 2 call centerjob-incumbent samples, with range-corrected true validities of .20 to .36 for a number of job performance criteria. Construct and predictive validity were found to be higher than for other scales consisting of CPI items designed to measure Openness or a related construct. Finally, norms were prepared for university undergraduate students (n = 1,847) and nonmanagement service-sector job applicants (n = 528).     

Validity of the MMPI subtle items

This article presents a review and critique of research on the Minnesota Multiphasic Personality Inventory (MMPI) subtle items in regard to their validity as measures of their scales' characteristics. The main categories of focus are their convergent and discriminative validity, the paradoxical effect, and the issue of whether removing these items would improve MMPI validity. Important problems in the methodologies are noted and discussed in some detail. These methodological concerns and the conflicting results across studies obscure firm conclusions on the validity of the subtle items, although strong support is lacking. However, more evidence is needed before deciding to remove these items from the MMPI; therefore, this article provides recommendations for future research.     

Further evidence on 16 PF distortion scales

Norms were obtained on a nationally representative sample of 4830 adult men and women for two empirically developed faking indices for Cattell's 16 Personality Factor Questionnaire (16PF) by Winder, Karson, and O'Dell (1975). These data provided convincing evidence that the cutoff suggested in the original work for the faking good scale was far too liberal and would routinely classify more than half of all 16 PF protocols as invalid. The faking bad cutoff appeared to be approximately correct. The correlations of the faking scales with the 16 PF primary factors were highly congruent with those reported in the development study and provided additional validity evidence for the two indices. Both indices were found to be sufficiently reliable to permit adjustments to be made in the primary trait scales when distortion is above average.     

Positive response distortion by police officer applicants: association of Paulhus Deception Scales With MMPI-2 and Inwald Personality Inventory Validity Scales

Interpretation of positive response distortion (socially desirable responding) in employment evaluations is an important validity issue. This study of police officer applicants examined the construct validity of the Paulhus Deception Scales (PDS)-Moralistic Bias (MB; exaggerated adjustment/agreeableness) and Egoistic Bias (EB; exaggerated power/ status)-in relation to validity scales of the Minnesota Multiphasic Personality Inventory-2 (MMPI-2; L, K, and S) and Inwald Personality Inventory (IPI; Guardedness). In regression analyses, MB was significantly associated with each validity scale (particularly L and Guardedness), whereas EB was significantly, but weakly, associated with L only. MB is consistent with response distortion as reflected in L ("perfect" adjustment/personality) and Guardedness (denial of shortcomings/faults). EB is a unique form of response distortion that is not reflected in MMPI-2 or IPI validity scales. The relevance of EB to self-assessment among police officer applicants is an important practical concern in personnel selection and an important theoretical question for future response distortion research.     

Detection of feigned mental disorders: a meta-analysis of the MMPI-2 and malingering

The validity of test data from multiscale inventories is dependent on self-reports that may be easily distorted by malingering. In examining the Minnesota Multiphasic Personality Inventory-2's (MMPI-2) role in the assessment of feigning, this review provides a conceptual analysis of the detection strategies underlying the MMPI-2 validity scales. The conceptual analysis is augmented by comprehensive meta-analysis of 65 MMPI-2 feigning studies plus 11 MMPI-2 diagnostic studies. For the rare-symptoms strategy, Fp (Cohen's d = 2.02) appears especially effective across diagnostic groups; its cut scores evidence greater consistency than most validity indicators. The data supported the F as an effective scale but questioned the routine use of Fb. Among the specialized scales, Ds appeared especially useful because of its sophisticated strategy, consistent cut score, and minimalfalse-positives. General guidelines are offeredfor specific MMPI-2 validity scales in the assessment of malingering with specific diagnoses.     

Malingering on the Personality Assessment Inventory: identification of specific feigned disorders

Profile validity is a primary consideration in the clinical assessment of psychopathology. Several indicators of negative impression management have been developed for the Personality Assessment Inventory (PAI; Morey, 1991) that can both indicate its presence and help differentiate effortful negative distortion from distortion arising from cognitive sets associated with psychopathology. In this study, we tested a method designed to delineate the specific Clinical scales relevant for interpretation of deliberately feigned disorders. We used associations between the Negative Impression Management (NIM) scale and Clinical scales in the normative standardization sample to derive NIM predicted scale scores in a regression framework. We contrasted these predicted scores with observed scores on Clinical scales to yield NIM predicted discrepancies hypothesized to identify those Clinical scales most salient for the interpretation of negative distortion. We found this method to be effective in identifying particular distortion on the relevant scales for individuals attempting to feign 3 specific diagnoses (major depressive disorder, generalized anxiety disorder, and schizophrenia).     

Factor structure of the basic personality inventory with incarcerated offenders

The item and scale factor structure of the Basic Personality Inventory (BPI) was examined in a sample of 486 offenders incarcerated for violent and sexual crimes. Separate principal-component analyses of the items for each of the 11 clinical scales, critical item scale, and social desirability scale indicated a one-dimensional factor solution for all scales except Depression and Persecutory Ideation. The Depression scale's two factors were Hopelessness and Depressive Affect and the Persecutory Ideation scale's two factors were General Paranoia and Perception of External Control. Although the factors for these two scales may assist in interpretation, the correlations between the factors and the total score of their respective scale were high. Confirmatory factor analysis of the 220 items from the 11 clinical scales supported the factorial logic of the scoring key. Analysis of the 11 clinical scales resulted in two factors: General Psychopathology/Adjustment and Antisocial Orientation. The results suggest that all but two scales can be viewed as unidimensional thereby allowing for a straightforward clinical interpretation. These analyses support the internal structure of the BPI and lend credence to external validity work with forensic populations.     

Assessing the likelihood of reliable workplace behavior: further contributions to the validation of the Employee Reliability Inventory.

This paper summarizes a number of studies in which the validity of the Employee Reliability Inventory, a preemployment screening instrument designed to assess the likelihood of reliable and productive workplace behavior, was examined. Criterion-related studies compared the scores of a broadly diverse group of job applicants with those obtained from an array of criterion and comparison groups, for whom there was documented evidence of reliable or unreliable behavior. Criterion-related evidence indicates that the six scales are effective in differentiating a variety of criterion groups with unreliable behavior from a number of different job applicant comparison groups. Construct-related evidence for the validity of an emotional adjustment scale is reported as well. The issue of response distortion in preemployment inventories is discussed, and data are reported which indicate that scores on all six scales appear to be functionally free from the potentially confounding effects of response distortion. These results are consistent with the original validation and cross-validation findings, which supported the validity of the six scales when assessing the likelihood of reliable behavior in a population of job applicants.     

Assessing MMPI Constructs Through the Measurement of Response Latencies

Responses to the Minnesota Multiphasic Personality Inventory (MMPI) were assessed with respect to their relevance to schema theory. The relation between scores on self-reported personality dimensions and the speed of processing test items associated with each dimension was examined. With previously derived factor analytic content scales, negative correlations were obtained between scale scores and mean latencies for endorsing relevant items, and positive correlations were found between scale scores and mean latencies for rejecting relevant items. A similar analysis completed on the traditional clinical scales revealed no such pattern. Results were interpreted as supporting the conceptualization of item responding as a content-based, schema-relevant process.     

Assessing MMPI constructs through the measurement of response latencies

Responses to the Minnesota Multiphasic Personality Inventory (MMPI) were assessed with respect to their relevance to schema theory. The relation between scores on self-reported personality dimensions and the speed of processing test items associated with each dimension was examined. With previously derived factor analytic content scales, negative correlations were obtained between scale scores and mean latencies for endorsing relevant items, and positive correlations were found between scale scores and mean latencies for rejecting relevant items. A similar analysis completed on the traditional clinical scales revealed no such pattern. Results were interpreted as supporting the conceptualization of item responding as a content-based, schema-relevant process.     

A Structural Analysis of the Impact of MMPI-2 on MMPI-1

Minnesota Multiphasic Personality Inventory-1 (MMPI-1) items modified for Minnesota Multiphasic Personality Inventory-2 (MMPI-2) are presented in pairs to illustrate modifications. MMPI-1 items deleted from MMPI-2 are presented, grouped by categories. Frequencies of items remaining in MMPI-2 for the following special scales are presented: Harris and Lingoes subscales of the clinical scales (Harris & Lingoes, 1955/1968); Wiggins Content scales (Wiggins, 1966); Tryon, Stein, and Chu (TSC) cluster scales (Stein, 1968; Tryon, 1966); Indiana Rational scales (Levitt, 1989); and selected other special scales. Frequencies of MMPI-1 items in the MMPI-2 validity and content scales are also presented, k is concluded that most of the special scales developed for MMPI-1 remain relatively intact in MMPI-2.     

Cross-sectional studies of personality in a national sample: 1. Development and validation of survey measures

Short scales were developed to measure three broad dimensions of personality in the National Health and Nutrition Examination Survey (NHANES I) Epidemiologic Followup Study. Items to measure neuroticism were selected rationally from the General Well-Being Schedule, and items to measure extraversion and openness to experience were selected by multiple regression from the NEO Personality Inventory. In a sample of 654 men and women from the Augmented Baltimore Longitudinal Study of Aging, all three short scales showed clear evidence of convergent and discriminant validity against full scales in self-reports, and against peer and spouse ratings. Because they were developed and validated on a sample with a comparable age range, the NHANES scales should be useful in examining age trends in these three personality dimensions in a national sample.     

Psychometric Properties of the Right Wing Authoritarianism Scale in Black and White South African Students

Recently Taylor and Boeyens (1991) showed that the South African Personality Questionnaire has inadequate internal consistency and factorial validity for use in the black population. This paper reports a study supporting their conclusion that scales developed with data from a white sample do not hold together psychometrically or conceptually when used with a black sample. In a sample of white students, Altemeyer's Right Wing Authoritarianism scale, initially developed in Canada, had an acceptable alpha of 0.83, but there was an unacceptable alpha of 0.43 in a black student sample. Item and factor analysis showed that for the black sample there was little structure and that the scale was effectively a set of heterogenous items. This paper presents a detailed discussion of the item and factor analyses and shows how certain items have different connotations for whites and blacks as a result of their different political and cultural histories. Although cultural differences in conservatism can be observed by looking at individual items, a comparison between black and white individuals on the higher order construct of Authoritarianism cannot be made. It is concluded that basic conceptual work based on the phenomenology of the concept being measured needs to precede the psychometric development of scales for use in the black community.     

Characteristics of items in the Eysenck Personality Inventory which affect responses when students simulate

A large sample of students completed Form A of the Eysenck Personality Inventory, and four subgroups were later asked to simulate extraversion, introversion, neuroticism or stability. It was found that subjects could simulate these four personalities successfully. The changes in individual item responses were correlated with the items' factor loadings, validity, response bias, and detectability. The different scales and types of item were considered separately. In some cases the changes in item responses when simulating introversion and extraversion were related to the extraversion validities and factor loadings of the items. More often, however, the behaviour of items under simulation was correlated with aspects of the items that made them more like an item from another scale and thus lessened their susceptibility to a particular type of simulation.     

The validity and utility of the positive presentation management and negative presentation management scales for the Revised NEO Personality Inventory

Schinka, Kinder, and Kremer developed "validity" scales for the Revised NEO Personality Inventory (NEO PI-R; Costa & McCrae) to detect underreporting-the Positive Presentation Management (PPM) Scale and overreporting-the Negative Presentation Management (NPM) Scale. In this investigation, the clinical utility of these scales was examined using the established validity scales from the Minnesota Multiphasic Personality Inventory-2 (MMPI-2; Butcher et al.) as the referent. The sample was composed of 370 psychiatric patients who completed the NEO PI-R and the MMPI-2 as part of a routine evaluation. Results indicated that response distortion compromised the utility of the NEO PI-R domain scales. Moreover, the PPM and NPM scales and an NPM-PPM index significantly differentiated invalid under-and overreporting groups from a valid responding group. The PPM and NPM-PPM index were adequate in classifying under- and overreporters, respectively.     

Special Abilities Assessment in Children

The faking-detection validity and incremental validity of response latencies to Minnesota Multiphasic Personality inventory (MMPI) items was investigated using an analog research design. One hundred undergraduates were assigned at random to five groups: each group received different faking instructions (standard, fake good, fake bad, fake good with incentive, fake bad with incentive). All subjects completed a computer-administered version of the MMPI. Content-determined response deviance scores and latencies of responses to Subtle and Obvious scale items were determined for each subject. The principal findings suggest that response latencies may have greater faking good detection ability than responses deviance scores, and that response latencies have statistically significant incremental validity for both the detection of faking good and faking bad, when latencies are used with response deviance scores obtained from Subtle and Obvious scales.