Sex and gender bias in self-report personality disorder inventories: item analysis of the MCMI-II, MMPI, and PDQ-R

There has been considerable controversy and research regarding sex bias in the diagnosis of personality disorders, but little has involved self-report inventories. Thus this study investigated items from the Millon Clinical Multiaxial Inventory-II (Millon, 1987), the Minnesota Multiphasic Personality Inventory (Morey, Waugh, & Blashfield, 1985), and the Personality Diagnostic Questionnaire-Revised (Hyler & Rieder, 1987). Subjects (N = 189) completed the Histrionic, Dependent, Antisocial, and Narcissistic scales from these inventories, along with the Bem Sex Role Inventory (Bem, 1974) and the Symptom Checklist-90-Revised (Derogatis, 1977). Items were considered to evidence sex or gender bias if they (a) failed to correlate with dysfunction and (b) exhibited sex or gender role differences. At least 13 items evidenced sex bias (76 items using a more liberal threshold). The majority were from Narcissistic scales; few Histrionic items evidenced sex or gender bias. Implications with respect to sex-bias assessment and item construction are discussed.     

Racial bias in prediction with the MMPI for a juvenile delinquent population

Relationships between MMPI scales and criteria were evaluated to determine if the MMPI is racially biased with a juvenile delinquent population. The MMPI was administered to 333 white and 107 black male juvenile delinquents, and criterion data were collected. The regression equations developed for Blacks and whites resulted in similar accuracies. However, the weights of the regression equations tended to differ for the two races. How they differed depended on the criterion predicted by the equations. Consequently, it was impossible to state unconditionally that the use of the MMPI with Black delinquents will have an adverse impact. In fact, one interpretation of the results suggested that the MMPI may not be racially biased for predictions with juvenile delinquents because the amount of constant error in prediction for a racial group decreased toward zero as the apparent objectivity of the criteria increased.     

A Canonical-Correlation Analysis of 16 Fear Factors and the MMPI

To determine clinical correlates of 16 fear factors of the Wolpe-Lang Fear Survey Schedule, the fear factor and MMPI scores of 92 psychiatric inpatients were subjected to a canonical-correlation analysis. The results produced three combinations of scores which yielded statistically reliable canonical-correlation coefficients: three fear factors related to psychotic levels of personality disorganization; two related to fairly directly experienced neurotic anxiety; and five related to neurotic levels of anxiety “bound” by somatic complaints. The results suggest different treatment approaches for different patterns of expressed fears. The interpretation of several of the factors (previously related to clinical status) awaits further research.     

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.     

Genuine memory bias versus response bias in anxiety

Low and high trait anxiety undergraduates were presented with physical‐threat, ego‐threat, positive, and neutral words. Following an orienting task promoting lexical—but not semantic—processing, unexpected free recall or recognition tests were presented. High anxiety participants showed increased correct recall of both types of threat‐related words, but also increased incorrect recall (intrusions) and incorrect recognition (false alarms) of these words. Furthermore, participants high in anxiety had reduced sensitivity (d′) for ego‐threat words, and reduced cautiousness (β) for physical‐threat words. This tendency to report threat‐related information regardless of prior presentation suggests that there is a response bias rather than a memory bias in anxiety. In addition, this bias is likely to be mediated by depression insofar as physical‐threat information is concerned, although the bias can be attributed to trait anxiety insofar as ego‐threat information is concerned.     

Accounting for successful control of implicit racial bias: the roles of association activation, response monitoring, and overcoming bias

Individuals who are primarily internally motivated to respond without prejudice show less bias on implicit measures than individuals who are externally motivated or unmotivated to respond without prejudice. However, it is not clear why these individuals exhibit less implicit bias than others. We used the Quad model to examine motivation-based individual differences in three processes that have been proposed to account for this effect: activation of associations, overcoming associations, and response monitoring. Participants completed an implicit measure of stereotyping (Study 1) or racial attitudes (Study 2). Modeling of the data revealed that individuals who were internally (but not externally) motivated to respond without prejudice showed enhanced detection and reduced activation of biased associations, suggesting that these processes may be key to achieving unbiased responding.     

Obtaining 16 PF Scores from the MMPI,and MMPI Scores from the 16 PF

Recent statistical analyses of the relation of the MMPI surface traits to the source traits in the 16 P.F. were extended to (a) the calculation of beta weights for estimating MMPI scale scores from the 16 P.F. and (b) deriving the MMPI item composition in terms of 16 P.F. scores from the MMPI, and vice versa. It has been argued from the specific nature of surface and source traits that additional diagnostic insight can be gained from their comparison, a principle which has been designated as depth psychometry.     

Race,social class,and response bias effects in grammatical processing

Black English and Standard English grammatical sentences were presented to 122 Black and White second graders in a modified cloze task. Both race and social class differences were sampled from within a single primary school, and both Black and White experimenters were utilized. No effects due to social class were found. Black subjects did better on Black English sentences than White subjects and showed fewer errors in the direction of Standard English syntax. An analysis of hit rates and false alarm rates suggested that Black subjects were more likely to emit Black English responses than White subjects but that both groups possessed equivalent sensitivity to grammatical differences. The results are not consistent with a deficit account of Black subjects' performance or with an explanation based on different dialect systems. Rather, it appears that Black and White children may differ only in their willingness to utilize specific response systems and may possess equal comprehension abilities in both dialect systems.This research was supported in part by funds made available to Bowling Green State University by NSF Institutional Grant No. GU-3619. Some of the findings were reported at the 83rd Annual Convention of the American Psychological Association, Chicago, 1975.     

Academic Achievement,Ability, and the MMPI Scales

This study investigated the relationship between personality traits and academic achievement at the university level, while holding scholastic ability constant. Although no distinctive traits characterized the achiever, the non-achiever showed pronounced tendencies in the direction of impulsivity, overactivity, and excessive sociability. He also scored significantly lower than the achiever in social responsibility.     

The effect of response bias on recall performance, with some observations on processing bias

We explored Roediger and Payne's proposal that response bias does not affect recall performance and that it is therefore not necessary to control for response productivity in recall studies. Two initial experiments, contrary to expectation, corroborated Roediger and Payne's findings: Forced recall did not produce more correct recalls than free recall, even though forced recall produced substantially more false alarms than did free recall. However, in succeeding experiments involving pictorial and verbal stimuli, reliable response-bias effects on recall were demonstrated. The stimuli yielding response-bias effects were those associated with higher probabilities of being guessed by chance. In addition, some of the data suggest that processing-bias effects (differential retrieval effort) may be unintentionally induced by instructions and may significantly affect recall memory. Consequently, it is necessary to assess or to control response-bias effects and, possibly, processing-bias effects in recall experiments in which level of recall is of interest.     

Behavioral bias and left-right response differentiation in the rat

To examine the proposition that lateral asymmetry facilitates left-right response differentiation in rats, we examined the relationships between the strengths of several behavioral biases and the scores on a learning task requiring left-right response differentiation. No support was found for a simple model positing a monotonic relationship between any behavioral bias and the learning scores. However, performance showed a U-shaped relationship to one behavioral bias. This finding conforms to a curvilinear model in which rats at either extreme of asymmetry are disadvantaged, at low degrees of asymmetry by a lack of navigational reference, and at high degrees by resultant strong position habits; moderately asymmetrical rats have neither disadvantage and are best able to use the asymmetry as a reference in processing left-right information.