A Comparative Review of Current Practices in Personality Assessment Norming

Exploring Potential Ethnic Bias Among MMPI-3 Scales in Assessing Personality Psychopathology

This study examined statistical bias in the measurement of personality psychopathology in the Latinx population using the Minnesota Multiphasic Personality Inventory-3 (MMPI-3). Data were extracted from two studies that yielded a composite data set of 103 White individuals and 250 Latinx individuals. All participants were administered the MMPI-2-Restructured Form-Extended Battery (MMPI-2-RF-EX) or MMPI-3 and the Personality Inventory for the DSM-5 Short Form (PID-5-SF). First, we conducted correlation analyses between theoretically overlapping scales of the PID-5-SF and the MMPI-3 among White and Latinx individuals. The majority of theoretically associated scales were found to be at least moderately associated in the total sample. In addition, Steiger's z-tests indicated that correlations were similar in magnitude across the White and Latinx ethnic groups. Hierarchical regression subsequently determined the presence of slope and/or intercept bias. Only one analysis (the MMPI-3 Anger Proneness prediction of PID-5-SF Negative Affectivity) indicated statistically significant intercept bias. No evidence of slope bias was found. In other words, these analyses indicated that the vast majority of the relationships between MMPI-3 scales and associated personality psychopathology constructs (as measured by the PID-5-SF) remained consistent across both ethnic groups. Overall, the results supported the appropriate cross-cultural use of the MMPI-3 to assess personality psychopathology.

Kids these days and kids those days: Investigating perceptions of children's social skills from 1988 to 2007

Behavior rating scales are frequently used assessment tools designed to measure social skills. Use of norm-referenced assessments such as behavior rating scales requires examiners and test publishers to consider when norms become obsolete and norm-referenced scores can no longer be validly interpreted. A fundamental factor influencing norm obsolescence regards changes in baseline levels of targeted traits within the population. Yet, limited research exists regarding how social skills may change at a population level over time as measured by established assessment tools. Thus, the present study investigates population trends in social skills of K-12 children as rated by parents, teachers, and students by concordantly linking the Social Skills Rating System (SSRS; nparent = 833, nteacher = 1215, nstudent = 4105) and the Social Skills Improvement System-Rating Scales (SSIS-RS; nparent = 2400, nteacher = 750, nstudent = 800) using validity samples collected during the development of the SSIS-RS (nparent = 240, nteacher = 221, nstudent = 224). Analyses evaluated differences between ratings on the standardization data from 1988 and 2007 by informant, sex, grade level, and sex by grade level. After applying linear linking techniques, we conducted a series of statistical comparisons that revealed a general upward trend of ratings for the 2007 sample compared to the 1988 sample, with important differences across sex, grade level, and informant. We conclude with a discussion of the implications of these findings for consideration and assessment of children's social skills.

Validation of the Mexican Spanish Version of the Scale of Positive and Negative Experience in a Sample of Medical and Psychology Students

The Scale of Positive and Negative Experience (SPANE) aims to measure affect with high transcultural validity. The bifactor model is the best theoretical option to represent affective balance, although it is not typically used in validation studies. The objectives of this research were to test a bifactor model vis-à-vis the traditional model composed of two correlated factors, to prove its invariance across sexes, and to provide evidence of concurrent validity. A nonprobability sample composed of 600 Mexican students of psychology and medicine was recruit. One-group and multigroup confirmatory factor analyses were carried out. The SPANE and the scales selected to assess depression, perceived stress, and satisfaction with life were applied. The bifactor model showed better goodness-of-fit indices than the two correlated factors model: Δχ²(11) = 121.436, p < . 001, Δχ²/Δdf = 11.04 > 5, ΔGFI = .034, ΔNFI = .025, ΔNNFI = .022, and ΔCFI = .023 >.01. The internal consistency for the general factor as well as for the factor of positive affect was excellent, whereas it was good for the factor of negative affect. The measurement model was valid across sexes. The general factor of affective balance had a very high correlation with depression, high with perceived stress, and medium with satisfaction with life. It is concluded that SPANE is reliable and shows evidence of validity among Mexican student of psychology and medicine, and the bifactor model is adequate to represent affective balance.