Psychological Assessment

Marsh, Herbert W.; Lüdtke, Oliver; Muthén, Bengt; Asparouhov, Tihomir; Morin, Alexandre J. S.; Trautwein, Ulrich; Nagengast, Benjamin

doi: 10.1037/a0019227pmid: 20822261

NEO instruments are widely used to assess Big Five personality factors, but confirmatory factor analyses (CFAs) conducted at the item level do not support their a priori structure due, in part, to the overly restrictive CFA assumptions. We demonstrate that exploratory structural equation modeling (ESEM), an integration of CFA and exploratory factor analysis (EFA), overcomes these problems with responses (N = 3,390) to the 60-item NEO–Five-Factor Inventory: (a) ESEM fits the data better and results in substantially more differentiated (less correlated) factors than does CFA; (b) tests of gender invariance with the 13-model ESEM taxonomy of full measurement invariance of factor loadings, factor variances–covariances, item uniquenesses, correlated uniquenesses, item intercepts, differential item functioning, and latent means show that women score higher on all NEO Big Five factors; (c) longitudinal analyses support measurement invariance over time and the maturity principle (decreases in Neuroticism and increases in Agreeableness, Openness, and Conscientiousness). Using ESEM, we addressed substantively important questions with broad applicability to personality research that could not be appropriately addressed with the traditional approaches of either EFA or CFA.

Thomas, Michael L.; Locke, Dona E. C.

doi: 10.1037/a0019229pmid: 20822262

The MMPI-2 Restructured Form (MMPI-2-RF; Tellegen & Ben-Porath, 2008) was designed to be psychometrically superior to its MMPI-2 counterpart. However, the test has yet to be extensively evaluated in diverse clinical settings. The purpose of this study was to examine the psychometric properties of the MMPI-2-RF Somatic Complaints (RC1) scale in a clinically relevant population. Participants were 399 patients diagnosed with either epilepsy or psychogenic nonepileptic seizures on the basis of video–electroencephalograph monitoring. The internal structure of the MMPI-2-RF was evaluated using taxometric, confirmatory factor analysis, and item response theory procedures. Data from 4 content-specific scales directly related to RC1 (Malaise, Gastrointestinal Complaints, Head Pain Complaints, and Neurological Complaints) indicated that the latent construct of somatization is a dimensional variable with a bifactor structure. However, consistent with the scale's construction, a unidimensional model also provided adequate fit. A 2-parameter logistic item response theory model better accounted for observed item responses than did 1- or 3-parameter models. Results suggest that the RC1 scale is most precise for T score estimates between 55 and 90. Overall, the scale appears to be well suited for the assessment of somatization.

Ketterer, Holly L.; Han, Kyunghee; Hur, Jaehong; Moon, Kyungjoo

doi: 10.1037/a0019511pmid: 20822263

In response to the concern that Minnesota Multiphasic Personality Inventory-2 (MMPI-2; J. N. Butcher, W. Dahlstrom, J. R. Graham, A. Tellegen, & B. Kaemmer, 1989; J. N. Butcher et al., 2001) Variable Response Inconsistency (VRIN) and True Response Inconsistency (TRIN) score invalidity criteria recommended for use with American samples results in an excessive number of exclusions in Asian samples (F. M. Cheung, W. Z. Song, & J. X. Zhang, 1996), we examined the cross cultural equivalence of the original VRIN and TRIN scales, and developed and validated Korean-specific VRIN and TRIN scales with Korean adult normative, clinical, and college samples. Although the results from item pair correlation analyses suggested the superiority of the Korean VRIN and TRIN over the original VRIN and TRIN, the mean comparison results and classification accuracy statistics using data with varying degrees of randomly inserted true and/or false responses did not reveal a strong advantage of one version over the other. We present and discuss plausible causes of the findings.

Noone, Jack H.; Stephens, Christine; Alpass, Fiona

doi: 10.1037/a0019512pmid: 20822264

Although a substantial proportion of the western population is approaching retirement age, little is known about how they are preparing for the future. Much attention has been paid to the consumption of educational material and retirement wealth in the present literature, but the process of retirement planning has been ignored. S. L. Friedman and E. K. Scholnick's (1997) theoretical model provided the basis for a comprehensive measure of retirement planning. According to their process theory, individuals develop an understanding of the problem, set goals, make a decision to start preparing, and finally undertake the behaviors needed to fulfill their goals. Fifty-two items were developed to assess each stage of the planning process for financial, health, lifestyle, and psychosocial retirement planning. These were tested on a population sample of 1,449 New Zealanders aged 49–60. Confirmatory factor analysis, bivariate correlations, and hierarchical regression provided support for the valid use of the measure. Necessary antecedents, such as the tendency to look to the future, and locus of control were significantly related to the Process of Retirement Planning Scale (PRePS). The PRePS also outperformed retirement planning measures used in the Health and Retirement Study (F. T. Juster & R. Suzman, 1995) after controlling for socioeconomic and psychological variables. This measure will enable social policy makers to determine which stages of retirement planning require support and intervention. The PRePS will also help to determine which domains of retirement planning predict well-being in later life and the factors which differentiate those who are planning from those who are not.

Cheng, Sheung-Tak; Li, Kin-Kit

doi: 10.1037/a0019579pmid: 20822265

Major life events and hassles have been considered 2 distinct constructs in the measurement of stress. Research also shows that chronic events are more impactful than time-limited ones. This study reports a new approach to measuring stress in which major life events are combined with recurrent hassles to form a single index—the Adolescent Stress Index (ASI). High school students (N = 365) in Hong Kong responded twice at a 3-month interval to measures of major life events and hassles, the ASI, and measures of physical and depressive symptoms. Data were analyzed using structural equation modeling. Results showed that the ASI predicted symptoms concurrently and prospectively above and beyond the effects of existing measures of major life events and hassles. The ASI is a viable instrument for documenting the cumulative impact of major and minor events in the lives of adolescents.

Samuel, Douglas B.; Ansell, Emily B.; Hopwood, Christopher J.; Morey, Leslie C.; Markowitz, John C.; Skodol, Andrew E.; Grilo, Carlos M.

doi: 10.1037/a0019580pmid: 20822266

Many personality assessment inventories provide gender-specific norms to allow comparison of an individual's standing relative to others of the same gender. In some cases, this means that an identical raw score produces standardized scores that differ notably depending on whether the respondent is male or female. Thus, an important question is whether unisex-normed scores or gender-normed scores more validly assess personality. Gender-normed and unisex-normed scores from the NEO Personality Inventory—Revised (P. T. Costa & R. R. McCrae, 1992) were examined in a large clinical sample, using 2 measures of personality disorder as validating criteria. Gender-normed scores did not obtain significantly higher correlations. In fact, for 2 personality disorders (antisocial and narcissistic), gender-normed scores yielded significantly lower correlations, suggesting that personality disorder pathology relates most closely to one's absolute level of a personality trait, rather than one's standing relative to others of the same gender. Ramifications of this finding for personality research and clinical assessment are discussed.

Purpura, David J.; Wilson, Shauna B.; Lonigan, Christopher J.

doi: 10.1037/a0019581pmid: 20822267

Clear and empirically supported diagnostic symptoms are important for proper diagnosis and treatment of psychological disorders. Unfortunately, the symptoms of many disorders presented in the Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev.; DSM–IV–TR; American Psychiatric Association, 2000) lack sufficient psychometric evaluation. In this study, an item response theory (IRT) analysis was applied to ratings of the 18 attention-deficit/hyperactivity disorder (ADHD) symptoms in 268 preschool children. Children (55% boys, 45% girls) in this sample ranged in age from 37 to 74 months; 80.4% were identified as African American, 15.1% as Caucasian, and 4.5% as other ethnicity. Dichotomous and polytomous scoring methods for rating ADHD symptoms were compared and psychometric properties of these symptoms were calculated. Symptom-level analyses revealed that, in general, the current symptoms provided useful information in diagnosing ADHD in preschool children; however, several symptoms provided redundant information and should be examined further.

Witt, Edward A.; Hopwood, Christopher J.; Morey, Leslie C.; Markowitz, John C.; McGlashan, Thomas H.; Grilo, Carlos M.; Sanislow, Charles A.; Shea, M. Tracie; Skodol, Andrew E.; Gunderson, John G.; Donnellan, M. Brent

Kennealy, Patrick J.; Skeem, Jennifer L.; Walters, Glenn D.; Camp, Jacqueline

doi: 10.1037/a0019618pmid: 20822269

The utility of psychopathy measures in predicting violence is largely explained by their assessment of social deviance (e.g., antisocial behavior; disinhibition). A key question is whether social deviance interacts with the core interpersonal-affective traits of psychopathy to predict violence. Do core psychopathic traits multiply the (already high) risk of violence among disinhibited individuals with a dense history of misbehavior? This meta-analysis of 32 effect sizes (N = 10,555) tested whether an interaction between the Psychopathy Checklist-Revised (PCL-R; R. D. Hare, 2003) Interpersonal-Affective and Social Deviance scales predicted violence beyond the simple additive effects of each scale. Results indicate that Social Deviance is more uniquely predictive of violence (d = .40) than Interpersonal-Affective traits (d = .11), and these two scales do not interact (d = .00) to increase power in predicting violence. In fact, Social Deviance alone would predict better than the Interpersonal-Affective scale and any interaction in 81% and 96% of studies, respectively. These findings have fundamental practical implications for risk assessment and theoretical implications for some conceptualizations of psychopathy.

de Kwaadsteniet, Leontien; Hagmayer, York; Krol, Nicole P. C. M.; Witteman, Cilia L. M.

doi: 10.1037/a0019696pmid: 20822270

An important reason to choose an intervention to treat psychological problems of clients is the expectation that the intervention will be effective in alleviating the problems. The authors investigated whether clinicians base their ratings of the effectiveness of interventions on models that they construct representing the factors causing and maintaining a client's problems. Forty clinical child psychologists drew causal models and rank ordered interventions according to their expected effectiveness for 2 cases. The authors found that different clinicians constructed different causal models for the same client. Also, the authors found low to moderate agreement about the effectiveness of different interventions. Nevertheless, the authors could predict clinicians' ratings of effectiveness from their individual causal models.