Factor analytic approaches to personality item-level data.

Factor analysis models have played a central role in formulating conceptual models in personality and personality assessment, as well as in empirical examinations of personality measurement instruments. Yet, the use of item-level data presents special problems for factor analysis, applications. In this article, we review recent developments in factor analysis that are appropriate for the type of item-level data often collected in personality. Included in this review are discussions of how these developments have been addressed in the context of two different (but formally related) statistical models item response theory (IRT: Hambleton, Swaminathan, & Rogers, 1991) and structural, equation modeling (Bollen 1989) for item-level data. We also discuss the relevance of item scaling in the context of these models. Using the restandardization data for the Minnesota Multiphasic Personality Inventory-2 Scale (cf. Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989), we show brief examples of the utility of these approaches to address basic questions about responses to personality scale items regarding: (a) scale, dimensionality and general item properties, (b) the "appropriateness" of the observed responses, and (c) differential item functioning across subsamples. implications for analyses of personality item-level data in the IRT and factor analytic traditions are discussed.

Incremental Fit Index Changes for Nested Structural Equation Models.

The assessment of structural equation model fit is quite critical for applications with finite samples. After reviewing the multiple fit indices in structural equation models, we present evidence on their behavior in simulation studies where sample size, estimation method, and model misspecification varies. Two sampling studies are reported with the first consisting of a confirmatory factor analysis model with a known population while the second utilizes an empirical data set whose population is only approximately known. Implications for the use of fit indices in the assessment of model fit are discussed.

Longitudinal patterns of formal and informal social support in an urban elderly population.

Stability and change in the use of formal and informal social support was assessed over a three year period among a representative sample of 1855 elderly urban participants in a longitudinal study of aging and health. Whether people received informal, formal, both types of support or no support was determined in baseline, 12 and 24 month personal interviews. Most respondents retained the same form of support across all three interviews. The extent of stability or change varied according to the form of support reported at baseline. Two-thirds of those with no social support continued without support and 40% of those using both informal and formal support continued to do so at subsequent assessments. The substitution of formal for informal support was infrequent and not statistically significant. Pair wise discriminant function analyses of groups of respondents with the most frequent longitudinal support patterns were performed to identify baseline health and social characteristics associated with stable use, the addition of another form of support or change to nonuse, over time. Changes in the use of support were influenced more by initial levels of health and functional status than by social and economic circumstances. Larger proportions of respondents dropped use of support then added an additional form of support. Respondents using both formal and informal support at baseline died across time in higher proportions than those in other support categories. The extent of stability in support use and nonuse, the character of changes in support use across time and greater attrition among those who were the heaviest users of support, serve to create a relative balance in the use of informal and formal support in this aging cohort, at least over the time frame over which respondents were followed in this study.

Unraveling the web of personality judgments: the influence of social networks on personality assessment.

Contemporary assessment models have focused on the degree to which self- and other reports of personality description agree in an effort to define consensus and agreement about personality attributes. In general, we believe that analyses of this type of data have been limited in that they tend to focus on both simple models (usually dyad-based) and simple aggregations of data (usually correlations between self- and other ratings). In addition, the behaviors used as stimuli in experimental settings lack the richness of behaviors in natural social settings. Here, we present some ideas from social network models in an effort to influence broader conceptualizations of agreement and consensus in assessment. Social network models provide a more complete description of interpersonal behavior beyond the dyadic level in both laboratory and natural settings. After defining some basic social network concepts, we go on to suggest the applicability of these concepts to personality assessment and, more specifically, to how these models might be used to study self-other agreement and consensus about personality judgments. Empirical data are used to illustrate social network concepts in the domain of personality assessment.

Childhood precursors of adolescent drug use: a longitudinal analysis.

Childhood and adolescent intrapersonal and interpersonal influences on older adolescents' drug use were examined with an interactional approach. Data were obtained when subjects were 5-10, 13-18, and 15-20 years old. Latent-variable confirmatory factor analysis was used to determine the fit of the variables to the latent constructs. Next, a structural model was hypothesized and tested to assess the influences of childhood and earlier adolescent factors on later adolescent drug use. Childhood aggression and parental sociopathy affected drug use in late adolescence. Unconventionality during early and middle adolescence had an important and pervasive impact on all aspects of middle and late adolescent functioning. Finally, both intrapersonal and interpersonal factors from childhood and early adolescence affected self-drug use during middle and late adolescence.

Characteristics of aggressors against women: testing a model using a national sample of college students.

Structural equation modeling was used to study the characteristics of college men (N = 2,652) who aggressed against women either sexually, nonsexually, or both. According to the model, hostile childhood experiences affect involvement in delinquency, leading to aggression through two paths: (a) hostile attitudes and personality, which result in coerciveness both in sexual and nonsexual interactions, and (b) sexual promiscuity, which, especially in interaction with hostility, produces sexual aggression. In addition, sexual and nonsexual coercion were hypothesized to share a common underlying factor. Although its development was guided by integrating previous theory and research, the initial model was refined in half of the sample and later replicated in the second half. Overall, it fitted the data very well in both halves and in a separate replication with a sample for whom data were available about sexual but not about nonsexual aggression.

Predictors of physician nonadherence to chemotherapy regimens.

Physician nonadherence to cancer treatment regimens may diminish treatment efficacy and compromise clinical research. The influence of clinical, demographic, and psychosocial patient characteristics on physician adherence to breast cancer chemotherapy was investigated, as was the role of the clinician's attitudes concerning the chemotherapy. One hundred seven women recently diagnosed with breast cancer were followed for 26 weeks of treatment. Fifty-six (52%) of the patients experienced unjustified modification for at least one chemotherapeutic agent. Stepwise multiple regression revealed independent contributions of increased patient age, treatment setting (clinic versus academic or community private practice), and stage of disease to physician nonadherence. Regimen complexity, delay in seeking treatment, and presence of psychiatric disorder did not contribute, in general, to unjustified regimen modifications. Patient psychological and psychiatric factors, however, did influence prescribing behavior for vincristine. Physician awareness of factors contributing to unnecessary treatment modifications may reduce the frequency of such behaviors.

Patient noncompliance with self-administered chemotherapy.

The contribution of patient factors to nonadherence with self-administered cancer chemotherapy along with its prevalence was studied. Fifty-one patients with breast cancer enrolled in protocols that included the orally administered medications Cytoxan (cyclophosphamide) and/or prednisone were interviewed and assessed at five points in time over a 6-month period. Two summary measures of patient nonadherence derived by self-report were developed: (1) dosage, the overall percentage of drug missed during 26 weeks of treatment; and (2) behavioral, the percentage of behavioral events or prescribing occasions on which a criterion level of drug was missed. Twenty-two patients (43%) met criteria for noncompliance according to both behavioral and dosage definitions. Univariate analyses showed more nonadherence in the clinic and private community settings than in the academic setting. Stepwise multiple logistic regression analyses assessed the contribution of patient demographic, psychologic, and physical symptom factors on patient noncompliance. In the regression analyses patient noncompliance was associated with (1) treatment location, more nonadherence in the private community-based treatment sector than in the academic setting; and (2) income, more nonadherence among those having lower incomes. In assessing total dose requirements in clinical research trials, rates of patient nonadherence need to be considered, and treatment location controlled.

Dimensions of the Need for Cognition: Subscales and Gender Differences.

The Need for Cognition (Cacioppo & Petty, 1982) scale has been used, predominantly in the persuasion literature, to assess the degree to which individuals tend to engage in elaborative thought. Using both statistically appropriate methods for the factor analysis of dichotomous variables and latent variable models, we report studies which discuss the development and validation of subscales of the NFC. The findings suggest that three subscales (Cognitive Persistence, Cognitive Confidence, and Cognitive Complexity), representing three different domains of elaborative thought, can be identified. A gender difference was obtained consistently on the Cognitive Persistence Scale: Women scored higher than men. Finally, using latent variable modeling methodologies, we present some initial exploratory evidence regarding cohort differences. We find no statistically significant relation between years of education and NFC subscale scores. Implications of the subscales and potential correlates of the gender difference found for Cognitive Persistence are discussed.

Assessing the Stability of Depression in College Students.

Two different approaches are used to assess the one-month stability of depressive affect in college students. First, a high retest correlation is demonstrated for a latent depressive affect construct using self-reports from the previous month on the Beck Depression Inventory, the Center for Epidemiological Studies-Depression Scale, and Zung's Self-Rating Depression Scale. This is true although the corresponding stabilities for individual scales are only moderate. Second, the predictive validity of depressive categorization based on these scales is examined using logistic regression techniques. Depressive categorizations are best predicted from the corresponding depressive categorization on the same scale at Time 1. There is little additional across-scale predictive information in the assessment of depressive categorization; this again suggests that a single factor model is most appropriate. Further, the association between assessments of dysphoric affect over one month in college students is less than previously established longitudinal associations between dysphoric affect categorizations over one year in older adult samples. This suggests that college students may be more affectively volatile than adults.

Prognostic validity of the Rorschach.

Fifty adult stutterers entering therapy at the UCLA Psychology Clinic were administered the Rorschach, with a Klopfer method inquiry, scoring, form level rating, and calculation of scores on the Rorschach Prognostic Rating Scale (RPRS). On the basis of independent clinician ratings of attitudinal or psychotherapeutic improvement, subjects were divided into groups of Improved Most (n = 21) and Improved Least (n = 29). Subjects were also divided into Continued (n = 43) and Dropped (n = 7). Logistic regression was employed to compare groups on the following Rorschach dimensions: Prognostic Score (RPRS); Human Movement; Animal Movement; Inanimate Movement; Shading; Color; Form Level. The Improved Most group was significantly higher in M, FM, Shading, and Productivity. The finding that M and FM discriminates between improvement groups corroborates results obtained in a previous study (Sheehan et al., 1954). The Rorschach movement variables, particularly M and FM, seem to be stable indicators of capacity for improvement in psychotherapy.

Factor Invariance Of Premorbid Social Competence Across Multiple Population Of Schizophrenics.

The social competence construct is important both in diagnosis and prognosis in schizophrenia. Empirical and theoretical work has examined the dimensionality of the construct as well as its invariance across different patient populations. Some recently published data by Zigler and Levine (1981) are re-analyzed through a simultaneous test of the factor structure of the Zigler-Phillips social competence scale across multiple samples. An invariant two-factor structure is established for groups varying as a function of patient gender and hospital type. Factor means differ across groups in a pattern consistent with previous evidence. Substantive interpretations of these results are discussed.

The evaluation and selection of adequate causal models: a compensatory education example.

Procedures for ascertaining relative model adequacy in latent variable structural relations models are discussed. Under diverse methods of estimation, this determination may be assessed using the chi square goodness of fit statistic, incremental fit indices for covariance structure models, and latent variable coefficients of determination. An example from evaluation research is taken (cf. Magidson, 1977; Bentler & Woodward, 1978). Numerical sensitivity of parameter estimates under alternative model specifications is demonstrated. Interpretive implications based on these procedures are discussed in terms of parameter sensitivity to alternative model specifications.