Jodie B. Ullman

Structural equation modeling: reviewing the basics and moving forward.

This tutorial begins with an overview of structural equation modeling (SEM) that includes the purpose and goals of the statistical analysis as well as terminology unique to this technique. I will focus on confirmatory factor analysis (CFA), a special type of SEM. After a general introduction, CFA is differentiated from exploratory factor analysis (EFA), and the advantages of CFA techniques are discussed. Following a brief overview, the process of modeling will be discussed and illustrated with an example using data from a HIV risk behavior evaluation of homeless adults (Stein & Nyamathi, 2000). Techniques for analysis of nonnormally distributed data as well as strategies for model modification are shown. The empirical example examines the structure of drug and alcohol use problem scales. Although these scales are not specific personality constructs, the concepts illustrated in this article directly correspond to those found when analyzing personality scales and inventories. Computer program syntax and output for the empirical example from a popular SEM program (EQS 6.1; Bentler, 2001) are included.

Classroom social experiences as predictors of academic performance.

A model linking childrens peer acceptance in the classroom to academic performance via academic self-concept and internalizing symptoms was tested in a longitudinal study. A sample of 248 children was followed from 4th to 6th grade, with data collected from different informants in each year of the study to reduce respondent bias. A path analysis supported the model; a lack of peer acceptance in the classroom in 4th grade predicted lower academic self-concept and more internalizing symptoms the following year, which in turn, predicted lower academic performance in 6th grade. An alternative path with internalizing symptoms predicting declines in peer acceptance was tested and received some support as well. Implications of the findings for schools are discussed.

Factor structure of the Penn State Worry Questionnaire: examination of a method factor.

The Penn State Worry Questionnaire (PSWQ) was originally designed as a unifactorial measure of pathological trait worry. However, recent studies supported a two-factor solution with positively worded items loading on the first factor and reverse-scored items loading on a second factor. The current study compared this two-factor model to a negative wording method factor solution among college students. A method factor model with all PSWQ items loading on a single worry factor and reverse-scored items loading on a negative wording method factor provided as good a fit as the two-factor model. This method factor alone did not predict a generalized anxiety disorder diagnosis. Finally, the psychometric properties of an abbreviated scale containing only positively worded items were examined. The PSWQ appears to measure a single unitary construct, but response patterns differ between positively worded and reverse-scored items. Theoretical implications for pathological worry and assessment-related issues are discussed.

Familism and psychological health: the intervening role of closeness and social support

Familism, a cultural value that emphasizes warm, close, supportive family relationships and that family be prioritized over self, has been associated with psychological health. The goal of this work was to fill a gap in the literature on how familism contributes to psychological health. Drawing from conceptual links between familism and close relationship processes, we hypothesized that familism contributes to better psychological health by facilitating closeness and social support. A university sample of U.S. women and men of Latino (n = 173), European (n = 257), and Asian (n = 642) cultural backgrounds completed measures of familism, closeness to family members, general perceived social support, and psychological health as indexed by perceived stress, general mental health, and depressive symptoms. Structural equation multiple-group modeling analyses found direct effects of familism on closeness to family members and perceived social support and an indirect effect of familism on better psychological health via greater closeness to family members and greater perceived social support. These effects did not differ by cultural background. Consistent with previous research, however, Latinos reported the highest levels of familism of the three cultural groups, and women reported higher familism and support as well as poorer psychological health than men. Discussion is focused on the implications of these findings for understanding the association of familism with psychological health and the relevance of the familism construct for diverse U.S. groups.

Cumulative Burden of Lifetime Adversities: Trauma and Mental Health in Low-SES African Americans and Latino/as

This study examined the utility of a lifetime cumulative adversities and trauma model in predicting the severity of mental health symptoms of depression, anxiety, and posttraumatic stress disorder. We also tested whether ethnicity and gender moderate the effects of this stress exposure construct on mental health using multigroup structural equation modeling. A sample of 500 low-socioeconomic status African American and Latino men and women with histories of adversities and trauma were recruited and assessed with a standard battery of self-report measures of stress and mental health. Multiple-group structural equation models indicated good overall model fit. As hypothesized, experiences of discrimination, childhood family adversities, childhood sexual abuse, other childhood trauma, and chronic stresses all loaded on the latent cumulative burden of adversities and trauma construct (CBAT). The CBAT stress exposure index in turn predicted the mental health status latent variable. Although there were several significant univariate ethnic and gender differences, and ethnic and gender differences were observed on several paths, there were no significant ethnic differences in the final model fit of the data. These findings highlight the deleterious consequences of cumulative stress and trauma for mental health and underscore a need to assess these constructs in selecting appropriate clinical interventions for reducing mental health disparities and improving human health.

Impact of marriage on HIV/AIDS risk behaviors among impoverished, at-risk couples: a multilevel latent variable approach.

Studies among normative samples generally demonstrate a positive impact of marriage on health behaviors and other related attitudes. In this study, we examine the impact of marriage on HIV/AIDS risk behaviors and attitudes among impoverished, highly stressed, homeless couples, many with severe substance abuse problems. A multilevel analysis of 368 high-risk sexually intimate married and unmarried heterosexual couples assessed individual and couple-level effects on social support, substance use problems, HIV/AIDS knowledge, perceived HIV/AIDS risk, needle-sharing, condom use, multiple sex partners, and HIV/AIDS testing. More variance was explained in the protective and risk variables by couple-level latent variable predictors than by individual latent variable predictors, although some gender effects were found (e.g., more alcohol problems among men). The couple-level variable of marriage predicted lower perceived risk, less deviant social support, and fewer sex partners but predicted more needle-sharing.

Trauma and Psychosocial Predictors of Substance Abuse in Women Impacted by HIV/AIDS

The purpose of this study was to estimate the relative contributions of trauma, chronic stress burden, depression, anxiety, social support, and social undermining in predicting alcohol and drug abuse, and whether ethnicity moderated these relationships. A multi-ethnic sample of 288 HIV-positive and HIV-negative women was recruited. Multiple group path analysis indicated that greater drug dependence was associated with being HIV+, more depression, and higher chronic burden. Trauma was related only to anxiety. Also, greater alcohol dependence was associated with more depression and more social undermining, and these effects were moderated by ethnicity. African American and Latina women evidenced different relationships between depression, social support and social undermining. Depression, social support and social undermining served as intervening variables in influencing the relationships between the other psychosocial variables and drug and alcohol dependence. The implications of these findings for alcohol and drug abuse research and services are discussed.

Spontaneous Cognition and HIV Risk Behavior

Theories of cognitive processes and risk behavior have not usually addressed spontaneous forms of cognition that may co-occur with, or possibly influence, behavior. This study evaluated whether measures of spontaneous cognition independently predict HIV risk behavior tendencies. Whereas a trait-centered theory suggests that spontaneous cognitions are a by-product of personality, a cognitive view hypothesizes that spontaneous cognitions should predict behavior independently of personality. The results revealed that spontaneous cognition was an independent predictor of behavior tendencies in cross-sectional analyses. Its predictive effect was stronger than drug use, a frequently emphasized correlate of HIV risk behavior in the literature, and comparable with sensation seeking in magnitude. The results suggested that a relatively spontaneous form of cognition may affect HIV risk behavior.

Violence prevention among HIV-positive women with histories of violence: healing women in their communities.

Experiences of past and current gender-based violence are common among HIV-positive women in the United States, who are predominantly from ethnic minority groups. However, culturally congruent, feasible interventions for HIV-positive women who have experienced past and/or current violence are not widely available. The Office on Womens Health Gender Forum has made several recommendations for responding to the National HIV/AIDS Strategy Implementation Plan, including recommendations to incorporate gender-based violence prevention into a comprehensive, gender-responsive national strategy. This paper draws on an example of a community-based project for HIV-positive women, the Healing Our Women Project, to illustrate how violence prevention can be achieved within peer-led and community-based programming. Strong community partnerships, responsiveness to community needs and local cultural norms, a trained workforce, and culturally competent care are programmatic cornerstones of gender-responsive services. HIV-positive women with histories of gender-based violence and risk factors for current and future violence deserve the highest quality gender-responsive services to ensure that they can address their health needs within contexts of safety and respect.

INGROUP REJECTION AMONG WOMEN: THE ROLE OF PERSONAL INADEQUACY

We examined predictors and outcomes of womens hostility toward other women. Based on a projection model, we hypothesized and tested the theory via structural equation modeling that womens sense of personal inadequacy, the tendency to stereotype, and general anger would predict hostility toward women, and hostility toward women would predict blaming victims of violence and poor relationships with a female friend. Participants were 464 college women with an average age of 28.08. All measures were pencil and paper. Womens hostility toward women served as an intervening variable between a personal sense of inadequacy, tendency to stereotype, blaming women victims of violence, and intimacy with ones best female friend.