Over the past 20 years researchers and health care practitioners have come to realize in addition to high prevalence rates, individuals with co-occurring disorders did not represent a homogeneous group (Drake, et al., 1998: 2001; Lehman, et al., 1994: 2000; Mueser, et al., 2000). It is essential to consider the heterogeneity of co-occurring disorders when considering new treatment modalities. Thus, it becomes pivotal to identify these differences for treatment approaches and program goals. Research shows that heterogeneity of treatment populations can be reduced through empirically-derived homogeneous groups based on multivariate analysis (Ries, et al., 1993; Lehman et al., 2000; Mueser, et al., 2000). The purpose of the current study was to address a significant void in knowledge on the heterogeneity of co-occurring disorders by determining if homogeneous subgroups exist within an outpatient population presenting for treatment and if so how many groups exist and what makes up group membership. Identification of subgroups can provide a mechanism to better understand the interrelationships between determinants that contribute to the etiology and problem severity at an individual and group level. Secondly, in an effort to improve service delivery, empirically-derived subgroups hold important clinical implications for treatment models. The exploratory research was conducted through a retrospective analysis seeking a parsimonious model of subgroups made up of individuals with co-occurring disorders entering an outpatient program using a latent class analysis (LCA). The best fitting statistical model in the latent class analysis was one in which the overall sample was composed of three (3) subgroups. The three-class model that included alcohol use, illegal drug use, education level and serious depression was identified as best fitting the data.

Late-life depression is a significant public and geriatric mental health concern and one of the most prevalent and common emotional disorders for all older United States citizens. This study examined how relationships with close friends and close relatives affected depressive symptomatology among African American and White older adults and explored how health, social and religious factors modified that relationship. The sample consisted of participants from the New Haven, CT cohort of the population-based longitudinal study, The National Institute on Aging project entitled “The Established Populations for Epidemiologic Studies of the Elderly” [EPESE]. Respondents were ages 65-75 and older (mean age = 72.3), of which approximately 82% were White and other Non-African American, 19% all African American and 2.2% other than White (N = 2,812). A linear multiple hierarchical regression analysis was conducted to test the association between the severity of depressive symptoms (Radloff, 1977; Engel and Schutt, 2005), close friendships and close relatives for community-dwelling older adults while controlling for demographic, health, social support and religious covariates. Data were analyzed using SPSS 16 for Windows with an assigned significance level of p = .05 (two-tailed). The results of the analysis of the correlation coefficients, (ΔR² of .002, F (2, 2324) = 3.185, p = .042) suggested our hypothesis to be true in that the relationship between social support and depression depended upon race. However, we specified that the relationship between social support provided by family and friends would each be stronger for African Americans than Whites. Contrary to this, outcomes indicated the relationship provided by family members to be the same for both races while the relationship provided by friends to be stronger for Whites. These findings are substantial in fulfilling the request for evidence-based empirical research. Increased scientific research is needed in comprehending how specific social support and health-related factors might impact depressive symptomatology particularly among large samples of African American and White older adults.

This simple interrupted time-series quasi-experimental study examined the effectiveness of a restraint reduction policy implemented in order to reduce the use of physical restraint in a residential treatment facility for children in the southeastern United States. Aggregate data on monthly physical restraint episodes from the agency were analyzed over a period of 4 years. A 22-month period was used as the baseline and the succeeding 26 months- when the restraint reduction policy was implemented- was the intervention phase. A regular regression model, estimated using ordinary least squares (OLS), modeled the effect of the policy change, and autoregressive integrated moving average (ARIMA) models were used to represent the autocorrelation structure of the residuals from the regression model, in the data analysis. Two ARIMA models, an ARIMA (1,0,0) and an ARIMA (0,0,1), were used to model the autocorrelation structure of the residuals from the OLS regression. The convergence of findings from these models suggested that the results of the analysis of the time-series data from this study were robust in a statistical sense because both models led to the same conclusion. There was a statistically significant decrease of about 1 restraint per child each month when the new policy of verbal de-escalation was implemented during the treatment phase. The rate of physical restraint was reduced by 70% with the implementation of the restraint reduction policy in both models. Although a significant reduction in the rate of physical restraints was associated with the implementation of the restraint reduction policy, interrupted time-series designs such as that used in this study are not strong enough for making cause-effect inferences.

Much previous work has demonstrated the importance of parenting for adolescents’ psychosocial functioning. Barber, Stolz, and Olsen’s (2005) recent monograph contributed in organizing and specifying the voluminous literature on parent-child/adolescent literatures by documenting specific paths among three commonly-studied dimensions of parenting (i.e., support, psychological control, and behavioral control) and three salient adolescent outcomes (i.e., social initiative, depression, and antisocial behavior). They did not, however, explore the possible mechanisms that underlie the consistent parenting effects they and others have found. Theory suggests that the adolescent self would be a logical mediator of the effects of parenting. Further, measures of the self, such as self-esteem, have consistently been linked to various aspects of adolescents’ psychosocial functioning (e.g., particularly depression and antisocial behavior). The relatively limited empirical work that has tested self as a mediator of the relationship between parenting and adolescent outcomes has typically not been conducted multivariately (i.e., with multiple parenting and adolescent outcome variables in the same model) or across time. Therefore, through a systematic extension of the recent parenting model (Barber et al., 2005), the present study contributes to the literature by theoretically specifying and testing one key mechanism of parenting’s effects on adolescents’ psychosocial functioning. In particular, this study proposed that self-esteem partially mediates the effects of both parental support and parental psychological control on adolescent outcomes, in males and females. The tests of this theoretical extension were conducted with longitudinal data, in order to lend greater support to a causal model. Results indicated that self-derogation (the negative component of global self-esteem) was an appropriate measure of adolescent self. Self-derogation mediated the effect of parental psychological control on adolescent depression and antisocial behavior. In other words, for the youth in this data set, the reason parental psychological control was associated with adolescent problem behaviors one year later was the degree of the youths' self-derogation which resulted from parental psychological control. The effects of parental support and behavioral control were not mediated by self-derogation. Implications and suggestions for future research are discussed.

Co-parenting, the joint participation and responsibility of a set of adults in childrearing (Talbot & McHale, 2004), contributes to our understanding of how family systems influence child outcomes. Co-parenting is more proximal to parenting than many measures of marital relationships in that it focuses only on those dyadic processes that are specific to parenting (e.g., child-rearing agreement, support/undermining of parenting practices). Because of this focus, it is also a more flexible construct for thinking about non-traditional parenting arrangements, such as family foster care. Rooted in a family systems theoretical framework, the unique contribution of co-parenting may help us better understand the relationship among marital dynamics, individual parenting, and ultimately, foster child outcomes. Therefore, it is imperative that the social work researchers and practitioners promote awareness of, and attention to, co-parenting in foster parents as an important facet of improving the well-being and care of children in foster care. To date, co-parenting has not been studied in foster couples and no measure of foster co-parenting exists. The goal of this dissertation is to contribute to the knowledge base of foster couples and co-parenting by initiating the study of co-parenting in foster couples. Specifically, this study examined the psychometric properties of the Casey Foster Applicant Inventory-Applicant – Co-parenting Scale, an instrument developed to measure foster parents’ co-parenting potential. Results of the EFA and CFA strongly supported a reliable 10-item, one-factor Co-parenting construct for foster mothers and foster fathers that is distinct from, but related to, marital quality and parenting quality in expected ways.

This dissertation critically examines the influence that environmental racial composition has on the racial identification development of African Americans adolescents. Theoretical models that describe the formation of racial identity are examined chronologically, and research compiled within social work literature on the topic of racial identification is examined. Although previous studies have examined how African American adolescents over the age of 15 deal with the complexities of racial identity development, to date there is no literature that examines how middle school age African American adolescents navigate the same waters. The purpose of this study was to examine the relationship between environmental racial composition and racial identity attitudes of African American adolescents. Specifically, do African American adolescents that live in a predominantly African American community racially identify themselves differently than African American adolescents that live in a predominantly White community? To determine this, the Multidimensional Inventory of Black Identity (MIBI) was given to 101 African American adolescents. A little more than half (57%) of study participants lived in a predominantly African American community in SC and the remaining 43% of participants lived in a predominantly White community in TN. The results from the study suggested that racial environmental composition (location) influenced the racial identity concepts of study participants, as there were significant differences in racial identity attitudes between participants in SC and TN. Results also indicated that the concepts of Ideology, Regard, and Centrality that the MIBI examines may be too complex for adolescents under the age of 15, and a new factor, Self- Importance, was discovered and study participants from SC had higher Self-Importance scores than participants from TN even when controlling for demographic differences. Study participants from SC also had greater variability in their Self-Importance scores than participants in TN. Finally, an interaction was found between racial environmental composition and gender pertaining to Self-Importance. Girls in SC had significantly higher Self-Importance scores than girls in TN.

The purpose of this study was to investigate the extent to which the word “pet” provided valid results in the Lexington Attachment to Pets Scale (LAPS) (Johnson, Garrity, & Stallones, 1992). A multiple-groups confirmatory factor analysis, using maximum likelihood (ML) estimation, was conducted to test the hypothesized factor structures in the specified measurement model. Respondents to the Original version (i.e. used the word pet) and the Personalized version (i.e. embedded the canine’s name as typed by participants) made up the two groups. Features of an electronic survey were necessary for this personalization and systematic alternative assignment to the two LAPS versions. A snowball sampling method utilized electronic mail to invite self-selected participants meeting the following criteria: 18 years or older; lived in the United States; and at least one dog living inside/outside their home and for whom they are responsible at least some of the time. According to collected human and canine demographics, the Original (n = 1,854) and Personalized (n =1849) groups appeared to be statistically equivalent (N = 3,703). The hypothesized measurement model generated a χ2 value of 4130.242, with 264 degrees of freedom and a probability of less than .001 (p <.001), suggestive of a lack of fit. However, goodness-of-fit indices were a consideration. Comparing the CFI (.95 vs. .862), TLI (.95 vs. .840) and RMSEA (<.05 vs. .063) cut off values with the results from this study respectively, reveals again, an inadequate fit. These results imply that the hypothesized measurement model was not consistent with the data and precluded specific tests of differential validity. While the results of this study made it inadvisable to further examine the differential validity associated with linguistic differences in the scale items, critical information was nonetheless identified. According to the data in the current study, the LAPS conceptualization of “pet attachment” (sic) may not be valid. Until further research provides stronger evidence, use of this scale could produce results that lead to invalid inferences. The original LAPS conceptualization and quantification of human and canine relationships is still a work in progress. Moreover, psychometric work is critically necessary before using the LAPS.

Prejudice, discrimination and hate crimes against lesbians and gay men are prevalent throughout the United States. Prejudice in public school settings, as manifested in teacher attitudes and behavior, is particularly problematic for lesbian, gay, bisexual and questioning or queer (LGBQ) students and LGBQ parents. Efforts to reduce prejudice for sexual minority groups have met with limited success. Creating safer and more inclusive school environments is essential for the wellbeing and learning of all students. Furthermore, the field of social work is guided by principles that necessitate intervention research that addresses harassment and violence based on sexual orientation prejudice. This dissertation examines research on prejudice against LGBQ populations in public school settings, theories that seek to explain prejudice, and interventions to reduce prejudice. Intergroup dialogue is tested as one promising intervention social workers can use to improve relationships regarding polarizing topics such as sexual orientation, and to improve school culture and climate for LGBQ populations. An experimental design using quantitative and qualitative data was employed. The study used an established dialogue protocol in a community setting, with multiple outcome measures. Quantitative results indicate dialogue participation resulted in statistically significant positive changes in feelings toward LGB people. Qualitative data analysis revealed positive changes in attitudes, feelings, behaviors, critical self-reflection and perspective taking were all generated as a result of intergroup dialogue participation. Significance of the study in terms of recruiting participants for research on controversial topics, education and training for teachers on LGBQ issues, and intergroup dialogue work are discussed.

The Air Force Family Needs Screener (FNS) has been utilized as a screening measure for risk of both child maltreatment and intimate partner violence (IPV) within the context of a military primary prevention program called the New Parent Support Program (NPSP) since 1988. However, this measure has only been validated with the initial pilot samples used in its development. The current study sought to assess the reliability, dependability, factor structure and predictive capacity of the FNS using a new sample of mothers. In addition, this study sought to develop shortened versions of the FNS in order to determine if they performed as well as or better than the original version of the measure. Results indicated the original FNS has adequate reliability and dependability, but the cut-score currently used in practice resulted in a fairly low sensitivity rate and corresponding high false negative rate. Since there was no evidence of measurement invariance of a common items model tested across four different population groups, three shortened versions of the FNS made up only of common items to all four groups were also tested. All shortened models appeared to perform as well as the original FNS, suggesting the FNS could be shortened, while retaining its predictive capacity. A shortened model made up of only thirty-six common items was recommended as a revised version of the FNS for implementation with the Air Force. Although not the most parsimonious model, this model retained the highest rate of sensitivity while still improving the acceptability of the measure for staff and respondents alike.

The focus of this study was to determine whether aggression adds incremental validity above and beyond the big five personality factors in predicting academic success. An archival data analysis was used in this study. The data vas consisted of all of the eight grade students who were present when the data was collected in a middle school in the SoutheasteIl1 region of the United States. The students completed the Personal Style Inventory-Adolescent (PSI-A), which is a 120 item survey instrument designed to measure the big five personality factors and aggression. Results from this study indicated that aggression does in fact add incremental validity above and beyond the big five personality factors in a sample of middle school students. The results of the study also indicated that the big five personality factors are significantly correlated with academic performance. More specifically, conscientiousness, openness, agreeableness, and emotional stability are significantly correlated with grades in the current sample. When aggression is added into the statistical model, conscientiousness, openness, and aggression are significantly correlated with grades. The limitations, implications and conclusions are also discussed. 11