Epidemiological studies have demonstrated that the prevalence of psychotic disorders is exceeded by that of sub-threshold psychotic experiences, which are phenomenologically similar to threshold psychosis but of less intensity or associated impairment. Recent research has highlighted the potential clinical significance of psychotic experiences with regards to psychological distress, service utilization, psychiatric comorbidities, and suicide risk. The aims of this three paper dissertation are to: 1) determine risk for suicidal behavior among respondents with psychotic experiences; 2) examine the prevalence of psychotic experiences among respondents with common mental disorders, and describe the clinical significance of these symptoms when occurring in the context of common mental disorders; and 3) evaluate factors associated with the persistence or remission of psychotic experiences in the general population. For all three papers, data were drawn from the Collaborative Psychiatric Epidemiology Surveys (n=20,013), composed of the National Comorbidity Survey-Replication, National Latino and Asian American Study, and National Survey of American Life. Psychotic experiences and other clinical variables were assessed using the World Health Organization Composite International Diagnostic Interview, version 3.0. Analyses consisted primarily of logistic regression models, with effect sizes calculated as adjusted odds ratios. Psychotic experiences were found to be associated with elevated risk for suicidal ideation and suicide attempts, and with multiple co-morbidities with common mental health conditions. The persistence of psychotic experiences over time was primarily associated with the type of symptom experienced (i.e. hearing voices) and with marital status. Co-morbid mental health conditions, although extensive, did not predict the persistence of psychotic experiences, although persistent psychotic experiences were associated with ongoing suicide risk. Together, these data support the clinical significance of sub-threshold psychotic experiences among a large general population sample of adults in the United States. The most clinically notable features of psychotic experiences are that they indicate drastically elevated risk for suicide attempts (particularly severe attempts with intent to die) and the presence of multiple co-morbid mental health conditions. These findings will have clinical utility in highlighting unique needs of individuals with sub-threshold psychotic symptoms, and will have public health value in identifying a significant risk factor for severe suicidal behavior that may be easily screened in the general population as well as in clinical settings.

This study used thematic narrative analysis to address the question: How do older adults who experience serious ongoing symptoms of schizophrenia understand and express stories of their personal survivorship in the face of life-course and present-time adversities? Framed by the developmental life course perspective and using major constructs of the theory of cumulative adversity and advantage to formulate a line of semi-structured questioning for narrative interviews about the life course experiences of older adults with schizophrenia who experienced ongoing illness symptoms, analysis of 31 interviews with six older adults with schizophrenia yielded findings across five central areas. Shared core themes included: 1) "My life as it is has value:" Narrating schizophrenia in later life; 2) "I have a key and live like a real person:" Homelessness and housing challenges in retrospect; 3) "There's not been jobs:" The meaning of employment; 4) "God told me how we're going to meet back up:" Narratives of relational conflict and loss, adjustment and renewal. A fifth area of findings developed the theory of cumulative adversity and advantage across the life course with schizophrenia. These results improve our understanding of the subjective experience of a highly vulnerable but grossly understudied and underserved population. Recommendations for focusing future research and development of more effective social work practice and policies are made.

In this dissertation I examine the Child Tax Credit (CTC), who gets it and who doesn't, paying particular attention to children under the age of three, its legislative and political history, and how it could be improved. At $56.4 billion per year, the Child Tax Credit (CTC) is nearly the largest U.S. federal expenditure on children and families, second only to the Earned Income Tax Credit (EITC), at $59.5 billion (JCT, 2011). Created in 1997, it has been expanded seven times in just the last decade. Yet in spite of America's federal commitment of dollars and legislative commitment of reform, little has been written about the CTC. I examine the literature first to see if cash and cash assistance matter, finding on balance that there is strong evidence that it does, particularly for young children; second to show that the U.S. underinvests in this domain in young children relative both to what is needed and to what other advanced industrialized countries do; and third to lay out the case that changes to the refundable CTC offer one opportunity to address this underinvestment. I examine the legislative history of the CTC, as I believe both the policy analysis and history need to be understood to inform the policy responses. Next, I examine whether the portion of the new safety net that was fashioned as tax policy is working as child policy - specifically, whether it is reaching our youngest children, with initial evidence that it may not be in the case of the CTC (Burman and Wheaton, 2007) yet may be in the case of the EITC (Dowd and Horowitz, 2011). Using the 2011 Current Population March Supplement, I examine empirical evidence of the age distribution of federal tax credits for children, finding that 29% of children under the age of three are in families with too little earnings to get the full CTC, as opposed to 20% of older children. Nearly 13% of children under the age of three are in families with no earnings and as such get no CTC or EITC, as opposed to 8% of older children. While the EITC may disproportionately benefit young children, poor young children are more likely to be left out eligibility of the EITC than their older counterparts. Since infants may or may not be eligible for any CTC or EITC, depending on their birth month, I suggest that as some have found a marriage penalty in parts of the tax code, that there may also be a "baby penalty." I use micro-simulation to examine the costs and benefits of alternative CTC policies. Here I find that while full refundability may be the optimum CTC policy, that there are other possibilities, including those that increase the phase-in of eligibility, that are less costly, and also substantially lower child poverty among young children, including doubling the CTC for young children, increasing the phase in, and using a look back provision to allow families to use their previous year's earnings to calculate their refundable CTC and EITC. Yet, only moving to full refundability would do anything for the 12.67% of young children in families with no earnings. Finally, I propose policy responses that are rooted both in the science of increased cash investments in young children, and in the politics of working legislatively to get there, suggesting that policy makers consider the question of age equity when examining the distributional effects of tax policies. Implications for research and policy are discussed.

Through the lens of general strain theory (Agnew, 1992), this dissertation examined the associations between arrests and the strains experienced by women with co-occurring substance use and posttraumatic stress disorders. Much of the research that has been conducted to better understand the experiences of female offenders shows that women in the criminal justice system are disproportionately affected by emotional and economic struggles such as substance use, trauma, depression, lower levels of education, lower employment achievement, and limited social support when compared to women not involved in the criminal justice system (Bloom, Owen, Covington, 2002; Chesney-Lind & Pasko, 2004; Chesney-Lind & Pasko, 2013; Freeman, 2000; Hayword, Kravitz, Goldman, & James & Glaze, 2006; Salisbury & Van Voorhis, 2009; & Warren, Hurt, Loper, Bale, Friend, & Chauhan, 2002). This study tested for associations between women's arrest and their strain experiences of drug use, alcohol use, posttraumatic stress disorder, depression, education, employment, and social support using data from the National Institute on Drug Abuse's (NIDA) Clinical Trials Network (CTN), Protocol #15 titled "Women's treatment for trauma and substance use disorders: A randomized clinical trial." Based on current research and general strain theory, I hypothesized that increased strain would be associated with a higher likelihood of arrest. In this study, I found that increased arrest was associated with increased levels of education and employment achievement. Increased education and employment achievement were also associated with increased severity of arrest type when crimes were categorized by no arrest, substance-related arrest, non-violent arrest, and violent arrest. Finally, I found that increases in alcohol and posttraumatic stress strain over time were associated with a higher likelihood of subsequent arrest. The results of the first two analyses were contrary to my hypotheses. One conclusion that can be drawn is that neither education nor employment strain are related to the increased likelihood of arrest or type of arrest. Alternatively, these results may show that the education and employment achievement scales were not adequate measures of strain because neither of the scores incorporated measures for subjective feelings of education or employment strain. It is also possible that the levels of education and employment achievement were a source of strain because they represented a failure of sorts for women who had hoped to attain higher levels of education and employment, especially after committing to a treatment program aimed at supporting them as they attempted to change their lives and begin their recovery from substance abuse. Results of the third analysis were consistent with my hypotheses. These results provide two important areas of focus for social work clinicians, policy makers, and researchers in their attempts to reduce women's criminal justice involvement. If severity of alcohol and posttraumatic stress strain were addressed, and subsequently reduced in treatment programs, these results suggest that this would help reduce women's likelihood of arrest. All of these results call for further testing of what constitutes women's strains and the relationships between these women-specific strains and arrest.

This study explores the dialectic between the older, linear-mechanistic approach of the clinical-normative-individual-system model, and the newer, ecological-systems approach of the life model. Theoretical issues are outlined as they have unfolded during several decades. The principal independent variable is clinical-orientation of clinician-subjects with regard to degree of adherence to linear-mechanistic and/or ecological-systems approaches. Secondary independent variables are duration-severity and interpersonal-context of client problem/situations, described in four situational vignettes. Hypotheses predict positive correlations between measures of each independent variable, and degree of linear versus ecological weighting to clinicians' assessments and intervention plans for each vignette. Data were collected in 1980 from 152 second-year graduate students in casework and direct practice, who represented an initial pool of 1,007 students from fourteen CSWE-approved schools which provided unrestricted cooperation, through lists of eligible students. Three instruments were utilized: (a) An informational questionnaire inquired about students' willingness to participate, and characteristics which would enable the researcher to determine eligibility, and identify extraneous effects. (b) The second sought graded measures of subjects' adherence to specific theoretical principles of linear or ecological approaches. (c) The final instrument sought repeated measures of type of assessment and intervention plan (linear or ecological), formulated in response to systematically varied vignette conditions. Findings include several positive correlations between self-rated clinical-orientation and assessment measures, and fewer positive correlations between clinical-orientation and intervention measures. The interpersonally isolated client whose problem/situation is chronic tends to pull judgments in the direction of linear-mechanistically weighted assessments and intervention plans, regardless of clinician's orientation. Implications for teaching the ecological approach are explored. Use of the clinical-orientation instrument for student self-observation is suggested. Research implications include refining of instrumentation, and comparison of seasoned and student clinicians, to test empirical applicability of the ecological approach.

This study examines the relative impacts of three groups of factors on the labor force participation of Army wives with children under the age of six and the implications for Army family policy. The three groups of factors are: (1) individual, family and socioeconomic factors, (2) features of Army life (eg. relocations, separations), and (3) satisfaction with overall support for Army families and select Army programs. The goals of this secondary analysis of the 1987 Annual Survey of Army Families (ASAF) are twofold. One goal is to analyze the combined effects of being a mother of a young child, being a wife of a Service member in the Army, and participating in the labor force. The second goal is to examine the linkage between the findings of this research and Army family policy. Log linear analysis is used to develop models that most parsimoniously represent the key factors that effect the odds an Army wife with young children will be in the labor force. First, a hierarchical log linear model is employed to estimate and select explanatory variables of labor force participation to be tested in a logit model. Second, the logit model is used to determine the log odds or chances that a woman has of being in the labor force based on the function of explanatory variables. Certain effects appear to be consistently significant and make a larger contribution to the logit models tested overall. Among the effects that increase the odds of an Army wife with young children being in the labor force are: whether the woman is black, whether the couple is geographically separated, whether the wife experiences problems with her overall adaptation to Army life, and whether she is dissatisfied with the overall support she perceives from the Army. The study concludes that a broader focus on balancing the demands of work and family life, rather than focusing primarily on facilitating the Army wife's access to jobs, should result in better outcomes for the Army and for Army families.

This exploratory study examines 10 categories of mental health services for any differential impacts they may have upon 6 measures of state psychiatric hospital utilization. These data for 1984, 1985 and 1986 are studied for the 40 catchment areas in Massachusetts, while controlling for salient demographic characteristics derived from the 1980 United States Census. Also examined for its impact upon rates of utilization is the Brewster v. Dukakis Consent Decree. The auspices of community mental health services are additionally included for control purposes to ascertain if systems predominantly operated directly by state personnel have different rates of hospital utilization than community systems that rely primarily upon contracts with vendors. The data indicate that considerable variance in rates of state psychiatric hospital utilization can be accounted for by the funding variables, while inclusion of the other independent variables allows for even more variance to be explained.

This study explored the relationship between the severity and types of psychosocial stressors and three major mental disorders. The data were derived from the field trials of the third edition of the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM-III), in which over 12,000 patients from all over the country were evaluated by over 500 clinicians. Two hundred forty-seven patients with Major Depression and 247 with Schizophrenia were randomly selected for this study, along with all 112 patients given the diagnosis of Schizophreniform Disorder, a disorder similar to Schizophrenia except for its brief duration. The number of psychosocial stressors recorded by the evaluating clinician for each subject was examined, and each stressor was classified according to whether it represented an entrance into or exit from the social field of the subject, whether or not it was desirable, whether or not its occurrence had been under the control of the subject, the number of Life Change Units it entailed, and what area of the subject's life is affected. These variables were then compared across diagnostic groups, for individuals with and without associated Personality Disorders. In addition, for each diagnostic group, the relationship between the subjects' highest mean level of adaptive functioning and the mean severity of their psychosocial stressors was examined, using the multiaxial system of DSM-III. Major findings that replicated those reported in the literature include that a greater proportion of individuals with Major Depression were reported to have experienced a greater number of stressors, undesirable events, entrances, and uncontrollable events, than individuals with Schizophrenia. Significant new findings include that, for Schizophrenia, the highest level of adaptive functioning in the past year and level of severity of stressors experienced prior to episode onset are positively correlated, while for Major Depression these variables are negatively correlated. The results for Schizophreniform Disorder are equivocal, with similar results to Major Depression for some stressor dimensions, and midway between the other groups on others. The implications for social work practice of these findings and further study of life events are great, for primary, secondary, and tertiary prevention of mental illness.

This study addressed the political participation of Older Americans. Political participation was defined as all behavior through which people directly expressed their political opinions and ideologies. This definition was broad enough to cover all expressions of political opinions--from discussing politics to contacting public officials. The main concern of this study was to distinguish between the social roles and status of the 'Young Old' (ages 55-74) and the 'Old Old' (ages 75 and over) and their effect on political participation. Little is known about the precise political behavior of these two age groups. This research examined the level of their political activity through multiple regression and factor analysis. A secondary analysis of University of Michigan 1984 pre- and post- National Election Study data provided the evidence for this study. The principal findings of this study were the social characteristics that differentiate those 'Young Old' and 'Old Old' who were more likely to participate in political activities. For the 'Young Old' they were: being married, educated, paying attention to the political campaign of 1984, community residence, "feeling closest" to blacks, belonging to interest groups, and having a strong feeling of political efficacy. For the 'Old Old' also being married, educated, belonging to interest groups, "feeling closest" to environmentalists were significant. Identification with the elderly as a group, was not a significant predictor of participation for the 'Young Old' and the 'Old Old'. This relationship held for all types of participation. No evidence existed to indicate that strictly age-based interests were prevalent among the 1984 elderly respondents.

Critical incidents of parent coping with their provocative children were observed over eight interviews with 27 at-risk parents whose demographic profiles typically matched that associated with the so-called "feminization of poverty". Following the Lazarus stress-appraisal-to-coping paradigm, relationships between child provocativeness and parent cognitive appraisal of the situation were analyzed, and the relationship of each of these respective social and psychological levels of stress to actual coping behavior studied. The role of anger--an emotion often associated with abuse--was also examined in relation to these stress and coping variables. And, finally, the temporal order of these components of the coping process was analyzed. Adaptiveness of parent cognition and coping behavior varied with the stressfulness of the situation when this was defined as child provocativeness. There were indications that the positive aspects of child provocativeness, parent cognition, and parent coping behavior went together, with child provocativeness being dependent on parent cognition and behavior rather than the other way around. Thus, it was concluded that abuse should be viewed as a transactional encounter which, while immediately triggered by provocative child behavior, is also dependent on preceding parent behavior, and parent cognitions. The implications were for prevention and intervention efforts which foster more adaptive levels of both cognition and behavior in parents. While all relationships were not statistically significant, support was found for the primacy of cognition in coping: the temporal order which Lazarus posits, i.e., that cognition precedes emotion which precedes actual coping behavior, was supported. It was recommended that findings be interpreted cautiously, with consideration of the small size and heavily minority makeup of the sample. It was also recommended that additional sources of stress in the parent-child relationship, and related parent cognitions and coping responses be identified in research. The PCE study design and instruments were seen as appropriate models for such expanded study. It was emphasized that in follow up studies involving similar minority samples, increased consideration be given to measurement and interpretation in light of cultural reality. The correspondence of cognitive perspectives with social work values, goals, and daily work at the interface of person and environment was noted, and recommendations were made for helping students and practitioners make the needed cognitive shift toward integrating such perspectives in practice.