Previous studies consistently find noted racial differences in stress and perceived social support, as well as evidence of a buffering effect of social support against stress in pregnancy. What has not been established, however, is whether this effect is equal for all women. Data from the Pregnancy Risk Assessment Monitoring System (PRAMS) are used to examine race differences in the moderation effect of social support on stress with two outcomes: birthweight and premature delivery. Consistent with previous findings, the results indicate that while more stressful life events did not significantly increase risk for low birthweight or prematurity directly, social support did act as a buffer. Furthermore, there was a moderately significant black-white difference in this buffering effect. Although greater social support significantly reduces negative effects of stress for white women, this effect is not seen for black women, suggesting that social support is not equally protective for black and white women.

In this dissertation, I examine the connection between fatherhood, work, social policy, and shifting ideals of masculinity in the United States and Canada. Drawing from 85 semi-structured interviews with fathers and their female partners, I argue that fathers today draw on newer forms of masculinity in how they think about and attempt to do fathering, but that structural factors supporting more traditional forms of masculinity allow them to fall back into less involved fathering. I was granted access to a multinational financial services firm and drew my sample of fathers from within this single firm. By sampling from within the same firm, I was able to control for some of the work-related variability that previous research has shown to influence father involvement. Each of the four substantive chapters focuses on one such structural factor, highlighting how it enables or limits father involvement. These factors are: the availability and use of paternity leave, the presence and involvement of extended kin in child care, the organization of child care, and the geographic separation of work and family life.

Drawing from the previous studies, I distinguish three components of social integration, including network integration (size, density, centrality), behavioral integration (frequency of interaction), and affective/cognitive integration (strength of positive thoughts and feelings about others). I propose a conceptual model and argue that these components of social integration, both at individual and organizational levels, should promote mental health. I extend previous studies, which have focused on the adult population, by focusing on the adolescent population. Specifically, I examine the effects of adolescent friendships on depressive symptoms. Using the National Longitudinal Study of Adolescent Health (N=11,023 adolescents at 105 schools), I demonstrate that greater degrees of network and behavioral integration are generally associated with fewer depressive symptoms, as expected. However, the relationships are weak, and many integration indicators do not have significant effects on depressive symptoms, after controlling for egocentric network sizethe most efficient measure of network integration. The effect of egocentric network size is mediated by affective/cognitive integration, consistent with the assumption frequently made in the literature that network and behavioral integration contribute to mental health by providing a sense of belonging, security, and comfort. The results also show that having friends who share the same socio-demographic characteristics strengthens a sense of belonging to school. However, this relationship varies across socio-demographic groups, and it also depends on student distributions at school, suggesting that adolescents in different groups and school contexts have unique motivations to become associated with similar others and derive different meaning from their friendships.

SOCIOLOGY<p> CHINESE AMERICAN FEMALE IDENTITY<p> JANELLE LEE WOO<p> Dissertation under the direction of Professor Peggy A. Thoits<p> This study examines the racial-ethnic identity formation of Chinese American women and assesses the utility of existing racial-ethnic identity development models and various social-psychological theories of identity formation. The social-psychological literature emphasizes the broader concepts of social self or social identity; these are socially constructed categorizations of self that reflect individual-level or collective-level identifications of the self with a membership group. A sub-field of social-psychology includes theories and models of racial-ethnic identity development. Unlike the broader social-psychological theories, ethnic identity theories stress the presence of racial-ethnic power dynamics that shape the course of racial-ethnic identity formation and its content. I explore whether Chinese American womens identity experiences are better described by developmental models or broader social-psychological approaches to identity.<p> To accomplish this, I used in-depth interviews with 30 women attending a public university in southern California. Through this analysis, I explored whether Chinese American women emulated the stages as suggested by ethnic identity development models and/or whether they more actively construct and negotiate their identities as suggested by broader social-psychological theories.<p> The interviews revealed that these women varied in their degree of ethnic identification. In addition, ethnic identity development models did not seem to apply well to these women. It appeared that broader social-psychological and sociological theories better described their identity formation. The respondents indicated that the imagined perceptions of others and/or their sense of shared group identity affected their views of racial-ethnic self more so than an unfolding series of steps from identification with the dominant group to identification with their own group. While some women followed the developmental path, the ethnic identity formation of the majority of the women involved a less sequential and internal process of discovery. In sum, these womens racial-ethnic identifications were highly contextually dependent and much more complex in their formation than suggested by developmental models, consistent with symbolic interactionist and self-categorization theories in sociology and psychology.<p> Approved_______________________________________Date_____________________

In this dissertation, I argue that cancer support groups may be usefully reconceptualized as subcultures with three distinct characteristics: ideology, norms and coping assistance. Ideology offers participants an alternative set of beliefs about cancer based on personal experience and intuitive understanding. Norms outline the range of suitable thoughts, feelings and behaviors for the specific situation as defined by the ideology. Together ideology and norms influence the types and extent of coping assistance offered by the groups to participants in need of support. I observed six different cancer support groups over the course of two years in order to document supportive transactions as they unfolded organically. By observing coping assistance in a naturalistic setting, I was able to focus on the normative and ideological influences on the types of supportive strategies offered by the groups rarely studied before.

Emotion management in the workplacethe control and expression of emotions in interactions with others―is required in many types of jobs. For example, in encounters with customers and clients, many workers must act cheerful and friendly, regardless of their true feelings. In addition, within the workplace, workers manage their emotions in order to act appropriately in interactions with coworkers and superiors. However, by concealing their true feelings and expressing false outward emotions, workers may suffer from feelings of inauthenticity and, in turn, experience psychological distress. Previous research suggests that these consequences vary by the workers status in the workplace hierarchy, occupation, self-concept, and attitude toward the emotion management that she performs. Using quantitative data from a mail survey of a random sample of 2,500 Tennessee state workers, I identify the conditions under which workers perform emotion management, the link between emotion management and psychological outcomes, how workers of varying statuses are differentially affected by emotion management, and the factors that affect the relationship between emotion management and distress. Whereas previous research has focused on occupational-level characteristics (e.g., service vs. non-service occupation), I find that job-level characteristics such as control over work, job complexity, and the amount of interaction a worker has with other people better predict the extent of a workers emotion management. In addition, workers who have higher status within the workplace perform less emotion management than those with lower levels of workplace status. Although the effects of emotion management at work are moderated by several factors―including a workers occupational prestige and self-concept orientation―my analyses reveal that, in general, emotion management has harmful consequences for the psychological well-being of workers because it increases feelings of inauthenticity and distress.

A key motivation in studying social movements is the belief by researchers that the characteristics, tactics and resources available to organizations within a movement facilitate or hinder the movements success. Relative to other processes within a movementsuch as emergence, participation, and recruitmenthowever, systematic research focusing on outcomes has only recently begun. This dissertation advances the study of social movement outcomes by providing a quantitative, historical examination of the influence of national resources on U.S. state woman suffrage outcomes. The mixture of successes and failures at the state level affords a unique look into movement outcomes and how state outcomes may have been affected by the availability of national resources. Using a resource mobilization framework, I investigate the effects of organizational ties between national and state organizations and provision of material resources to the states by the national. Through event history analysis, I find that, net of factors such as political and gendered opportunities and framing of movement arguments, resources delivered by the national organization to the states do not help explain the likelihood of winning woman suffrage at the state level. The primary contribution of my work is thus a partial re-writing of the resource mobilization story, at least for the U.S. suffrage movementnet of other factors, resources are not useful in the final stage of a movements history, specifically the policy change of voting rights for women.

This dissertation examines how identity processes affect teens mental health as they move through adolescence and transition into young adulthood. Using seven waves of panel data in the National Youth Survey (N = 1725), I test three main hypotheses. First, I argue that reflected self-esteem plays an important role in the development of adolescents sense of self. Adolescents whose reflected self-esteem is positive should experience better well-being than those persons with negative reflected self-esteem. Second, positive reflected self-esteem should be directly related to greater salience of and more time spent in the role-identity that is attached to the source of the reflected self-esteem. Finally, the relationship between reflected self-esteem and mental health should be medicated by the salience and the amount of time spent in the role. The first two hypotheses are strongly supported, but not the third. Implications of these findings are discussed. By understanding the identity process among adolescents, we may better explain how it is that teens in the United States come to avoid or to suffer detrimental and possibly debilitating mental health problems.

Although gender is the strongest correlate of delinquent behavior, most delinquency theories were formulated to explain male delinquency, and the academic literature has ignored status offenses such as running away, although this deviance comprises a substantial and controversial role in girls offending. This dissertation adds to the literature by examining the relationship of running away and status offending to concurrent delinquency and criminal behavior throughout the life-course. Also, while Power-Control Theory is reported to be sensitive to girls lives, this theory has not been sufficiently tested or used in an explanation of exclusively status offenses. Examining Power-Control Theory will uncover explanatory variables relevant to status offending, as well as any differences in the correlates between delinquent and status offending. The results suggest running away is related to juvenile delinquency and adult offending in a similar manner for both males and females, and, abuse experiences increase subsequent offending among runaways. However, the results provide little support for Power-Control, although some facets such as taste for risk and parental control were significant. In fact, daughters from command-class mothers actually had lower rates of runaway behavior. Three main conclusions are contributed: First, status offending is significantly related to concurrent and future deviance, possibly by providing contact with delinquent peers. Most status offenders, like most delinquent offenders, desist as they age into adulthood. Parental control and risk taking attitudes, variables often found in the delinquency literature, are also related to status offending. Second, abuse histories are significant for boys as well girls. Abuse has mostly been studied as an important factor for only female offending, yet this dissertation found abuse histories should be a focus of study in male delinquency as well. Third, this dissertation provides further evidence liberation-based theories are an inadequate explanation of female deviance. Although these theories have received little empirical support, this explanation is still accepted and wide-spread in academic theories and the popular literature and media. Suggestions for future research and policy implications conclude this paper.

This study explores how emotion culture relates to class position and beliefs about masculinity. What are the cultural ideologies, values, norms, and vocabularies about emotion that structure mens lives? And, do these dimensions of emotion culture vary by socioeconomic status? In-depth interviews were conducted with a split sample of eighty white, heterosexual men (forty upper-middle class, forty working class) to obtain their perspectives on masculinity and emotion. I outline and discuss dominant and secondary ideologies about stoicism and five emotions (sadness, anger, fear, love, and happiness) as they emerged from the interviews. Additionally, the compensatory masculinity model proposes that men will exaggerate certain traits of masculinity in the absence of other traits. While my research finds some support for this model regarding expressions of anger, it also challenges the theory vis-à-vis stoicism. Alternative explanations for class similarities and differences are also introduced.