▼ In recent decades, societal expectations about marital roles became generally more egalitarian. Despite this general shift some segments of society, such as evangelicals, have resisted this contemporary trend by retaining hierarchical gender roles. However, recent research suggests that some evangelicals soften their gendered marital model by incorporating more egalitarian elements. I use a symbolic interactionist perspective to explore how the perceived sanctity of marriage makes perceptions of gender and gendered obligations more salient to spouses in covenant marriages than to spouses in standard marriages. I use quantitative and qualitative data from a panel survey of newlywed covenant and standard couples in Louisiana (Marriage Matters, 1997-2004). In the quantitative analyses, I routinely find across all gender, religious, and marital attitude indices that covenants are more traditional than standards and this traditionalism is not an artifact of their greater evangelism. The qualitative findings illuminate the core meanings covenants attribute to their traditionalism and a key finding shows that they have a rigid hierarchy, but soften this hierarchy by constructing a warm, respectful interpretation of gender differences. Covenants believe that their marital status aides them to focus on each other more completely. They also view their marriage as being more Godly than do standards. Covenants use their beliefs in God and religion to justify their traditional attitudes. Conversely, standards do not seem to be as adept in discussing gender. They are less likely to share a guiding story about how gender should be manifested in marriage even though gender roles and displays are evident in their marriages. The core of the qualitative findings show a major distinction between standards and covenants. Advisors/Committee Members: Sanchez, Laura.

▼ Using data from the 1980, 1990, and 2000 U.S. Censuses, this project measures the association between educational attainment and outmigration from the State of Michigan. I begin by using descriptive statistics to document basic trends over time, focusing especially on the states where two groups, the college educated and the less than college educated, have moved. Intergroup comparisons are also presented and discussed. Next, logistic regression is used to analyze how Michigan outmigration differs by educational group after statistically controlling for select demographic, economic, and social characteristics. This study concludes by offering policy recommendations that Michigan officials should consider in an effort to reduce the Michigan exodus. Specific study results indicate that the destination states of Michigan outmigrants have been relatively constant regardless of educational level or census year. However, one key finding to emerge was that the number of college educated outmigrants was nearly twice that of the non-college educated group. Study results support the notion of a Michigan brain drain and also identify key characteristics that are likely contributors to outmigration. More specifically, results from logistic regression models indicate that even when controlling for educational status, the most powerful predictors of outmigration are age, previous migration, number of children under five and 18, and tenure. The magnitude of some of these differences changed slightly time, however, the directions remained consistent. Advisors/Committee Members: Goza, Franklin.