Contact My SSW Intranet

Main menu

Special Seminar: Statistical Issues in Studying Social and Ethnic Disparities (Educ)

To see course dates, times or section specific details, please log in.

SW839, Section 001

Researchers in education, social work, human development, mental health, and health share a strong interest in understanding social and ethnic disparities in outcomes. Persistent social and ethnic gaps in birthweight, early reading achievement, early pregnancy, violent victimization, adult health, and mortality, for example, pose major challenges to policy and scholarship across the disciplines. This seminar will consider recurring statistical issues that arise in studying these disparities.

One set of issues concerns individual and contextual components of inequality. For example, an ethnic gap in educational achievement will entail a gap between children of different ethnicity who attend the same school; but this ethnic gap may also reflect mean differences between schools that are, to some degree, ethnically segregated. The within-school gap may vary from school to school. Similarly, neighborhoods will vary in ethnic composition, further complicating our understanding of the meaning of ethnic inequality. Statistical methods are needed that account for individual and contextual components of inequality.

A second, related set of issues concerns causal inference. Causal mechanisms creating inequality may operate at the level of the person or family, but they may also operate at the level of the social context, that is, the school, neighborhood, or firm. Experimental interventions may randomize the person or the context to alternative treatments; non-experimental studies must cope with selection biases that can operate at each level.

A third set of issues involves measurement. Protective and risk factors at the level of the person or the context may be measured with error. Methods for assessing the quality of social settings are much less developed than are methods for assessing individuals, though scientists will often disagree on the validity of individual measures as well.

This seminar will consider these issues as they arise in the work of the participants. While the first several weeks will lay a foundation through readings, lectures, and discussions, the bulk of our time will be spent considering methodological issues as they arise in specific studies in which the participants are engaged. During lab sessions outside of the regular class meeting time, participants will work with an advanced doctoral student to carry out analyses in response to seminar discussions. Participants will have multiple opportunities to present work in progress and to pursue analytic questions and issues that arise from seminar discussion. The aim is to push the research of the participants forward while creating a higher level of understanding of foundational issues.

Semester: Fall 2002
Instructors: Raudenbush, Stephen W., Verbitsky, Natalya
U-M Class #: 33959
Program Type: Residential
Credits: 3 Credit Hours

Course Codes

W:Social Work is not the home dept; home dept in parenthesis, contact home dept with questions
X:Social Work is the home department of this course

Contact Us Press escape to close