teaching

This graduate level course introduces students to a set of principles of survey design that are the basis of standard practices in the field. The course uses the concept of total survey error as a framework, examining the features of survey design and how they contribute to survey error. Topics include: basic survey sampling concepts, alternative modes of data collection including computer assisted interviewing, effects of nonparticipation and nonresponse, development and evaluation of standardized survey instruments, survey management, survey interviewing, the role of the survey interviewer, post-survey processing of data, and research ethics. The course covers the basics of survey design and implementation and provide concepts and tools for evaluating surveys.

This graduate level course is designed to introduce students to the newest and most state-of-the-art methods of survey data collection, an appreciation of their history and development, as well as their rapidly developing research programs. Purposes both methodological (to improve the instrument and forward the state-of-the-art in survey techniques) and substantive (addressing research questions in sociological frameworks) are addressed.

This experiential and participatory course teaches students survey research best practices and allows them to apply these techniques working on a team of survey researchers in the environment of a production survey. The course is intended to extend students’ understanding of survey research techniques by giving them first hand experience operationalizing concepts, writing survey questions, and designing and testing a survey instrument. ​The course offers students hands-on experience in planning and conducting a survey, focusing on sampling, questionnaire development, and interviewing techniques. Students will put survey best practices into action in the development of a survey instrument by planning and moderating focus groups. They will evaluate questions by conducting cognitive interviews and completing behavior coding of recorded interviews. Students also participate in a sample listing exercise in the Boston area and conduct face-to-face pretest interviews.

The course builds a framework of statistical techniques to help students choose the most appropriate statistical technique/model to answer their research given the characteristics and limitations of their data. Students will be introduced to the general linear model and the generalized linear model for categorical dependent variables (logistic regression, multinomial logistic regression) ordinal dependent variables (ordinal logistic regression), and count outcomes (Poisson regression, negative binomial regression, and their extensions). More advanced approaches will be introduced, including structural equation modeling and models for grouped and nested data (cluster, multilevel, hierarchical models). Students will also be introduced to approaches for handling missing data.