Developing a Spatial Sampling Frame for the U.S. General Social Survey

Data collections in social science increasingly seek to link contextual information on neighborhoods, local political jurisdictions, characteristics of the natural environment, economic and cultural opportunities, and toxic exposures and public health issues to individual survey respondents. The growing need to link multiple levels of analysis in individual data collections leads to questions about how best to conceive of this interface and how to best sample individuals within spatial contexts so that a wide variety of places are represented. This project headed by Professor Kevin Leicht, with co-PI Naresh Kumar from the Department of Geography (funded by the National Science Foundation) develops a spatial cluster sampling methodology that maximizes variation in known geographic characteristics across geographic space while maintaining the cost advantages of geographic cluster sampling for the collection of individual survey data. The project data collection specifically uses the Chicago MSA as a laboratory for testing this new sampling methodology.