Trial and Tribulations of Using Official Data for Sampling at Micro-Geographic Units of Analysis

• In Event: Crime and Place: The Baltimore Longitudinal Hot Spots Study: Methodological Challenges and Lessons Learned

Abstract

In June 2013, the Baltimore Longitudinal Hot Spots Study began the process of identifying all eligible households on a sample of street segments to collect over 3,700 face-to-face surveys of residents in Baltimore City, MD. Traditionally, social science survey research methods rely upon official data as a means to provide a sampling frame of all potentially eligible survey respondents. This presentation discusses the advantages and disadvantages of using official data on occupied households from government agencies to provide accurate information for survey sampling at micro-geographic units of analysis. Since accurate sampling frames ensure that the data collected are representative of the target population, the presentation will discuss the limitations of relying upon official data to determine whether or not street segments are eligible for inclusion based upon the number of occupied households on a given street segment. Further, it will discuss the necessary steps that should be taken to ensure that household sampling frames are accurate for survey research at mice-geographic units of analysis.

Authors

• Matthew Nelson, George Mason University

• David Weisburd, George Mason University / Hebrew University

• Justin Ready, Arizona State University

• Brian Lawton, George Mason University

• Breanne Cave, George Mason University

• Amelia Haviland, Carnegie Mellon University