Testing Methods for Improving Response Rates in Local Level Victimization Surveys

• In Event: Studying Victimization Patterns with Local Level Victimization Surveys

Abstract

Victimization surveys represent one of the three core methods for measuring crime. Traditionally, criminology has largely relied on the National Crime Victimization Survey for victimization data. There is a need, however, for victimization data at the local level to complement other sources of crime data (e.g., police incident reports, self-reports). Challenges in conducting public surveys in the era of cell phones, coupled with low base rates of victimization, make local level victimization surveys expensive and difficult to achieve representative samples. This paper presents the results of several experimental tests of different methods for conducting local level victimization surveys. The experiments were embedded in a local victimization survey employing a mixed-mode, address-based sampling design. The study tested the impact of various modes (paper, phone, web) and incentives on response rates. The results are framed in the context of developing more cost efficient methodologies while yielding reasonable response rates and representative samples.

Authors

• Edmund F. McGarrell, Michigan State University

• Chris Melde, Michigan State University