Associate Professor of Sociology and of Survey Research and Methodology

Ph.D. University of Michigan

Areas of Specialization:

  • Survey Methodology
  • Nonresponse Error
  • Measurement Error
  • Interviewer Effects


    Curriculum Vitae

    My research examines nonresponse, measurement and coverage errors that occur in surveys. First, I examine issues related to why people don’t participate in surveys, and what we can do through survey design features and statistical fixes to address nonresponse. For example, in a 2012 article, my coauthors and I experimentally found that people are more likely to participate in a mode that they said that they prefer than a nonpreferred mode. This is something that had long been hypothesized, but had not been experimentally evaluated. I also examine the intersection between nonresponse and measurement errors in sample surveys. Sometimes the actions that survey organizations take to increase response rates affect the quality of data from those who required additional effort. Second, Dr. Jolene Smyth (also in the Sociology department) and I are investigating how people answer survey questions in self-administered surveys through the use of eye tracking technology. In particular, we are interested in whether people who have lower literacy levels use different methods of answering questions than people with higher levels of literacy. Third, I am highly involved in the analysis of various types of paradata from surveys, such as keystroke files, response timing data, and call records. I have authored three overview chapters in a recent volume published by Wiley on paradata (Frauke Kreuter, ed.), and am co-Principal Investigator on a large NSF-funded grant on the use of paradata for evaluating measurement error in sample surveys. Finally, my recent research has examined methods for selecting a respondent within a household in self-administered surveys. Random selection of respondents within households is critical to maintaining a probability sample of the population. Self-administered modes don’t have an interviewer to help with that process, complicating this process tremendously.

    I teach courses related to survey methods and to quantitative analysis. I teach an online course that provides an introduction to survey methods, and in-person courses on applied sampling methods, analysis of complex survey data, and total survey error. I also teach the undergraduate statistics course, and will be teaching the graduate introductory regression course.

    Dr. Kristen M. Olson