Total Survey Error and Modes
Introduction
Design based surveys are based on the central idea that you start a study by drawing a random sample from your target population. You then approach these people, and ask them to participate in your study. People typically answer some questions, after which you record the responses in a database, and then conduct analyses.
There are many steps in this process where bias or error can be introduced. The Total Survey Error framework can be used to conceptualize these errors, and assess where in the survey process things go wrong. In order to illustrate the TSE framework, we discuss the choice of survey mode (whether you use Internet, mail, telephone or in-person interviews) and its impact on survey error.
Before class, make sure to finish the take-home exercise
Literature
- Groves, R. M., & Lyberg, L. (2010). Total survey error: Past, present, and future. Public opinion quarterly, 74(5), 849-879.
- Biemer, P. P. (2010). Total survey error: Design, implementation, and evaluation. Public Opinion Quarterly, 74(5), 817-848.
Lecture
Total Survey error and survey modes Slides
Class exercise
- Review your adopted survey.
- What TSE component would you focus on. What does this mean for survey design (joint exercise) Class Exercise
Take Home Exercise
Drawing simple random samples. Bring your results to the next class.