Investigating Interviewer Falsifications – A Quasi-experimental Design

Abstract

This paper discusses the strengths and limitations of experimentally-orientated research in sociology. In principal, results from experiments cannot be generalized. Nevertheless they allow controlling for certain conditions that are impossible to measure during a survey. This paper gives an example: It presents findings from a quasi-experimental research on interviewer falsifications. Interviewers conducted real standardized interviews and subsequently falsified corresponding interviews in the lab. This enables a comparison of real and falsified survey data. Our research design has the limitation that there is no proof that actual falsifiers in an actual survey fieldwork environment would have falsified interviews in the same way as did the participants of our study. However, only this quasi-experimental approach allows us to know for sure which interviews are falsified.