Placebo Statements in List Experiments
12 Pages Posted: 9 May 2017 Last revised: 10 May 2017
Date Written: May 9, 2017
List experiments are a widely used survey technique for estimating the prevalence of socially sensitive attitudes or behaviors. This paper shows a potential vulnerability in their design: in a conventional list experiment, the treatment group respondents see a greater number of items (N 1) than the control group respondents (N). This can lead to mechanical inflation in the mean response of the treatment group, simply because respondents see a greater number of items. We test this through a placebo control statement that is by design false for all respondents. Using original data, we find the N 1 item placebo group mean to be greater than N standard control group mean, even though there should be no difference in means between the groups. This clearly indicates the presence of mechanical inflation. Further, we show that this effect is most pronounced among elderly and low socioeconomic status respondents. We recommend using the essentially costless placebo statement in list experiment control groups as a preventive measure against potentially upward biased findings.
Keywords: List experiments, item count technique, survey design, placebo statement, respondent fatigue.
JEL Classification: C81, C83
Suggested Citation: Suggested Citation