Trap questions (also bogus questions) are meant to check if respondents are speeding when filling out the survey and therefore cheat. We occasionally pose a trivial question (eg. To choose a certain answer at for this purpose created question) to respondents. If a certain respondent chooses wrong answer we can assume that the respondent was not paying attention to the content of the question.
Use of trap questions is recommended in panels and when we award respondents for responding to the survey. We want to eliminate cases where respondent only wants to satisfy the surveyor by answering quickly or choosing random answers. It is recommended that such questions are not too difficult, boring, unusual or offensive.
Study by Baker – Prewitt and Miller (2009) shows that more than 10% of the respondents are not successful at such questions. This research also shows that we can increase quality of the research by using trap questions:
- Respondents use more time to finish the survey.
- Less respondents speed through the survey.
- Less respondents choose the same answer at different questions (straightlining).
- Smaller likelihood for respondents to miss next trap question.
- Less mind cheating.
- Respondent’s answers show higher correlation with measurement scale.
We can also increase respondent’s attention by using a combination of positive and negative statements. This encourages respondents to be more careful when reading statements and be more thoughtful before answering an individual statement.