Notice: the WebSM website has not been updated since the beginning of 2018.

Events

Name Did I Do That? How Trap Questions Can Hurt Data Quality
Organized by Keith Phillips, ARF Re:Think Plus
From 17.04.2015, 12:30
To 17.04.2015, 13:30
Description

Without an interviewer present, it is generally believed that online self-completion surveys are more likely to have disengaged participants which have led to data quality concerns. One of those solutions the industry has come up with is asking quality control (or trap questions) to determine if a participant is paying attention. The biggest factors determining what proportion of participants are thrown out of a data set are the type of trap questions that are being used and how many trap questions are being asked. However, the fundamental assumption of a trap question is that the behavior within the trap is indicative of the participant's behavior throughout the survey, but if inattention is a random phenomenon then the exclusion from the trap could be meaningless. Furthermore, specific trap questions are designed to disengage participants and therefore, by design, create a systematic exclusion within the data, which can impacte data outcomes.

 

 

Print