The Douglass Results Blog

AI is a valuable research tool that can refine questions and build more dynamic attribute lists. However, it can be detrimental to your online survey when humans use it to develop survey bots for financial gain. Bots are especially prevalent when conducting research among IT and Business professionals.

During a recent project, we were “stung” by a bot. It had managed to squeak through and answer a multitude of times, causing us to stop field and reassess the screening process at every level. What we found is the need to employ a combination of new and traditional approaches which we now employ for all our online surveys. Here’s what we recommend:

On the front end: employ traditional screening questions along with a staggered termination (letting a respondent answer a few more questions after being identified as not qualified).

Within the survey: create checks across questions and terminate the respondent if the responses from both questions do not align.

On the back end: check for duplicate responses, unusually fast completion times and a run of completes within seconds of each other.

We suggest conduct quality control checks on a daily or every other day basis.

Don’t let these pesky bots corrupt your data!