Dear REDCap community,
I recently saw an article by Pozzar et al. (2020) that describes that bots and illegitimate survey responses are becoming a common problem in online-recruitment studies, and it discusses ways to prevent low-quality responses. I'm curious what REDCap techniques you'd recommend to minimize the likelihood of bots and illegitimate responses for studies that involve online recruitment and paid surveys in REDCap. Here are some of the ideas they (and others have) presented:
-use automated survey invitations to send each respondent a unique survey link
-require respondents pass CAPTCHA (REDCap can do this on public surveys, not on unique survey links)
-include hidden items in a survey that will be seen by computers but not by respondents (I assume that the @HIDDEN-SURVEY action tag would accomplish this); if the items are completed, they were likely completed by a bot
-include a timestamp at the beginning of each instrument; REDCap already records the survey completion time; with the timestamp of the beginning and completion of each instrument, you can calculate how long it took each participant to complete the survey, and exclude responses that were unreasonably fast; also, you can determine their timezone, which can help if you have geographical exclusion criteria
-include and require a response to open-ended questions
-include items with directives (e.g., "Select the second option below")
-include pairs of items that can be compared for consistency
-collecting verifiable information (e.g., address, zipcode)
-designate the respondent's email address or some other identifier as a Secondary Unique Field in REDCap to prevent multiple responses from the same person
What other ideas do you have for how to prevent bots and illegitimate responses for studies that involve online recruitment and paid surveys in REDCap?
-Isaac