Background: I am offering a modest monetary incentive for the completion of an online survey and have publicized the survey on social media in order to reach the adolescent and young adult population. I have been inundated with 500+ fake responses (indicated by inconsistent responses, extremely quick response times, or made-up facilities in their open-ended responses). I realize this treads a fine line - I do not want to "cherry pick" the responses that I delete so I would prefer to have a concise way to identify spam responses or (even better) prevent them from occurring.
I have tried adding a CAPTCHA (did not work). I also recently added a password which could only be obtained upon emailing the research team. This also did not work - the fake respondents identified the password and the spam started back up again. I am hesitant to take the survey off of social media since it was the bulk of my participant recruitment plan.
Hi Kathryn.
I am looking for a similar answer. I have had to shutdown 2 Qualtrics surveys because of this. I have a captcha which prevents bots but there's nothing to prevent actual people from responding to my eligibility questions in a way that would allow them to advance to the main survey and onto the e-gift card form ($10 for completing online survey). The length of time to complete the survey raised my suspicion and the email addresses in the online gift card form alerted me that there were significant problems. There was a flood of responses in the wee hours of the morning making a clear delineation between "real" respondents and "fake" respondents. There were other patterns in response that I won't go into detail here. I have opened another survey since then and made sure to put quotas on all my surveys to prevent the flood of response. My next step is to do a screening call and then send the participant the link to a survey that only they can take. However, that doesn't address what to do with the data. I have contacted my IRB and Qualtrics (repeatedly) and explained the incident. I still would like to hear what others have been doing. Thanks for asking this!
Hi Kathryn,
I was searching for answers to a similar question, and stumbled across your post, I know it's from a while ago. I'd love to hear about what you did. Did you decide to compensate anyone that looked suspicious? I just did an online survey and I believe someone went in and completed it 8 or so times based on unusual repeated answers to eligibility questions, but it's hard to prove it. Apparently qualtrics has some embedded data codes that you can use for fraud detection but you have to put it into your survey before you publish it, so it's too late for me on that one now.
Also, I'm a humec Cornell alum too!
- Badges
- Science topic
- Similar topics
- Social Science
- Media
- Social Media