When Research Becomes All About the Bots: A Case Study on Fraud Prevention and Participant Validation in the Context of Abortion Storytelling
Michaela Krawczyk, Katie A. Siek · 2024
Effective fraud prevention and participant validation are essential for ensuring data quality in today’s highly-digitized research landscape. Increasingly sophisticated bots and high levels of fraudulent participants have generated a need for more complex and nuanced methods to combat fraudulent activity. In this paper, we share our experiences with fraudulent survey responses, which we encountered in our work around abortion storytelling, and the multi-stage protocol that we developed to validate participants. We found that effective fraud prevention should start early and include a variety of flagging methods to encourage holistic pattern-searching in data. Researchers should overestimate the amount of time they will need to validate participants and consider asking participants to assist in the validation process. We encourage researchers to be transparent about the interpretive nature of this work. To this end, we contribute a Participant Validation Guide in supplemental materials for community members to adapt in their own practices.