Preventing and Eliminating Bots and Participant Fraud in Online Surveys: Two Case Studies from International LGBTQ+ Social Research

Participant fraud from bots and ineligible participants poses a growing threat to international social research, requiring bespoke mitigation strategies. This paper presents two case studies of international LGBTQ+ surveys compromised by fraud and describes preventative and eliminative mitigation. Case study 1 (N = 1,707) describes a preventative screening process in a survey about LGBTQ+ leisure spaces, combining Qualtrics security tools with geolocation and email address checks. Case study 2 (N = 3,681) describes an eliminative strategy in an LGBTQ+ video gaming survey, using 10 fraud indicators analyzed using hierarchical and K-means cluster analysis. Both studies found fraud rates of 43-45%, illustrating the value of tailored, multi-layered strategies. Some indicators (e.g., IP address blacklisting, attention checks) were less effective, possibly due to privacy-seeking behaviours and neurodiversity in LGBTQ+ samples, but unique demographic characteristics could support sample validation. Strengths and limitations of preventative and eliminative approaches are compared, and practical recommendations offered.