Mitigating Fraud in Online Surveys: A Methodological Approach

Online surveys offer data collection benefits and pitfalls, especially when utilizing crowdsourcing platforms such as Research Match and Prolific for recruitment.

The purpose of this methodology report was to describe strategies used to fortify an online REDCap research survey in which fraudulent and suspicious responses were identified and mitigated.

A two-pronged approach was designed to identify initial design limitations and engage in evidence-based redesign, which included scam alert features, study design changes, survey structure improvements, and crowdsourcing platform considerations.

A proactive, eight-step data cleaning protocol was designed and implemented in collaboration with the institutional review board and REDCap data experts. The same fraudulent records were identified within multiple steps of data cleaning, suggesting fraudulent records demonstrate multiple suspicious indicators.

Exciting opportunities in online data collection come with the risks of compromised data quality, resource waste, and damage to population health. Evidence-based protocols must be proactively designed to deter and detect early fraudulent results.