Bad Actors in Research
Compensating participants for their time is an ethical and well-established practice in research. It acknowledges the value of their contribution and often helps recruit individuals who would otherwise be difficult to reach. But alongside these benefits lies a growing challenge: participants who misrepresent themselves to gain entry into studies for compensation.
This problem is especially pronounced in studies that rely on self-reported eligibility through online recruitment or remote interviews. Some individuals falsify demographic details to meet inclusion criteria. Others create duplicate accounts or provide misleading information to participate multiple times. In more concerning cases, groups of bad actors share strategies to “game” screeners and access stipends, skewing recruitment and participation data.
The implications are serious. Misrepresentation not only contaminates datasets but also jeopardizes the validity of research findings. When ineligible participants are included, results no longer accurately reflect the population under study. This can misdirect policy, dilute the effectiveness of interventions, and undermine public trust—particularly in sensitive areas like mental health, veteran care, or underserved community research where data accuracy is essential for resource allocation.
Bad actors can be difficult to spot, but there are telltale signs. Inconsistent responses across surveys or interviews, repeated use of contact details or IP addresses, and suspiciously fast completion times all warrant closer scrutiny. In studies requiring language fluency, participants who struggle to articulate coherent responses or comprehend basic instructions may require additional verification to ensure they meet eligibility criteria.
In virtual interviews, the signals can be more subtle. Participants refusing to turn on their cameras without a clear reason, giving scripted or robotic answers, or repeatedly claiming “connectivity issues” to avoid deeper questioning are all potential warning signs. Clusters of applicants arriving from unusual sources or closed online forums may also indicate coordinated fraud. Even emails with suspicious combination of a name followed by a series of numbers may indicate a fake account or participant.
Preventing misrepresentation starts with layered screening. This can include open-ended eligibility questions, staggered compensation (providing full payment only after participation is verified), and using tools to detect duplicate entries or unusual activity patterns. For studies requiring language fluency, incorporating brief, unscripted verbal or written assessments during screening can help confirm that participants meet the communication standards necessary for valid participation.
For virtual studies, requiring video participation when feasible, verifying basic identity information, requiring cameras to be on, and using rapport-building conversation early in interviews can also help validate participants. Researchers can further protect their work by carefully monitoring referral sources, avoiding overly public recruitment channels, and working with institutional review boards (IRBs) to balance participant privacy with study integrity.
The inclusion of ineligible participants isn’t just a logistical problem. It wastes time, drains funding, and compromises the very purpose of research: to generate valid, actionable knowledge that can improve lives. Addressing this issue doesn’t mean creating insurmountable barriers for honest participants, especially those from marginalized groups, but it does require a thoughtful approach to verification and screening.
The goal is simple: protect the integrity of the research while keeping studies inclusive and ethical. In doing so, we preserve not only the quality of our data but also the trust of the communities we aim to serve.
Consider the following:
How to Spot Bad Actors:
Inconsistent responses across screening, consent, or survey instruments.
Suspiciously fast completion of complex surveys or consent processes.
Repeated use of email addresses, phone numbers, or IP addresses across multiple entries.
IP addresses originating outside the target region or multiple submissions from a single IP.
Generic or nonsensical open-ended responses intended to bypass screening.
“Perfect” responses to eligibility questions that seem tailored to gain entry.
Sudden spikes in applicants from unusual or closed online forums.
Inability to provide consistent or plausible details during verification.
Significant difficulty articulating responses or understanding instructions in studies requiring language fluency.
Virtual Interview Red Flags:
Refusal to turn on video without a reasonable explanation.
Scripted, robotic, or identical answers to screening and interview questions.
Frequent claims of “connectivity issues” used to avoid deeper questioning.
Multiple participants appearing to be in the same environment or using identical virtual backgrounds.
Evasion when asked clarifying or follow-up questions.
How to Prevent Bad Actors From Entering Your Study:
Use multi-step, layered eligibility screeners with open-ended questions.
Verify participant information through follow-up emails, phone calls, or live pre-interview checks.
Track unique participant data (IP addresses, device fingerprints, email verification).
Validate participant location using geolocation data when geography is relevant.
Provide full compensation only after participation and data integrity are confirmed.
Embed attention-check questions within surveys to detect disengaged or fraudulent participants.
Require cameras for virtual interviews when feasible, and incorporate casual conversation for verification.
Include short, unscripted verbal or written assessments when language proficiency is an eligibility criterion.
Avoid overly public recruitment channels such as social media (this will guarantee bad actors); use controlled or vetted sources.
Track referral sources and investigate unusual surges in applicants.
Collaborate with IRBs to design verification strategies that balance participant privacy with research integrity.
In the end, make sure you are accounting for the extra time and increased workload you will have to contend with if you are recruiting through the general public and an unknown pool of participants. Good luck and be careful out there…
