Safeguarding engagement with under-represented groups: Ethical considerations for mitigating fraudulent participation in health research

Abstract

Members of under-represented, marginalized groups experience disparate health outcomes, warranting research that improves equitable engagement in research and subsequent care practices. However, in our online study conducted during the COVID-19 pandemic, we encountered systematic signs of fraudulent behavior corroborated by recent reports from other research teams, with differing approaches to mitigating fraudulent participation. We highlight our experiences during recruitment and pilot focus groups, along with our actions for mitigating further recruitment of fraudulent participants. Then, we synthesize our learnings with other scholars’ experiences to provide recommendations for the study development, recruitment, screening, and data collection stages of research. We also offer a critical analysis of current recommendations and their implications for reinforcing barriers to engaging under-represented groups in research. The recommendations provided emphasize the importance of preventing fraudulent participation in quantitative and qualitative research through adopting multiple strategies, while balancing considerations for meaningfully engagement with under-represented groups to ensure equitable and inclusive research.

Introduction

Members of various under-represented social identity groups in research, including those based on race, ethnicity, gender, sexuality, and disability, experience patterns of discrimination in healthcare and disparate health outcomes (Hill et al., 2024; Krahn et al., 2015; Marrone, 2007; Ussher et al., 2023). These experiences have been noted as factors that impact differences in distress, quality of healthcare, and subsequent health outcomes (Major et al., 2013). This warrants research with under-represented groups to identify strategies for improving equitable engagement and care.

Our team conducted a qualitative study (Duong et al., 2025) to explore strategies for meaningful engagement, specifically with under-represented adolescents and young adults (AYAs) in cancer research. Our study was conducted virtually in 2022 due to the COVID-19 pandemic, when global social distancing and public health directives were still active across Canada (Dekker and Macdonald, 2022). The use of online research has grown since the onset of the pandemic, as it allows for greater reach for recruitment over large geographic areas, requiring less time and resources, and provides greater accessibility and convenience for participation (Hoskins et al., 2025; Pullen-Sansfaçon et al., 2024). However, fraudulent participation has been identified as a novel transdisciplinary issue, growing in parallel with the expanded use of online research (Davies et al., 2023; McLachlan et al., 2024), which we encountered as an unanticipated challenge during the pilot phase of our study.

Fraudulent participation has been identified as a phenomenon where individuals deliberately provide false information about their personal characteristics, or computer bots impersonate human beings, to gain access to a study as an eligible participant (Drysdale et al., 2023; Kumarasamy et al., 2024). It is speculated that fraudulent participants capitalize on the anonymity afforded in online studies specifically to misrepresent their personal characteristics and avoid being flagged as ineligible, with various researchers identifying financial incentives as a potential motivator for this deception (e.g. Chandler and Paolacci, 2017; Davies et al., 2023; McLachlan et al., 2024; Mistry et al., 2024; Teitcher et al., 2015). The COVID-19 pandemic had major economic impacts across the globe, contributing to rising unemployment across various sectors and global poverty (Yonzan et al., 2022). Furthermore, the COVID-19 pandemic exacerbated intersecting systemic inequities: affected employment sectors were more women-dominated, social distancing measures affected jobs with greater racialized representation, and temporary, low-income, and self-employed workers with greater job insecurity were disproportionately affected (Jackson and Victor, 2021; Maestripieri, 2021). These impact the financial situation of fraudulent participants, who may seek online research opportunities to mitigate situations of financial precarity (Roehl and Harland, 2022). For instance, the fraudulent participant confronted by Owens (2022) described their intersecting experiences as a woman with a disability and immigrant without legal status, who faced a pandemic-related job loss and participated in online studies “out of true necessity.” Overseas currency conversion rates can also incentivize potentially fraudulent participation from abroad, where seemingly small compensation amounts in the researcher’s local context can be deemed significant in other countries (Teitcher et al., 2015). As trends in health and socioeconomic inequities are expected to persist post-pandemic (Garcia Rojas et al., 2025; The British Academy, 2021), this context may continue to motivate fraudulent participation in online research.

Researchers are actioning various strategies for mitigating fraudulent participation in quantitative survey research, such as the Reflect, Expect, Analyze, Label (REAL) framework (Lawlor et al., 2021) for identifying and managing suspected participants and their data, or institutional research board (IRB)-approved protocols (e.g. Davies et al., 2023). However, fraudulent participation is also growing in individual and focus group-based qualitative research (e.g. Kumarasamy et al., 2024; McLachlan et al., 2024). Qualitative inquiry involves an additional element of vulnerability from participants who disclose personal narratives and lived experiences during interviews and focus groups to inform a research question. Therefore, fraudulent participants can threaten the safety of genuine participants in focus group-based qualitative research, who trust that they are in a safe space to share their confidential experiences with others (Davies et al., 2023).

Fraudulent participation also threatens data integrity, as the dishonest information provided by potentially fraudulent participants informs erroneous conclusions and recommendations for real-world practices and policies (Drysdale et al., 2023; Hoskins et al., 2025; Kumarasamy et al., 2024; Mistry et al., 2024; Pullen-Sansfaçon et al., 2024). Fraudulent research, a related concept where research is based on falsified or altered data, already contributes to public distrust in science; fraudulent participation can exacerbate this issue (Parker et al., 2022). Essential efforts to include marginalized and under-represented populations can further be thwarted when fraudulent participants take away from spaces meant for the representation of authentic voices, needs, and priorities (Reed et al., 2024).

Fraudulent participation threatens ethics principles of honesty, rigor, and respect, which can foster distrust in researchers as well (Hoskins et al., 2025; Teitcher et al., 2015). Researchers are institutional agents with an obligation to foster and especially restore trust in populations with extensive histories of mistreatment and exclusion by research institutions. Some discuss the benefit of increased skepticism and suspicion to detect fraudulence (Parker et al., 2022; Roehl and Harland, 2022). However, researchers’ distrust in their participants can perpetuate harmful power dynamics when determining who is a genuine participant, especially when decisions are made without transparency and ethical consideration, which exacerbates barriers to inclusion in research for under-represented populations. In participatory research, researcher distrust can also harm ongoing relationships with communities, which requires an openness and sensitivity to their needs (Drysdale et al., 2023). Therefore, our experiences and subsequent learnings are discussed to inform future research practices that balance meaningful engagement with under-represented groups in health research, while ensuring data integrity.

Our experience

Our team conducted a pilot study to identify barriers and enablers to engagement of under-represented AYAs in cancer research. The study materials, including recruitment materials, scripts, and surveys, were approved by the Health Research Ethics Board of Alberta – Cancer Committee (HREBA.CC-22-0126), and piloted in virtual focus groups. Eligible participants were between 15 and 39 years of age (National Cancer Institute, 2025), diagnosed with cancer within this age range, lived in Canada, and self-identified as part of an under-represented or under-researched social identity group. Under-represented identities included Indigenous, racialized, ¹ 2S/LGBTQIA+, ² and living with a disability. The study was promoted through posters on Instagram, Facebook, and X, with details on eligibility criteria, $50 CAD compensation via gift cards, a contact email, and eligibility survey link. The eligibility survey was administered through the Research Electronic Data Capture (REDCap) platform. It included yes-no questions on whether they met specific criteria, including “Do you identify as a member of an under-represented or under-researched group?,” with a follow-up text box to self-describe. If deemed eligible, they would provide an email address at the end of the eligibility survey, and a research team member would contact the provided email address to provide a consent form, demographic questionnaire, and schedule the participant’s focus group.

Three focus groups of 7–8 under-represented AYAs with cancer (N = 23) were conducted via a secure institutional Zoom videoconferencing account. Each focus group was facilitated by two researchers and one of our patient partners – people with lived experience of cancer as under-represented AYAs, one of whom also had clinical experience serving this population. Most focus group participants identified as Black (78%) and gay, lesbian, or bisexual (78%), with intersections with disabled and transgender identities. Preliminary barriers to engagement, that participants identified broadly in cancer research, included discrimination by researchers and healthcare providers, while enablers included the accessibility of online platforms and connection to people with shared lived experiences. As advertised, participants were compensated with digital gift cards upon completion of the focus groups.

Our recruitment and focus groups consistently showed signs of potentially fraudulent participation in line with other literature on this topic. Signs included a quick succession of near-identical, generic emails lacking salutations, sent shortly after recruitment materials were posted on social media (Davies et al., 2023; Hewitt et al., 2022; McLachlan et al., 2024; Pellicano et al., 2024; Pullen-Sansfaçon et al., 2024; Woolfall, 2023). For instance, emails would only say “I want to participate” or “I would be a good fit for this research.” Systematic discrepancies were identified between potentially fraudulent participants’ eligibility survey and demographic questionnaire responses (Davies et al., 2023; Kumarasamy et al., 2024; Pellicano et al., 2024). For instance, 14 respondents identified as First Nations or Métis in a question on Indigenous identity, but selected “Black/African/Caribbean” in a separate question on race and ethnicity – “Aboriginal/Indigenous (First Nations/Inuit/Métis)” was notably not chosen. Furthermore, a substantial number of email responses showed receipt of emails in time zones outside of Canada, mainly GMT+3 (Moscow/Arabia Standard Time) and GMT+1 (Central European/West Africa Time). This substantiates other researchers’ reports of suspicions that potentially fraudulent participants can be residing in one geographic location but claim to be elsewhere (Owens, 2022; Pellicano et al., 2024; Pullen-Sansfaçon et al., 2024). Lastly, potentially fraudulent participants tended to have phone numbers with area codes outside the target geographic location for participants; these phone numbers were traced to platforms used for multiple voice over internet protocol (VoIP) phone numbers (O’Donnell et al., 2023; Teitcher et al., 2015).

During the focus groups, all participants maintained anonymity by declining to turn their cameras on or blaming faulty cameras (Hoskins et al., 2025; Kumarasamy et al., 2024; McLachlan et al., 2024; Pellicano et al., 2024). Participants were often not able to provide clarifying details about their lived experiences, even when probed (Mistry et al., 2024; Pullen-Sansfaçon et al., 2024; Sefcik et al., 2023). For instance, when asked about enablers to participating in research, a participant wrote “multiple choice questions” in the Zoom chat box, then followed with “yes I meant they will be helpful” after being prompted for further detail. Some responses identified by co-facilitators were perceived as odd; for instance, when asked to introduce themselves, one participant had an identical word-for-word introduction as another participant from a different focus group. Another participant had posted laughing emojis and “wewe ongea kwanza nikuskie” to the Zoom chat box, which was manually translated from Swahili into English to mean “You speak first, I’ll listen to you”; this may have been meant as a private message to another participant, which would suggest some coordination between participants beyond the call for crafting responses to the focus group questions. While Pozzar et al. (2020) discusses their experience of potential human coordination in fraudulent online survey research, our experience highlights how online platforms for qualitative data collection can also facilitate this coordination.

The methods of engagement for under-represented groups, identified in both our pilot focus groups and our overarching study (Duong et al., 2025), included accessible participation through virtual platforms and supporting preferences for anonymity and privacy. However, based on our experiences with our pilot focus groups, we are concerned that methods to enable the participation of marginalized people in research could also be methods that enable fraudulent participation. Furthermore, the considerable similarities in the fraudulent participatory approaches identified between our team and others indicate that this is an evolving issue beyond individual circumstances, with fraudulent participants showing an increased sophistication and organization in their behaviors. While financial motives likely remain with rising income insecurity (Owens, 2022), organized efforts can also allow fraudulent participants to sway study results to reflect particular outcomes (McLachlan et al., 2024; O’Donnell et al., 2023). Furthermore, bots may have been used, as they facilitate mass fraudulent activity at rapid paces, with increasing capacity for bypassing verifications like CAPTCHAs (Jones et al., 2021; Pellicano et al., 2024; Teitcher et al., 2015). Pozzar et al. (2020) also discusses server farms, where large collections of computer hardware can be in a single facility, providing remote access to many virtual private servers that can facilitate the simultaneous completion of multiple online surveys, interviews, and focus groups. These concerns warranted action for our study. Though, due to ethics requirements, all focus group participants were still fully compensated for their time.

Our actions

After our pilot focus groups, debriefs were conducted with our patient partners. Our partners shared their lived experiences with past focus groups, along with insights on their experiences with co-facilitating our pilot groups. Their experiences as cancer survivors provided confidence in identifying participants as fraudulent, addressing ethical dilemmas as researchers who could have been excluding genuine participants from research meant to serve them.

Based on our patient partners’ input, several IRB-approved modifications were made. Firstly, a screening protocol was developed to enhance the integrity of the data. Participants were instructed to first complete the eligibility survey. Screenings with participants were then conducted by the study coordinator through videoconferencing (e.g. Zoom) or by phone, following an approved script to explain the study purpose and format, review participants’ format preferences for participating, and address questions or concerns (see Supplemental Material 1). Though, we did not require cameras to be on, due to concerns about excluding potentially genuine participants with technological barriers. Patient partners also recommended a systematic approach to verifying participants’ eligibility, by comparing responses to identify discrepancies that would indicate potential fraudulence early in the research process. For instance, if a participant stated during the screening call that they were in a Canadian province, but their email responses showed a different time zone, the fraudulent participant could be excluded prior to obtaining consent and collecting data. Participants’ data was excluded if discrepancies were identified in their responses between the eligibility survey, screening call, demographics questionnaire, and emails with the research team (see Supplemental Material 2).

Secondly, individual interviews were implemented as an alternate modality for participation in the post-pilot study, to allow for researchers to build one-on-one rapport with each participant, which would help participants feel more comfortable with disclosing further details that can support verification of genuine and potentially fraudulent participants. This was reflected in updated IRB-approved recruitment materials, consent forms, the study protocol, and focus group and interview scripts; these scripts were also updated with additional clarifying questions based on our pilot experiences.

Third, our team also explored removing compensation amounts from recruitment posters but maintaining them in the screening call and consent forms for transparency, after obtaining IRB approval for revised recruitment materials. However, recruitment halted as a result, so this was reintroduced in another accepted modification. A third approved modification involved the removal of the contact email on recruitment materials, to minimize the large influxes of responses.

Later, in recognition that trust is bidirectional, one research team member (JD) created a professional Instagram page, recording a video to introduce themself to the online AYA cancer community. They explained their lived experiences and reasons for researching this population, and invited connection through direct messages. By emphasizing engagement with the community, rather than recruitment for a specific study, this Instagram page demonstrated an authentic, trustworthy research space for members of the AYA cancer community to connect with. This page also allowed the research team member to connect with reputable patient advocates and community organizations, increasing the likelihood of recruiting authentic participants.

After adopting these strategies, we observed a noticeable decline in the frequency of fraudulent incidents; after introducing the Instagram page, the incidents stopped. Potentially fraudulent participants continued to send email inquiries, eligibility survey responses, and Instagram messages; however, engagement with authentic participants showed noticeably different response patterns that were more commonly seen in other cancer studies, as corroborated by our patient partners. This made it easier to flag potentially fraudulent participants earlier before proceeding to consenting and data collection.

These strategies were time- and resource-intensive, yet they were not foolproof. Even with the screening mechanisms in place, we had flagged a few cases. One individual attempted to undergo the screening twice under different names but was caught when they turned their camera on. The screening was still completed out of courtesy, but discussion with the Principal Investigator led to the decision to not proceed with an interview and cease contact, as the individual was already ineligible after their first screening. Another individual had completed the demographics questionnaire right before their interview, and discrepancies in their survey data were only identified afterward. This participant’s data was subsequently removed from analyses, as their potential fraudulence meant their data may not reflect the experiences of authentic under-represented groups. Therefore, data retained and reflected in quantitative demographics and qualitative themes would represent genuinely eligible participants, whose experiences were relevant to our research question. Overall, out of 77 expressions of interest, 57 were flagged as fraudulent, two were uncertain due to a lack of response after completing the eligibility survey, and 18 were authentic, though one had dropped out before the interview due to personal reasons.

Recommendations and ethical considerations

There are several recommendations for addressing fraudulent participation in online research that we propose, alongside other recommendations in the literature (see Table 1). These recommendations are based on a proactive approach involving multiple strategies, to prevent as many fraudulent participants as possible. The following strategies include those that have been employed by other researchers, along with novel strategies and considerations for inclusive health research. It is important to note that different approaches to identifying and addressing fraudulent participants have unique ethical and contextual considerations. Furthermore, identifying multiple signs of potential fraudulence and having multiple strategies in place can increase the effectiveness of one’s approach to fraudulence (Davies et al., 2023; Kumarasamy et al., 2024; Lawlor et al., 2021; Pozzar et al., 2020; Teitcher et al., 2015).

Table 1.

Summary of recommendations for addressing fraudulent research participants.

Study component	Recommendations
Study development	• Establish a protocol for identifying and managing potentially fraudulent participants and data, and submit to the IRB • Partner with patients and community members, whose insights can support identification of fraudulent behavior while minimizing harm to genuine participants
Recruitment & advertisement of compensation	• Build trust with community partners and groups to facilitate recruitment • Disclose financial incentives to ensure equitable recruitment and transparency • Consider using physical gift cards, gift cards restricted to the country of residence of eligible participants, non-financial incentives, or pro-rated compensation for multi-step data collection
Screenings	• Incorporate screening surveys (with reCAPTCHA functions, single-use personalized links) and/or calls to build rapport and support verification • During screening calls, consider requiring cameras to be on for a few minutes to visually confirm unique participants • Establish training, supervision, and regular meetings to support research personnel conducting screenings and navigating uncertainties around identifying fraudulent participants • Ensure ineligible participants know what avenues to contact if they have concerns around unfair exclusion (e.g. with the research team’s IRB)
Ongoing data quality checks	• Review data as it is collected to identify discrepancies, allowing for follow-up clarification
Broader practices	• Connect with IRBs to raise awareness • Establish communities of practice to share timely solutions and best practices

Study development

Mitigation strategies should start at the study development phase, as strategies to prevent, identify, and manage potentially fraudulent participants and their data can be outlined in protocols and approved by IRBs for ethical practices (Davies et al., 2023; Kumarasamy et al., 2024; McLachlan et al., 2024; Pozzar et al., 2020; Ridge et al., 2023). In protocols, procedures to determine genuine, suspected, and fraudulent participants should be clearly outlined to ensure classifications are systematic throughout the research process and across study team members. The research team’s approach for managing data should also be specified – whether taking a “cynic” approach that tends to exclude, a “skeptic” approach that proceeds with caution, or a “seeker” approach that includes fraudulent data to explore other research questions (Flicker, 2004). If researchers want to track potential participants’ IP addresses to verify their geographic location, IRB approval would be required for the collection, storage, and/or destruction of identifying information that is not relevant to the study’s objectives (McLachlan et al., 2024; Reed et al., 2024; Roehl and Harland, 2022; Sefcik et al., 2023). Furthermore, online research reporting guidelines should include a section on mitigating fraudulent responses, to ensure researchers are systematically developing protocols to proactively minimize this risk, and support sharing best practices in the research community.

Taking a community-engaged approach to research informed by the experiences of patient partners and clinicians would strengthen a proactive approach to mitigating fraudulent participation, including in health research. As we found with our experiences, patient partners and clinicians hold unique insights, along with lived and/or professional experiences that can inform what is considered atypical for patients’ communication styles and behaviors (Davies et al., 2023; McLachlan et al., 2024). These insights can shape criteria outlined in study protocols for identifying potentially fraudulent participants, ensuring that researchers take appropriate measures to balance potential concerns around privacy, data quality, and minimizing harm to patients and community members in research (Drysdale et al., 2023; Jones et al., 2021).

While we involved patient partners from the start, ensuring the early involvement of our IRB with proactive considerations for fraudulent participation could have prevented the need for several rounds of ethics modifications, which can contribute to delays in data collection while awaiting IRB approvals. Though, researchers may still need to update recruitment protocols, trying different approaches that work best for their target population, as well as provide documentation to their IRB of how fraudulent participants are addressed, depending on institutional requirements.

Recruitment and advertisement of compensation

Many researchers report experiences of fraudulent participation when recruiting through social media and online platforms (e.g. Davies et al., 2023; Hewitt et al., 2022; Hoskins et al., 2025; McLachlan et al., 2024; O’Donnell et al., 2023; Reed et al., 2024); however, recruitment through these avenues is still beneficial. Many researchers have found social media to enable recruitment of “hard-to-reach” populations, including those from marginalized and stigmatized backgrounds (Russomanno et al., 2019; Topolovec-Vranic and Natarajan, 2016); furthermore, patients themselves report being engaged with each other, healthcare providers, and community organizations through these platforms (Antheunis et al., 2013; Duong et al., 2025; Farsi et al., 2022). Therefore, taking additional steps to foster relationships built on trust with community partners and groups can enable recruitment of genuine participants connected to these groups (McLachlan et al., 2024; Owens, 2022). Snowball sampling is another strategy with an element of trust, though some researchers caution the potential increase in risk of recruiting batches of fraudulent participants (e.g. Mistry et al., 2024; Sefcik et al., 2023).

Many researchers recommend caution with how much information is disclosed on recruitment materials, particularly the level of detail regarding financial incentives, to address the potential incentive for deception (e.g. Chandler and Paolacci, 2017; Mistry et al., 2024; Sefcik et al., 2023). Researchers with past experiences of fraudulent participants often opted to exclude disclosure of monetary amounts from their public advertisements and instead disclose this amount during screening calls (e.g. Davies et al., 2023; Hoskins et al., 2025; Kumarasamy et al., 2024; Reed et al., 2024). However, compensation and transparency are core enablers for participants from under-represented backgrounds to engage in research broadly, as these factors demonstrate respect for participants’ time and contributions (Drysdale et al., 2023; Duong et al., 2025). Removing transparency around compensation can perpetuate barriers to access for under-represented groups with a history of exploitation in research and systemic disparities (Cheung et al., 2021; Pellicano et al., 2024). Furthermore, there are conflicting reports on whether disclosing a lower incentive amount on recruitment materials, or mentioning the general possibility of compensation instead, is enough to prevent fraudulent participation (Drysdale et al., 2023; Pozzar et al., 2020; Teitcher et al., 2015). Therefore, we recommend that disclosure and minimal deception is important to practice as researchers, and the implementation of complementing mitigation strategies can still be sufficient to address fraudulent participation.

To address concerns around financial incentives as motivators for fraudulent participation, alternative or additional strategies can ensure genuine participants are still compensated. Mailing physical gift cards to any valid mail address has been identified as a useful strategy for ensuring only one gift card is administered per participant, so long as participants are informed of this prior to providing consent (Kumarasamy et al., 2024; Pozzar et al., 2020). Gift cards can also have restricted validity to the country that genuine participants are expected to reside in (Davies et al., 2023; Mistry et al., 2024; Pellicano et al., 2024). Non-financial incentives can be used to provide compensation that supports the researchers’ target population, such as classroom resources for teachers or access to training resources for community members, though it is recognized that this may lower interest in genuine participants (Lawlor et al., 2021; McLachlan et al., 2024). Lastly, prorated compensation practices can be used to match the steps of participation, such as completing a survey, then a follow-up interview in mixed-methods studies (Kumarasamy et al., 2024; McLachlan et al., 2024).

Screenings

Screening procedures prior to obtaining informed consent have been a highly recommended practice in quantitative and qualitative research alike. Online screening surveys can help researchers flag overtly ineligible participants (Hoskins et al., 2025). Online survey platforms also have features such as collection of IP addresses and other metadata to identify participants’ locations, reCAPTCHA functions to block bots, and single-use personalized links to prevent individuals from attempting to participate multiple times (Chandler and Paolacci, 2017; Davies et al., 2023; Jones et al., 2021; Kumarasamy et al., 2024; Lawlor et al., 2021; Pozzar et al., 2020; Reed et al., 2024). However, the inaccessibility of reCAPTCHA functions, how students and family members may share IP addresses, and the use of VPNs due to privacy concerns, should be considered (Pellicano et al., 2024; Teitcher et al., 2015).

Some researchers exercise caution with using self-identification of identity characteristics on surveys, and/or required identification documents to verify eligibility instead (Chandler and Paolacci, 2017). However, this requirement of evidence to prove one’s identity can potentially damage rapport and participants’ trust in the research process, especially if asked to prove identities that are under-represented (Kumarasamy et al., 2024; Lawlor et al., 2021). If requiring verification documents, consent forms should clearly state what information will be collected, its use for verification, and steps to ensure it will not be published or made public (Mistry et al., 2024; Roehl and Harland, 2022). The amount of information collected should be the minimum needed to determine eligibility, to avoid unnecessary participant burden (Jones et al., 2021).

Many researchers emphasize the importance of follow-up screening calls, introducing an additional point of contact to build rapport and trust between researchers and participants, as well as gain additional information for verification with screening survey responses (e.g. Hoskins et al., 2025; Kumarasamy et al., 2024; McLachlan et al., 2024; Owens, 2022). This can also address limitations in eligibility surveys, which potential participants can retake to “learn” answers that indicate eligibility to participate (Roehl and Harland, 2022). We found that screening calls allow participants to provide context to their survey responses, review the consent form together, and address questions, to ensure the participant was a good fit for the study. In one study, some participants expressed their appreciation for this additional security, as it increased their comfort with enrolling in a study after connecting with a study team member and knowing protections were in place (Reed et al., 2024). Video calls are especially beneficial, as requiring participants’ cameras to be on for a few minutes allows researchers to visually confirm unique participants (McLachlan et al., 2024; Mistry et al., 2024; Pellicano et al., 2024; Roehl and Harland, 2022). However, potential barriers to screening calls include connectivity issues, poor video quality, and other forms of unreliable phone or internet access to virtual screenings. Therefore, researchers should exercise caution around excluding participants based on these factors, as this can affect the inclusion of individuals who may face broader financial barriers (Hoskins et al., 2025). Furthermore, confidentiality concerns may be especially important for marginalized groups, such as transgender people who may be more readily recognizable in areas with low representation, or participants who are hesitant to discuss sensitive topics depending on the degree of privacy in their location (Drysdale et al., 2023; Pullen-Sansfaçon et al., 2024). During screening calls and in consent forms, rationales for gathering identifying information should also be clearly outlined, with transparency on whether participants are deemed eligible or not, whether fraudulent participants receive compensation, and who they can contact if there are concerns around unfair exclusion (Jones et al., 2021; McLachlan et al., 2024; Mistry et al., 2024; Pozzar et al., 2020; Teitcher et al., 2015).

Decisions to include or exclude participants at the screening stage can come with ethical dilemmas. Determining fraudulent participants involves challenging the authenticity of their lived experiences, based on subjective interpretations and dispositions (Hoskins et al., 2025; Woolfall, 2023). Therefore, researchers are at risk of prematurely excluding genuine participants and silencing their voices with stringent eligibility criteria and screening processes (Lawlor et al., 2021). These processes can reinforce patterns of social exclusion, particularly for participants who are already under-represented in research, and foster bidirectional distrust, negatively affecting the researcher-participant relationship (Hoskins et al., 2025; McLachlan et al., 2024). Building relationships and rapport is especially important for the quality and richness of data on under-represented groups, as participants must feel safe when disclosing their lived experiences when they have been historically mistreated by health research and systems (Drysdale et al., 2023; Hoskins et al., 2025). Therefore, being proactive in outlining one’s approach to identifying and mitigating potential fraudulence is key, along with practicing reflexivity to recognize how one’s positionality impacts the research process.

Overall, it is important to balance data quality mechanisms, while minimizing the burden on participants to demonstrate their eligibility to engage in research. Furthermore, the level of detail provided to potential participants regarding data collection and use should be balanced, to respect individuals’ right to informed consent while ensuring that fraudulent participants do not learn of researchers’ specific methods to detect fraud (Teitcher et al., 2015). Excess detail could also deter genuine participants, as strategies such as collecting IP addresses could be viewed as an excessive breach of one’s right to privacy. Therefore, considerations of whether a strategy’s benefit outweighs the potential harms of deception are necessary; in the case of IP addresses, the researcher’s appraisal could lead to the adoption of other strategies in lieu.

To further support decision-making around potential fraudulence during screenings, training and supervision could be provided for research personnel conducting screening calls, so these individuals develop confidence in identifying potentially fraudulent participants while on call (McLachlan et al., 2024). Another strategy is to implement a pause between screenings and enrollment, as research personnel can debrief potential concerns with the principal investigator or team (Hoskins et al., 2025; Jones et al., 2021). Documentation in a reflexivity journal can also support systematic approaches in decision-making (Roehl and Harland, 2022).

Ongoing data quality checks

Signs of fraudulence can still be detected beyond the screening stage, even with mitigation strategies in place (Pozzar et al., 2020; Roehl and Harland, 2022). Therefore, having ongoing data quality checks is important to confirm consistency in participants’ details, allowing researchers to follow-up with participants about potential discrepancies instead of forgoing whole datasets (Davies et al., 2023; Kumarasamy et al., 2024; Owens, 2022; Pellicano et al., 2024; Roehl and Harland, 2022; Teitcher et al., 2015). For instance, Mistry et al. (2024) reported how they asked participants to confirm their pre-interview questionnaire responses at the start of their virtual interview; if their answers were inconsistent, the researcher informed them that they were ineligible to participate. In quantitative research methods like demographic surveys, attention checks can be used to ensure conscientious responses and minimize random responses from bots, increasing data integrity (Marjanovic et al., 2014). There can also be adaptations to response patterns over time in qualitative research, as potentially fraudulent participants learn what is required to participate in the study; however, there is an element of subjectivity in determining whether there are patterns between participants and interviews or focus groups (Davies et al., 2023). Having a consistent set of research team members involved in reviewing participants’ responses can help with ensuring a rigorous, standardized process to identifying potential fraudulence (Kumarasamy et al., 2024; Pullen-Sansfaçon et al., 2024). Regular meetings and debriefings can also support research staff with navigating uncertainties around identifying fraudulent participants (Pellicano et al., 2024; Ridge et al., 2023).

Broader practices

Fraudulent participation in research is becoming more prevalent across disciplines and methodologies, yet many researchers adopt ad hoc approaches in silo. Long-term comprehensive education, training, and resources is warranted, to ensure researchers are well-equipped to proactively and ethically address this evolving problem (Hewitt et al., 2022; Sefcik et al., 2023). This is especially important with the use of bots and the potential organization of fraudulent participants, along with the rise of artificial intelligence, which presents new opportunities for fabricating more sophisticated and realistic descriptions of personal characteristics and experiences (Pellicano et al., 2024). For instance, fraudulent participants can manipulate their audiovisual presentation and superimpose filters to impersonate other individuals via deepfakes (Sandotra and Arora, 2024), which can threaten the validity of online interviews and focus groups. As knowledge and awareness expands around these issues, resources can be developed collaboratively across scholars through communities of practice to allow for the real-time sharing of solutions and best practices (Kumarasamy et al., 2024; McLachlan et al., 2024). Connecting with one’s IRB can also support awareness-raising (Mistry et al., 2024).

Researchers should also report their experiences of fraudulent participation as separate from attrition for increased transparency. Attrition reflects a participant’s choice to withdraw from a study, while excluding participants and data due to potential fraudulence reflects the researcher’s evaluation of their eligibility. This transparency will allow other research teams to grasp the extent of fraudulent participation and justify their mitigation strategies.

Conclusion

Fraudulent participation is a growing problem in online research, including research in health contexts. There are existing strategies and recommendations for mitigating this issue, but they can exacerbate barriers to participation for members of under-represented groups. It is imperative that researchers balance strategies for ensuring data integrity, while recognizing the need to foster safe environments for patients and community members to participate, that address the unique barriers that under-represented groups face in research. As strategies for engaging in fraudulent participation evolve, we call upon fellow researchers to work together to bring attention to this issue and proactively create safe spaces to promote meaningful engagement with their participants.

Supplemental Material

sj-docx-1-rea-10.1177_17470161261434262 – Supplemental material for Safeguarding engagement with under-represented groups: Ethical considerations for mitigating fraudulent participation in health research

Supplemental material, sj-docx-1-rea-10.1177_17470161261434262 for Safeguarding engagement with under-represented groups: Ethical considerations for mitigating fraudulent participation in health research by Jenny Duong, Iqra Rahamatullah, Tristan Bilash, Caitlin Forbes, Sharon H. J. Hou, Brianna Henry, Sheila N. Garland, Jacqueline L. Bender, Perri R. Tutelman and Fiona S. M. Schulte in Research Ethics

Supplemental Material

sj-docx-2-rea-10.1177_17470161261434262 – Supplemental material for Safeguarding engagement with under-represented groups: Ethical considerations for mitigating fraudulent participation in health research

Supplemental material, sj-docx-2-rea-10.1177_17470161261434262 for Safeguarding engagement with under-represented groups: Ethical considerations for mitigating fraudulent participation in health research by Jenny Duong, Iqra Rahamatullah, Tristan Bilash, Caitlin Forbes, Sharon H. J. Hou, Brianna Henry, Sheila N. Garland, Jacqueline L. Bender, Perri R. Tutelman and Fiona S. M. Schulte in Research Ethics