03 · What You Need to Know
Responding to Suspicious Research Services Within a Collaborative Project
Why Is a Collaborator's Choice of Research Service Your Concern?
Collaborative research often involves shared responsibility for the evidence and claims presented in a publication.
A researcher may not personally perform every technical task, but the author team should understand how consequential research activities were conducted and whether the reported methods are accurate.
If a collaborator engages an external provider, the service may influence the dataset, analysis, manuscript, or publication process.
For example, a questionable data collection service may compromise observations used by every coauthor, even when only one collaborator negotiated the contract.
However, the existence of an external provider does not automatically make the study unreliable.
The appropriate question is whether the provider's actual work complied with the research design and relevant ethical requirements.
Does Using a Questionable Service Mean Your Collaborator Committed Misconduct?
No. Suspicion about a provider and evidence of misconduct by a collaborator are different matters.
A collaborator may have legitimately hired a company for editing, transcription, statistical consultation, or participant recruitment.
Alternatively, the collaborator may have misunderstood the provider's services, failed to verify its methods, or knowingly participated in deceptive activities.
These possibilities have different implications.
Formal misconduct findings require assessment under the applicable institutional or regulatory framework.
Researchers should therefore distinguish an unusual commercial arrangement from evidence that someone fabricated data, falsified results, plagiarized material, or violated other research integrity requirements.
What Should You Establish First?
The initial objective should be to identify the service's actual involvement in the research.
Relevant questions include what tasks were delegated, who authorized the arrangement, what materials were transferred, which deliverables were received, and how the outputs were incorporated into the study.
Researchers should also determine whether the provider's work was permitted under the approved protocol, institutional rules, and relevant publication policies.
For example, hiring a statistician to review model assumptions is different from purchasing a dataset falsely represented as newly collected observations.
The service's advertised activities may justify scrutiny, but they do not necessarily establish what occurred in the particular project.
Which Types of Service Involvement Require Different Responses?
| Service involvement |
Primary concern |
Appropriate initial assessment |
| Language editing |
Undisclosed substantive writing or altered scholarly claims. |
Compare manuscript versions and evaluate contribution requirements. |
| Statistical analysis |
Unjustified methods, unverifiable outputs, or manipulated results. |
Review scripts, model specifications, and analytical documentation. |
| Data collection |
Participant authenticity, consent, and data provenance. |
Examine collection procedures and available source records. |
| Qualitative coding |
Interpretive accountability and confidentiality. |
Review the coding process and analytical responsibilities. |
| Manuscript preparation |
Ghostwriting or misrepresentation of intellectual contribution. |
Establish who developed the manuscript's substantive content. |
| Publication assistance |
False reviewer identities or improper editorial influence. |
Review authorized submission activities and relevant declarations. |
| Authorship arrangements |
Purchased authorship or inappropriate allocation of credit. |
Assess actual contributions against applicable authorship criteria. |
The appropriate investigation depends on the actual activity. A concern about statistical validity should not automatically be treated as evidence of purchased authorship, and a questionable authorship arrangement does not independently prove that the data were fabricated.
Should You Ask the Collaborator for an Explanation?
In many circumstances, a factual and non-accusatory request for clarification may help resolve uncertainty.
For example, you might ask which tasks the provider performed, whether analytical scripts are available, or how the delivered observations can be traced to genuine collection activities.
However, direct discussion is not always the safest or most appropriate first step.
If there is a credible risk that evidence could be destroyed, participants could be harmed, or the person raising the concern could face retaliation, confidential institutional advice may be preferable.
Researchers should follow relevant reporting policies and avoid conducting unauthorized investigations.
The principles discussed in whether to confront someone before reporting suspected misconduct are relevant when deciding how to proceed.
What Evidence Should Be Preserved?
Relevant evidence may include contracts, invoices, service descriptions, manuscript versions, data files, analytical scripts, correspondence, and records of decisions about the provider's work.
Researchers should preserve materials to which they are legitimately entitled to have access.
They should not access a collaborator's private accounts, copy restricted participant records without authorization, or alter original research materials.
Documentation should distinguish confirmed facts from suspicions.
For example, a provider's website advertising completed manuscripts is evidence of an advertised service, not necessarily proof that the collaborator purchased one.
Preserving authentic records helps authorized reviewers establish what occurred without relying exclusively on recollections or accusations.
How Can You Verify Externally Supplied Data?
Verification should focus on whether the data have credible provenance and whether collection followed the approved methodology.
Depending on the study, relevant evidence may include recruitment records, eligibility procedures, collection logs, original exports, instrument versions, and documented quality checks.
For example, if a contractor supplied 500 survey responses, the research team should be able to explain how those respondents were recruited and how the final dataset was constructed.
Unusual response patterns may justify examination but do not independently establish fabrication.
The procedures discussed in verifying third-party research data collection can help structure this assessment.
What if the Provider Supplied Statistical Results?
Researchers should determine whether the analysis can be explained and adequately verified.
Useful materials may include variable definitions, sample-selection rules, data transformations, model specifications, diagnostic results, and analytical scripts.
A provider's promise of statistically significant findings is a warning sign because scientific conclusions should follow the evidence rather than a predetermined outcome.
However, receiving significant results from a commercial analyst does not itself establish manipulation.
The team should examine whether the selected methods are appropriate and whether the results can be reproduced or otherwise verified.
Where technical expertise is insufficient, independent statistical review may be necessary.
What if the Service Wrote Part of the Manuscript?
Writing assistance may be legitimate when the contribution is appropriately handled and the authors remain accountable.
However, substantive writing that develops interpretations, theoretical arguments, or scholarly conclusions may require authorship consideration or acknowledgment.
The team should establish whether the provider merely improved expression or introduced intellectual content.
Payment does not automatically eliminate attribution obligations.
Similarly, a collaborator's failure to disclose a writing service does not necessarily establish that the entire manuscript was ghostwritten.
The distinction between legitimate research assistance and ghostwriting should guide the assessment.
What if Purchased Authorship Is Suspected?
Purchased authorship involves payment for authorship credit without satisfying the relevant scholarly contribution and accountability criteria.
If a collaborator arranged such a transaction, the concern may affect the accuracy of authorship declarations and the integrity of the publication.
However, paying a research specialist for legitimate work is not equivalent to purchasing authorship.
Researchers should examine the actual contribution, financial arrangement, and governing journal requirements.
The implications of purchased authorship should be evaluated independently of unrelated concerns about data or methods.
What if the Service Manipulated Peer Review?
Some publication services may offer to handle manuscript submission or communicate with journals.
Legitimate administrative assistance differs from providing false reviewer identities, concealing conflicts, or improperly influencing editorial decisions.
If suspicious submission activity is identified, relevant records should be preserved and the journal's editorial office may need to be informed.
However, the author team's knowledge and authorization should be assessed separately from the provider's conduct.
A compromised review process does not automatically establish that the underlying research data are false.
The issues associated with fake and manipulated peer review require appropriate editorial assessment.
What if Confidential Participant Information Was Shared?
External services may receive identifiable datasets, interview recordings, transcripts, or other sensitive research materials.
Researchers should establish whether the transfer was permitted under participant consent, ethics approval, institutional policy, and applicable data-protection requirements.
If unauthorized processing or disclosure is suspected, the concern should be referred promptly through appropriate institutional privacy and security channels.
Researchers should not wait for proof of publication fraud before addressing a credible risk to participant confidentiality.
Privacy and research misconduct investigations may involve different authorities and obligations.
The responsibilities associated with contractor handling of confidential participant information remain relevant even when the service's scientific work appears acceptable.
Should You Stop Using the Provider's Work Immediately?
Temporary restrictions may be appropriate when credible concerns affect data authenticity, participant protection, or the reliability of consequential analyses.
For example, the team may pause submission of a manuscript while verifying an externally supplied dataset.
However, suspicion alone does not necessarily justify discarding all materials produced by the provider.
Researchers should assess which outputs are affected and whether independent verification is possible.
Original records should be preserved rather than silently deleted.
Where the concern involves participant safety or ongoing unauthorized data processing, more urgent containment may be necessary.
Should You Withdraw Your Name From the Manuscript?
Authorship withdrawal may be appropriate when a researcher cannot responsibly approve or accept accountability for the manuscript.
However, removing one's name does not necessarily resolve the underlying research integrity problem.
If fabricated data or manipulated editorial processes are suspected, the issue may still require institutional or journal assessment.
Researchers should also distinguish unresolved concerns from established misconduct.
Before making an authorship decision, it may be appropriate to seek confidential advice and determine whether the evidence can be verified.
Where the manuscript has already been submitted or published, authorship changes generally require journal procedures rather than unilateral alteration of the record.
What if the Collaborator Refuses to Provide Documentation?
Refusal may have different explanations, including contractual restrictions, confidentiality concerns, unavailable records, or unwillingness to cooperate.
Researchers should identify what information is necessary for evaluating the study and whether it can be obtained through authorized channels.
For example, a provider may legitimately withhold participant identities while still being expected to explain recruitment procedures and data provenance.
Where essential evidence remains unavailable, the team should consider whether the affected research can support the intended claims.
Refusal alone does not prove fabrication, but unverifiable evidence may be unsuitable for publication.
When Should You Report the Concern?
Reporting may be appropriate when credible evidence suggests a serious integrity violation, when necessary verification cannot be obtained, or when participant protection requires institutional intervention.
Researchers do not necessarily need conclusive proof before raising a good-faith concern through an authorized channel.
However, allegations should distinguish observed facts from interpretations and unresolved suspicions.
Appropriate recipients may include a research integrity officer, institutional ethics committee, privacy office, or journal editor, depending on the issue.
The principles governing evidence needed before reporting suspected misconduct help clarify this distinction.
What if the Collaborator Is Your Supervisor or a Senior Researcher?
Power relationships may complicate direct discussion and reporting.
A junior researcher may depend on a collaborator for supervision, employment, references, or future research opportunities.
In such circumstances, confidential advice from an appropriate institutional authority may be preferable to immediate confrontation.
Researchers should preserve evidence lawfully and avoid actions that could compromise an authorized investigation.
Institutional procedures should provide fair assessment and appropriate protections against retaliation.
However, available protections vary across jurisdictions and institutions, so researchers should verify the procedures that actually apply.
What if the Manuscript Has Already Been Submitted?
If unresolved concerns materially affect the manuscript's integrity, the author team may need to inform the journal and request an appropriate pause or assessment.
The communication should accurately describe the concern without asserting conclusions that have not been established.
For example, the team may explain that the provenance of externally collected observations is under review.
The journal may request documentation, suspend editorial consideration, or refer relevant matters to the institution.
Researchers should not submit false explanations or conceal consequential uncertainties simply to preserve the manuscript's progress.
What if the Article Has Already Been Published?
Published research may require correction, expression of concern, retraction, or another editorial response if serious problems are established.
However, discovering that a collaborator used a questionable service does not automatically mean the article must be retracted.
The appropriate action depends on the service's actual involvement and whether the findings or publication process remain reliable.
Authors should cooperate with institutional and editorial assessment and preserve relevant records.
Publication reliability and individual misconduct responsibility should be evaluated separately.
Can You Be Held Responsible for a Collaborator's Conduct?
Coauthorship may involve accountability for the published work, but it does not automatically establish personal participation in another person's misconduct.
Under the International Committee of Medical Journal Editors' recommendations, authors have responsibilities concerning their contributions and the appropriate resolution of questions about the work's accuracy or integrity.
Other disciplines may apply different authorship standards.
Researchers should therefore distinguish their obligation to address unreliable findings from any determination of personal culpability.
Failing to respond appropriately after discovering credible concerns may create separate accountability problems.
How Can Research Teams Prevent Similar Problems?
Research teams should establish expectations about external services before data collection, analysis, or manuscript preparation begins.
Agreements may specify who can engage providers, what approval is required, how data may be shared, and what documentation must be retained.
Authorship and disclosure expectations should also be discussed early.
Researchers should avoid arrangements that conceal substantive contributions or prevent verification of consequential outputs.
These procedures cannot guarantee that every collaborator will act appropriately, but they can reduce uncertainty and support timely corrective action.
Watch Out
Do not accuse a collaborator of fabrication solely because they used a suspicious commercial service. Equally, do not ignore unresolved questions about data authenticity, authorship, or participant confidentiality merely to avoid conflict within the research team.
