03 · What You Need to Know
Shared Responsibility Does Not Mean Identical Responsibility
Research integrity begins with individual responsibility
Every researcher has responsibility for their own conduct. A team member cannot normally justify fabricating data, misrepresenting an analysis, plagiarizing another person's work, or concealing a known integrity problem merely by saying that a supervisor instructed them to do it.
The Singapore Statement on Research Integrity expresses this principle directly by stating that researchers should take responsibility for the trustworthiness of their research. The European Code of Conduct similarly places accountability among its fundamental principles of research integrity.
Individual responsibility is therefore the starting point. Collaboration does not dissolve personal professional judgment.
Team members are responsible for the work they actually perform
Responsibilities should ordinarily follow roles and contributions.
A researcher collecting interview data may be responsible for following the approved protocol, documenting collection accurately, protecting confidential information, and raising problems encountered during fieldwork. A statistician may be responsible for the integrity of the analysis they conduct and for communicating assumptions or limitations. A data manager may carry specific responsibilities for access, documentation, security, and version control.
This does not mean that each person operates in a sealed compartment. Team members often need enough understanding of adjacent work to communicate effectively and recognize obvious problems. But it is generally unreasonable to assume that every contributor possesses the expertise, information, or authority to independently verify every component of a complex project.
Research leaders usually have additional responsibilities
Principal investigators, project leaders, laboratory heads, and supervisors often have responsibilities beyond their own direct technical contribution.
These may include establishing appropriate procedures, assigning roles, ensuring that team members receive necessary training, providing reasonable supervision, supporting adequate research records, addressing conflicts, creating channels for raising concerns, and responding when integrity problems emerge.
The National Academies identifies neglectful or exploitative research supervision as a detrimental research practice. The European Code of Conduct likewise places responsibilities on senior researchers, research leaders, and supervisors to support appropriate training, supervision, and a research environment conducive to integrity.
Exactly what a supervisor owes will depend on context. A principal investigator supervising a novice student performing an unfamiliar procedure may reasonably need closer oversight than a project leader collaborating with an experienced specialist who independently leads one technical component.
Delegating a task does not always delegate oversight
Research teams necessarily divide labor. A principal investigator cannot personally pipette every sample, verify every transcript, inspect every line of code, and independently repeat every statistical calculation.
Delegation is therefore normal. Abdication is different.
If a research leader assigns an important task, reasonable oversight may still include selecting appropriately qualified personnel, ensuring they understand expectations, establishing quality-control procedures, reviewing outputs at an appropriate level, and responding to warning signs.
The appropriate level of oversight depends on the risk, complexity, experience of the person performing the work, and consequences of error. Research integrity does not require omniscience. It does require reasonable governance of work for which a leader has supervisory responsibility.
Delegation
Assigning research work to another qualified person while maintaining whatever communication, supervision, verification, and accountability are reasonably required by the role.
Abdication
Treating assignment of a task as though it eliminates any responsibility to establish appropriate processes, respond to warning signs, or exercise oversight that the role reasonably requires.
Authorship creates responsibilities, but those responsibilities are not necessarily unlimited
Authorship is not merely a reward for participation. It communicates contribution and responsibility for published work.
The Singapore Statement states that researchers should take responsibility for their contributions to publications, funding applications, reports, and other representations of their research, and that authors should meet applicable authorship criteria.
The National Academies likewise emphasizes authorship as a mechanism for assigning both credit and responsibility. Clear authorship practices are especially important when errors or misconduct allegations arise.
However, disciplinary and journal authorship standards differ. In large interdisciplinary collaborations, an author may not personally possess the expertise to verify every technical detail contributed by every specialist. Responsible authorship therefore requires understanding and meeting the accountability standard applicable to the publication rather than assuming either that authors are responsible for nothing outside their own paragraph or that every author personally certifies every raw observation.
Teams should make responsibilities explicit before something goes wrong
Many integrity disputes become difficult because nobody established who was responsible for what.
Before data collection begins, teams should clarify matters such as who maintains primary records, who controls dataset versions, who verifies data entry, who has authority to exclude observations, who performs and checks analyses, who manages approvals, who communicates with participants or collaborators, who decides authorship, and who handles corrections or concerns.
The larger and more interdisciplinary the team, the more useful explicit documentation becomes. Assumptions that are obvious to one discipline may be completely unfamiliar to another.
This is particularly important in collaborative research involving different institutions, where policies, data systems, authorship conventions, and supervisory structures may differ.
Responsibility can change when someone becomes aware of a problem
A researcher may have had no responsibility for creating an error but acquire a new responsibility after discovering it.
Suppose a co-author learns that a central analysis cannot be reproduced from the underlying data. The co-author did not perform the analysis and may have had no reason to suspect a problem previously. Once the problem becomes known, however, simply ignoring it because "that was the statistician's part" may be inconsistent with responsible research conduct.
The appropriate response depends on the seriousness and evidence. It might involve asking questions, preserving records, alerting appropriate team members, seeking independent methodological review, contacting institutional officials, or helping correct the research record.
Awareness therefore matters. Responsibility is not determined only by who caused the original problem.
Shared responsibility does not create automatic collective guilt
The phrase "the whole team is responsible" is useful when describing a culture of integrity, but dangerous if interpreted as automatic culpability.
Suppose one researcher secretly fabricates a set of measurements despite reasonable procedures and supervision. Other team members may have had no knowledge of the fabrication and no reasonable opportunity to detect it. Their membership in the team does not automatically establish misconduct.
Whether a researcher can be responsible for misconduct committed by another team member requires attention to their own conduct, knowledge, duties, participation, and response.
Watch Out
Do not confuse collective responsibility for creating an integrity-supporting research environment with collective culpability for a specific act. Responsibility should be traced to roles, actions, knowledge, and applicable duties rather than assigned merely because names appear on the same project.
Junior researchers still have responsibilities, but power differences matter
Students, research assistants, and early-career researchers remain responsible for their own conduct. Being junior does not make fabrication or falsification acceptable.
Yet power relationships affect what people can reasonably do when problems arise. A doctoral student asked by a powerful supervisor to use a questionable method may face consequences that a senior colleague would not. Responsible research environments should therefore provide safe routes for advice and reporting and protect good-faith concerns from retaliation.
A junior researcher confronted with an instruction they cannot defend should not assume that seniority converts the instruction into accepted practice. The appropriate response to a methodologically or ethically questionable request may involve asking for the rationale, documenting the issue, seeking confidential advice, or escalating through appropriate institutional channels.
Institutions also carry responsibilities for research integrity
Research integrity is not sustainable if responsibility is placed entirely on individual researchers while institutions provide inadequate policies, training, infrastructure, supervision, or mechanisms for addressing concerns.
The National Academies describes integrity as dependent on both individual and collective adherence to core values and emphasizes the role of organizations in fostering environments that support responsible research. It also identifies inadequate institutional policies, procedures, or capacity for fostering research integrity as potentially detrimental practices.
Under the U.S. Public Health Service research misconduct framework, institutions receiving relevant PHS support have defined responsibilities for handling research misconduct allegations. ORI also describes HHS and covered institutions as sharing responsibility for the integrity of the research process.
Institutions cannot outsource every integrity responsibility to principal investigators any more than principal investigators can outsource everything to junior researchers.
