03 · What You Need to Know
Decide Influence One Research Decision at a Time
There is no single correct level of stakeholder influence
Stakeholder engagement encompasses different relationships. A researcher may seek focused feedback on participant materials, maintain a recurring advisory group, collaborate with partners on selected workstreams, or undertake co-production with deliberately shared power and responsibility.
It would make little sense to apply the same decision structure to all of these arrangements.
The level of influence should follow from what the project claims the partnership is for. Consultation implies that researchers seek and consider advice. Partnership implies a more substantial role. Co-production makes stronger claims about sharing power.
Problems arise when the language promises one relationship and the governance provides another.
Meaningful influence requires the possibility of changing something
A partner does not have meaningful influence merely because researchers listen politely.
If a decision has already been made and cannot change, inviting partners to “decide” it creates an illusion of authority. If every stakeholder recommendation can be rejected without explanation, the relationship may function more like consultation regardless of the terminology used.
This does not mean every recommendation must be accepted. It means there should be decisions for which partner contributions can genuinely affect what happens.
This distinction is central to recognizing when consultation begins to become tokenistic.
Different decisions can justify different distributions of influence
Instead of asking how much power partners should have over “the study” as one undifferentiated object, break the project into decisions.
| Decision |
Potential partner influence |
Important considerations |
| Research priorities |
Potentially substantial or shared |
Whose problems and unanswered questions should receive attention? |
| Research question |
Potentially substantial or shared |
Relevance, existing evidence, feasibility, ethics, and scientific contribution |
| Outcome selection |
Potentially substantial |
What matters to affected people, validity of measures, and research purpose |
| Recruitment procedures |
Potentially substantial |
Accessibility, trust, feasibility, sampling requirements, ethics, and fairness |
| Specialized analysis |
Collaborative input with technical responsibility appropriately assigned |
Methodological expertise, prespecified procedures, assumptions, and analytical integrity |
| Interpretation |
Potentially collaborative or shared |
Evidence, contextual knowledge, competing explanations, and limitations |
| Dissemination |
Potentially substantial or shared |
Audiences, accessibility, accurate communication, and partner priorities |
These are examples rather than universal allocations. A community statistician may have considerable analytical expertise. A principal investigator may also possess lived experience relevant to the topic. Roles should be assigned according to the actual people and responsibilities involved rather than stereotyped categories.
Expertise should influence decisions without becoming a hierarchy of human worth
Some decisions genuinely require specialized expertise.
A complex statistical model should not be chosen by majority vote among people unfamiliar with its assumptions. An ethics requirement cannot simply be waived because partners find it inconvenient. A validated measure should not be altered casually if doing so destroys the basis for interpreting its scores.
But specialized expertise should be applied to the question it can answer.
A statistician may determine whether a model is appropriate but may not know whether the outcome being modeled matters to patients. A clinician may understand treatment mechanisms but not the full burden patients experience. A researcher may understand sampling theory but not why a proposed recruitment location is distrusted locally.
Partnership works best when expertise gives people authority to contribute where their knowledge is relevant without turning one form of expertise into universal authority over the project.
Being affected by a decision can justify substantial influence
Expertise is not the only basis for influence. Researchers should also consider who bears the consequences of a decision.
Participants experience study burden. Communities may live with how findings characterize them. Practitioners may be expected to implement an intervention. Patient partners may experience outcomes that researchers treat as abstract variables.
Those affected interests can justify meaningful influence even when partners do not possess formal research credentials.
This principle helps explain why patient and public involvement values experiential knowledge rather than requiring public contributors to qualify as professional researchers first.
Some responsibilities cannot simply be transferred
Research partnership occurs within institutional and legal structures.
A principal investigator may carry formal responsibility to a funder or institution. Researchers may have obligations concerning participant safety, research integrity, privacy, data security, regulatory compliance, or ethics approval. Certain decisions may therefore have boundaries that cannot be changed merely through an informal partnership agreement.
Researchers should identify these constraints early.
“I have final responsibility” should not become a blanket justification for overruling partners on every issue. But “we are equal partners” should not be used to pretend that formal accountabilities do not exist.
Decision authority should be discussed before decisions become contentious
Many partnerships operate smoothly while everyone agrees. Their governance becomes visible only when they do not.
Suppose community partners want to remove a survey question that researchers consider essential. Or researchers want to publish a finding that an organizational partner considers damaging. Or patient partners want an outcome prioritized that would require redesigning the study.
Who decides?
Projects benefit from establishing decision processes before such conflicts occur. The arrangement might involve consensus where possible, delegated authority for particular workstreams, majority decisions for some matters, specified final responsibility for others, or escalation procedures when agreement cannot be reached.
No single mechanism fits every partnership.
Consensus is useful but should not become compulsory
Collaborative research often values consensus, but insisting that everyone agree can create its own problems.
Less powerful partners may remain silent to preserve relationships. A dominant organization may frame disagreement as obstruction. Researchers may keep discussing an issue until exhausted partners eventually accept the academic preference.
Sometimes disagreement is real and should remain visible.
A transparent partnership can record different views, explain how a final decision was made, and preserve dissent where it matters. The aim is legitimate decision-making, not unanimous enthusiasm.
Power includes more than voting rights
A project may announce equal voting while leaving deeper forms of power untouched.
Who controls the budget? Who receives salary for attending meetings? Who writes the agenda? Who understands the technical documents? Who owns the data infrastructure? Who can speak directly to the funder? Who controls publication submission? Who has time to attend a three-hour meeting on a weekday?
Formal decision rules matter, but material resources, information, language, institutional status, and access also determine who can exercise influence.
This is why meaningful partnership may require compensation, accessible information, preparation, capacity building, and changes to how meetings and decisions are organized.
Partners should not receive responsibility without corresponding authority
One particularly unfair arrangement gives partners substantial work and accountability while researchers retain the authority.
For example, community partners may be expected to recruit participants, maintain trust, defend the project locally, and explain controversial decisions that they had little role in making.
If partners are expected to carry responsibility for an aspect of the research, they should ordinarily have appropriate influence over the decisions associated with that responsibility.
Similarly, substantial work should prompt consideration of appropriate compensation for research partners rather than assuming partnership means unpaid labor.
Influence over interpretation does not mean control over findings
Partners can contribute substantially to interpreting evidence, particularly when they possess contextual or experiential knowledge researchers lack.
But no partner should have unlimited authority to decide what the evidence says.
The distinction between meaningful influence and outcome control is important for improving research relevance while preserving evidence-based findings.
Partners can challenge interpretations, propose alternatives, identify missing context, and disagree with researchers. Conclusions should ultimately remain defensible in relation to the data and methods.
07 · A Quick Checklist
Before Deciding How Much Influence Partners Will Have
For major study decisions, check:
What form of relationship have you actually promised: consultation, collaboration, partnership, or co-production?
Which decisions can partners genuinely change?
Whose expertise is relevant to each decision?
Who will experience the consequences of each decision?
Which ethical, methodological, legal, regulatory, or institutional responsibilities constrain decision authority?
Are partners being given responsibility without sufficient authority or resources?
Do all partners understand how disagreements will be resolved?
Can minority or dissenting views be recorded rather than forced into consensus?
Do compensation, information access, meeting arrangements, and other resources allow partners to exercise the influence they formally possess?