03 · What You Need to Know
Different people know different things about what makes a question important
Researchers have legitimate expertise in identifying research questions
Researchers know the literature, theoretical debates, methodological possibilities, unresolved contradictions, measurement problems, and limitations of existing evidence. Those forms of expertise are indispensable.
A researcher may recognize that an apparently practical question has already been answered convincingly, that a proposed comparison cannot support the desired inference, or that a seemingly narrow methodological issue has large consequences for an entire evidence base.
Community engagement should therefore not be framed as correcting the assumption that researchers have nothing useful to contribute to priority setting. They do.
The more defensible argument is that scientific expertise may be necessary without always being sufficient.
People affected by the research possess knowledge researchers may not have
Lived experience can reveal which problems actually create burdens, which outcomes matter in practice, how institutions operate on the ground, what barriers conventional measures overlook, and whether a proposed intervention solves a problem people actually have.
This knowledge is not merely anecdotal decoration added after a literature review. For some research questions, it is evidence needed to define the problem correctly.
Formal priority-setting initiatives demonstrate how different stakeholder groups can contribute to identifying unanswered questions. The James Lind Alliance, for example, brings patients, carers, and clinicians together in Priority Setting Partnerships to identify and prioritize uncertainties about treatments and care. Its model is designed specifically to give these groups influence over research agendas rather than leaving priorities entirely to researchers and funders.
The model comes from health research and should not be treated as a universal template for education, social science, engineering, or other fields. It illustrates a broader principle: the people who experience a problem can possess relevant knowledge about which uncertainties deserve attention.
Ask what kind of question you are trying to choose
Not every research question requires the same decision process.
A highly technical question about the mathematical properties of an estimator may legitimately be driven primarily by methodological expertise. A study asking which barriers prevent a community from accessing a public service relies much more heavily on contextual and experiential knowledge. Research intended to develop an intervention for community use may require yet another arrangement.
| Type of question |
Whose knowledge may be especially important? |
Why? |
| Foundational theoretical or methodological question |
Researchers and relevant technical experts |
The central uncertainties may depend heavily on disciplinary or methodological knowledge |
| Question about population experiences or priorities |
People with relevant lived experience alongside researchers |
The importance and meaning of the problem may not be visible from published literature alone |
| Question about professional or service practice |
Service users, practitioners, researchers, and other affected stakeholders |
Different groups may see different implementation problems and outcomes |
| Intervention intended for community implementation |
Researchers, intended users, implementers, and relevant community partners |
Scientific effectiveness and real-world usefulness may depend on different forms of knowledge |
| Policy question affecting an underrepresented population |
Researchers, affected populations, policymakers, implementers, and relevant experts |
The decision combines empirical evidence with practical consequences and competing values |
The point is not to construct a committee for every hypothesis. It is to recognize when deciding what matters requires knowledge beyond the literature.
The literature can reproduce earlier researchers' priorities
Researchers are taught to identify gaps by reviewing what has already been published. That is sensible, but it creates a subtle circularity.
If previous research systematically prioritized certain questions while neglecting others, a literature-driven gap analysis begins inside the boundaries created by those earlier priorities. Researchers may identify what is missing from the existing academic conversation without asking whether the conversation itself is focused on the right problems.
This issue becomes especially relevant for populations that have historically had little influence over research agendas. A literature may contain many studies about a population while still neglecting questions that population considers important.
Research priority setting can therefore involve more than filling empty cells in a literature matrix. It can involve questioning how the matrix was constructed.
Underrepresentation does not automatically tell you what the population wants studied
Researchers should avoid another shortcut: "This population is underrepresented, so studying anything about it is inherently responsive to its needs."
Underrepresentation identifies a possible evidential problem. It does not tell you which research questions members of the population prioritize. This is why underrepresentation alone is not sufficient justification for every new study.
A population might prefer research on access to services while researchers repeatedly study attitudes. It might want evaluation of institutional practices while researchers focus on individual deficits. It might already be exhausted by descriptive studies documenting a problem everyone recognizes.
If population priorities form part of your rationale, those priorities should be established rather than assumed.
Community priorities are not automatically singular or obvious
"Ask the community" sounds simple until researchers have to decide whom to ask.
Communities contain disagreements, inequalities, institutions, generations, professions, political interests, and different experiences of the issue. Formal leaders may not represent people with less institutional power. Advocacy organizations may emphasize different concerns from unaffiliated community members. People who have time to attend research meetings may differ from those whose circumstances make participation difficult.
Community-engagement guidance therefore cautions against assuming that a single representative can speak for an entire population. CIOMS recommends understanding community dynamics and power inequities and seeking participation from relevant sectors of the community. This guidance was developed for health-related research, but the underlying representational problem is broader.
Community involvement does not reveal one authentic answer waiting to be discovered. It brings additional perspectives into a decision that may remain contested.
Research priorities involve values as well as evidence
Choosing what to study is partly an empirical decision and partly a value judgment.
Researchers can estimate prevalence, characterize uncertainty, evaluate feasibility, and identify potential consequences. Those data can inform priority setting. They cannot, by themselves, determine whether reducing one uncertainty is more important than addressing another.
Priority decisions may involve severity, number of people affected, inequality, feasibility, urgency, scientific opportunity, cost, neglectedness, and likely benefit. Different stakeholders may reasonably weight these considerations differently.
Making those judgments explicit is usually preferable to presenting research priorities as though they emerge mechanically from a literature review.
Researchers still have responsibilities they cannot delegate
Community involvement does not remove researcher responsibility for scientific validity, ethical conduct, methodological appropriateness, or honest interpretation.
A community may strongly prefer a question that available methods cannot answer credibly. Stakeholders may request conclusions the evidence cannot support. Different groups may advocate incompatible priorities.
Researchers should explain these constraints rather than pretending every preference can become a feasible study. Shared priority setting works best when methodological expertise and experiential knowledge constrain and inform each other.
Funders and institutions also shape what gets studied
Researchers and communities are not the only actors setting priorities. Funding calls, journal interests, institutional strategies, policy agendas, commercial incentives, available datasets, and career structures all influence which questions become research projects.
This matters because a study described as "community-driven" may still be constrained by what a funder is willing to support. Conversely, funding mechanisms can deliberately create opportunities for populations and stakeholders to influence research priorities.
Transparency is useful here. Researchers should distinguish questions that genuinely emerged from community priority setting from questions developed academically and later refined through consultation.
Early involvement matters when priorities are supposed to influence design
If researchers claim that community priorities shape the study, involvement must occur while the research question can still change.
Seeking feedback after funding, protocol development, and ethics approval may improve implementation but cannot retroactively turn the original question into a community-selected priority.
This is why some population-specific questions require community involvement before study design.
The timing should match the claimed influence.
Priority setting can improve fairness without making every decision democratic
Inviting affected populations into research priority setting can change whose concerns enter the evidence agenda. This may be particularly important where research-based decisions affect populations that have historically had little influence over the knowledge used to govern services, interventions, or policies.
That can contribute to fairer population research. Yet fairness does not necessarily require every scientific decision to be settled by vote.
Different decisions may appropriately allocate authority differently. Communities may have strong influence over which outcomes matter, researchers over analytical validity, ethics bodies over participant protections, and funders over resource constraints. The important issue is whether the allocation of decision-making authority is justified rather than merely inherited.
Beware of token priority setting
Researchers sometimes invite community members to a workshop, ask them to rank a predetermined list of researcher-generated questions, and then describe the resulting project as community-led.
The exercise may still be useful, but the description overstates what occurred.
Meaningful priority setting requires clarity about who generated candidate questions, who participated, how priorities were ranked, how disagreements were handled, what criteria were used, and what influence the results actually had over the research agenda.
Watch Out
Do not invite people to identify research priorities if there is no realistic pathway for their priorities to influence what gets studied. Consultation without decision relevance can create participation burden while giving researchers little more than an engagement paragraph for the methods section.
Sometimes the most important question challenges the researcher's original framing
A researcher may approach a population wanting to understand why people fail to use a service. Community members may respond that the more important question is why the service was designed in a way that excludes them.
That shift can be intellectually uncomfortable because it changes the unit of explanation. It may also produce a stronger study.
This is one reason researchers should remain alert to deficit framing in population-specific research. Who participates in setting the question can influence where researchers look for the problem.
