01 · The Question
How Can Two People Hear the Same Research Invitation So Differently?
“Would you like to participate in my study?”
Asked by a stranger, that may be an uncomplicated invitation.
Asked by your professor, supervisor, physician, thesis adviser, caseworker, employer, or someone who controls an opportunity you need, the same words can carry considerably more weight.
Nothing in the sentence identifies the difference. The difference exists in the relationship.
Power affects research participation because people do not make decisions in a social vacuum. Researchers and potential participants may differ in authority, expertise, institutional status, economic resources, control over services, professional influence, or the ability to impose or absorb consequences.
The presence of a power difference does not automatically invalidate consent. It does mean that researchers should examine whether the less powerful person can realistically choose differently from the more powerful person's preference.
03 · What You Need to Know
Power Is More Than Formal Authority
Start with control, not titles
Researchers often identify power relationships through familiar categories:
teacher and student, employer and employee, physician and patient.
Those categories are useful because the authority is easy to see. But power can exist without a formal supervisory title.
A senior academic may influence a junior researcher's career. A community leader may determine access to local organizations. A service provider may help someone navigate essential benefits. A researcher may control authorship, funding, laboratory access, recommendations, visas, opportunities, specialist knowledge, or entry into a professional network.
The useful question is therefore:
What can this person influence that matters to the potential participant?
That question reveals forms of power that an organizational chart may miss.
The Belmont Report connects authority directly to voluntariness
The Belmont Report states that voluntary consent requires conditions free of coercion and undue influence. It specifically observes that unjustifiable pressures commonly occur when people in positions of authority or commanding influence urge a course of action, particularly where possible sanctions are involved.
This does not mean authority itself is unethical.
Teachers need authority to teach. Supervisors need authority to manage. Clinicians possess expertise and responsibility within care. Organizations need decision structures.
The ethical problem arises when power belonging to one relationship leaks into the research decision.
Power can operate through punishment
The clearest mechanism is a threatened disadvantage.
A professor threatens a lower grade.
A supervisor threatens loss of leave.
A service provider threatens loss of an entitled program.
A clinician threatens ordinary care.
OHRP describes coercion as an overt or implicit threat of harm intentionally presented to obtain compliance. It gives threatened loss of needed health services as one example.
Here, the powerful person uses control over a valued outcome to make participation the route to avoiding harm.
Power can also operate through reward
The same hierarchy can influence participation in the opposite direction.
Instead of threatening disadvantage, an authority figure can make participation appear to unlock advantage.
A supervisor may control promotion. A professor may offer academic credit. A program administrator may offer access to an opportunity.
OHRP distinguishes these mechanisms. Its employee guidance gives loss of employment benefits as an example associated with coercion while noting that a decision to participate leading to job promotion may constitute undue influence. Its student guidance permits reasonable academic rewards in appropriate circumstances but requires comparable non-research alternatives where specified.
This is why power analysis should ask both:
“What could this person lose by refusing?”
and:
“What could this person gain only by agreeing?”
Power can operate through uncertainty
Sometimes nobody has promised or threatened anything.
The participant simply does not know what the powerful person will think.
OHRP's employee guidance is especially revealing here. It recognizes that employees may question whether participation affects performance evaluations or job advancement even when this is only the employee's perception. Employees may view employers as authority figures to whom they must show deference, undermining freedom of choice.
Uncertainty itself can therefore influence behavior.
If the participant thinks, “I doubt my supervisor would punish me, but why risk it?” the power relationship has already entered the decision.
Dependency magnifies power
The same authority may matter differently depending on how dependent someone is on it.
A patient with many alternative clinicians may experience a research invitation differently from someone who depends on a single specialist. An employee in a secure position may experience a supervisor differently from someone whose contract expires next month. A student taking one elective course may be situated differently from a doctoral student whose adviser controls progression and recommendations.
The 2024 Declaration of Helsinki recognizes dependent relationships as requiring particular caution in medical research. When a potential participant is in a dependent relationship with the researcher or may consent under duress, informed consent must be sought by an appropriately qualified individual independent of that relationship.
Dependency is therefore not simply a characteristic of the powerful person. It describes how consequential that relationship is to the less powerful person.
Expertise is a form of power too
Power is not limited to sanctions and rewards.
Researchers often possess knowledge participants do not. Clinicians understand medicine. Scientists understand the study design. University researchers may be perceived as experts. Technical terminology can make participants dependent on the researcher's explanation of what the study means.
Expertise is necessary for research, not an ethical defect.
But information asymmetry can become problematic if researchers use their credibility to discourage questions, exaggerate benefits, minimize uncertainty, or make their recommendation seem beyond reasonable disagreement.
Informed consent exists partly because participants need understandable information to make their own decision rather than simply deferring to the expert.
Institutional power can matter even when the individual researcher has little authority
A junior researcher may personally control nothing important about a participant.
The researcher's institution might.
Imagine recruitment conducted inside a university, hospital, employer, government agency, school, residential facility, or service organization. Participants may respond not only to the individual recruiter but to what the institution represents.
A survey bearing an employer's logo may look official. A study distributed through a school administration may resemble a required school activity. Research conducted in a clinic may look like part of care.
Researchers should therefore consider institutional as well as interpersonal power.
Social status can affect whose preferences dominate
Power differences can also arise from social or community structures that are not formally part of the research organization.
A respected community leader may strongly influence members. A family elder may exercise substantial authority. A senior professional may shape norms within a field.
The Belmont Report even identifies manipulation through the controlling influence of a close relative as an example of undue influence.
The ethical analysis should therefore avoid assuming that only employment contracts and academic grades create meaningful power.
Gatekeepers can transfer their authority into recruitment
Researchers frequently need permission from organizations or community representatives before approaching potential participants.
Permission to access a population is not the same thing as consent on behalf of the people within it.
A school principal may permit recruitment without being able to volunteer students. A company may authorize a study without deciding that employees must participate. A community leader may welcome researchers without obligating community members.
If the gatekeeper communicates research as an expectation, their authority can create pressure even when the research team does not.
Power differences can affect withdrawal as well as enrollment
A participant may successfully say yes voluntarily and later discover that saying no is harder.
They may want to withdraw but worry about disappointing the professor, clinician, employer, or researcher who invested time in them.
Researchers may also unintentionally invoke sunk costs:
“You've already completed five visits.”
“We really need your final interview.”
“Your data will be unusable if you leave now.”
Those statements may describe genuine research inconvenience, but they can also make the participant feel responsible for protecting the study.
Power-sensitive research preserves voluntariness throughout participation rather than only at enrollment.
Power differences are relational, not fixed participant traits
It is tempting to classify some people as “vulnerable participants” and everyone else as ordinary autonomous adults.
That can oversimplify the problem.
A person may be powerful in one setting and dependent in another. A senior executive can become a dependent patient. A professor can become a student. A financially secure person can depend heavily on a specialist service.
OHRP requires additional safeguards when participants are likely to be vulnerable to coercion or undue influence, while its guidance repeatedly emphasizes the contextual nature of influence.
Researchers should therefore examine the particular relationship and circumstances rather than treating vulnerability as a permanent label attached to certain categories of people.
A power difference does not automatically make participation involuntary
This qualification is essential.
If every unequal relationship made voluntary participation impossible, enormous areas of educational, organizational, clinical, and community research would become ethically impossible regardless of safeguards.
The more defensible approach is proportional.
Identify the power. Determine how it could affect the decision. Reduce avoidable connections between that power and participation. Add safeguards appropriate to the remaining risk. Then allow capable people to make their own decision where the applicable ethics framework permits it.
Watch Out
Do not treat “power imbalance” as a magic phrase that automatically invalidates consent. Specify the power: what does one party control, how does the other party depend on it, and how could that relationship alter the consequences of participation, refusal, or withdrawal?
