Researching your own workplace or your own patients at Walden: conflict of interest, dual roles, and the board's questions

Why does the board care where you work?

Two ideas converge on your building. The first is undue influence: 45 CFR 46.116 requires consent to be sought only under conditions that minimise the chance of coercion or undue influence, and 45 CFR 46.111 requires extra safeguards wherever participants are open to exactly that pressure. An invitation from a person who writes your schedule, signs your evaluation, or manages your care is never just an invitation — the relationship answers before the person does. The second is Walden's own posture. Its public guidance treats the researcher's professional role as a standing source of pressure — a manager recruiting subordinates is the example its own delays FAQ gives — and its list of people needing special care includes a researcher's patients and a researcher's subordinates. Its standing advice to anyone planning to collect data in their own workplace: bring the dual roles to an IRB office-hours consultation early, because managing them is a design question, not a disclosure formality.

None of this makes insider research improper. It makes it a design problem — and the board's questions are the specification. Your training points the same way: CITI's human-research curriculum carries modules on research involving workers and employees, and on conflict of interest in studies with people. The board assumes you have met these ideas; the file should read like it.

What is a dual role, exactly?

You hold a dual role whenever the person a participant meets in the study is also a person who holds standing in their life outside it. The pairs the Walden board sees most:

Only one of these pairs involves formal authority; peer pressure, care relationships, and evaluative roles raise the same structural concern, which is why "I am not anyone's supervisor" rarely ends the conversation.

Which questions will the board put to you?

When a Walden application names the researcher's own site, the board's letter tends to circle the same territory. Draft your answers before filing; better, build them into the application so the letter never needs writing.

• What role do you hold at the site, and over whom? Plain words: title, reporting lines, whether any prospective participant reports to you, is graded by you, or receives care from you.
• Who recruits? If the honest answer is "I do, in person, to people who know me," expect the board to press. The strong answer names a neutral channel or a third party.
• Who consents, and where? Consent taken by the manager in the manager's office fails the circumstances test no matter what the form says.
• How does a refusal stay invisible to you? This is the question beneath the others. If you can see who declined, the pressure exists whether or not you would ever act on it.
• Who sees identifiable data? Especially where you would recognise handwriting, voices, or stories. "The researcher, with codes" is often not enough when the researcher knows the room.
• What does the site say? Written permission from someone with authority over you and the setting — not from you, and not from a peer doing you a favour.
• What are participants told? The consent document must say, in plain language, that taking part or declining changes nothing about employment, evaluation, care, or standing — the voluntariness language 45 CFR 46.116 requires, aimed at your specific relationship.

What does a separation plan look like?

The plan is a short set of design choices moving power out of the researcher's hands wherever the roles touch. Not every study needs every device; every insider study needs some, chosen to fit where your authority sits.

1. Neutral recruitment. The invitation reaches people through a channel that is not you: a site coordinator forwards the email, a flyer sits in the break room, a staff newsletter carries the notice. No face-to-face asks from the person with the dual role, and no recruiting inside supervised time.
2. Invisible refusal. Interested people contact the study; the study never contacts a named list you can watch. For surveys, an anonymous link with no completion roster. For interviews, scheduling that does not pass through your office.
3. Consent at arm's length. Where authority is real, someone without it conducts consent — or, at minimum, consent happens in private, in writing, with explicit language that the dual-role holder will never know who declined.
4. Blinded or brokered data. A data steward at the site pulls and de-identifies records before you see them; a transcriptionist strips names before analysis; codes are held by someone who is not you where the design allows it. Access you hold as an employee is used only as the approved file describes — a chart you may open for care is not thereby open for research.
5. Roles on paper. The application states which hat you wear at each step, and the consent document repeats it in participant-facing words: when you are the researcher, you are not the manager, the clinician, or the instructor — and what participants say cannot follow them into that other relationship.

Each device answers a specific board question, which is why the plan belongs in the application, not in your reply to the board's letter. A file arriving with the separation designed reads as a researcher who saw the problem; one without it reads as hope. The difference is usually a full review cycle — and if the letter has already come, the plan is still the answer: fold it into the response as answering Walden IRB revision requests lays out.

What about your own patients?

The clinician-researcher pair deserves its own paragraph, because the pressure runs deepest where care is involved. A patient asked by their own provider hears the request inside the care relationship — agreeing feels like being a good patient, declining feels like a cost — so boards look for recruitment and consent arriving from outside that relationship, and for consent language severing study from care in words a worried person can believe: care continues identically either way, and nothing said in the study reaches the chart. Records raise a second layer: identifiable health information sits under privacy rules binding the site itself, so record access runs through the site's own approval — the letter of cooperation naming which records, and whatever agreement the site's policies require for data leaving its systems. Walden's red-flag guidance also names the honest fallback when separation cannot be built: a secondary analysis of data the site already creates in its own operations. And where the project reads as practice improvement rather than research, the determination — as always at Walden — belongs to the board, not to the person hoping to skip a step.

Does an insider study need anything extra in the file?

Nothing exotic — the standard enclosures, each carrying the separation plan where it touches them. The application describes your role and the plan; the recruitment materials show the neutral channel doing the inviting; the consent document carries the voluntariness and no-consequences language aimed at your actual relationships; the letter of cooperation is signed above you; the data-security plan names the steward, the de-identification step, and the code key's holder. The application checklist walks the full enclosure set, and the consent form requirements cover the elements the board reads most closely — read both with your dual role in mind, because at an insider site every document is also a document about you.

What to do next

Write the honest paragraph first: your title, who reports to you, who you treat or teach among the people you hope to study, and what you can already see or open at work. That paragraph decides how much separation the design needs — and whether this is a file you can reasonably carry alone. If you would rather have it read cold, that is the desk's first move on every insider file: we make the determination call, write the dual-role plan, build every enclosure to carry it, and answer the board until the approval letter arrives — how it works describes the engagement. Send your paragraph and whatever documents exist through the free application review, and we will tell you plainly how exposed the file is and what the plan must contain — before the board asks in writing. The study stays yours; the decision stays the board's.