Running the project on your own unit: dual roles and the questions the WGU board will ask
Most WGU capstones happen where the person already works, which is exactly why they are worth doing: you know the problem, you know the people, and the improvement lands somewhere real. The board is not opposed to any of that. It is asking a narrower question — whether someone who answers to you can comfortably say no to you — and the file either answers it with arrangements or leaves the reviewer to imagine the worst.
Working on your own unit is allowed and ordinary. WGU's application asks how you are related to participants and treats those you supervise or teach as needing extra care, so the file must show who invites, who consents, and who never learns who declined.
Can you run a capstone project where you already work?
Yes, and the paperwork assumes you might. WGU's application asks you to describe the population including the relationship between participants and the person running the project, offering as examples the relationship between someone and those they teach, between a supervisor and those they supervise, and between professional colleagues. A form that asks the question outright is not a form expecting the answer "none".
The same form goes further in its protected-populations question, which lists among others those employed by, or taught by, the person conducting the project, adding that in a school setting such participants are frequently involved and must be discussed as such. The second hat is anticipated. What is not anticipated is a file that stays silent about it, then describes a recruitment method only a manager could use.
One extra gate applies where the workplace happens to be the university itself. WGU's policy sends any proposal that gathers data from its own population — those enrolled, its employees, its graduates, the employers of those graduates — or that reaches for university records about them, to institutional research leadership for approval before the board sees it, with the capstone instructor enabling that step. If your project touches WGU's own people or records, that route comes first.
What is the board actually worried about?
Not your character. The approval criteria are structural. A board must find that selection of participants is fair, taking account of the setting in which the work happens; that privacy and confidentiality provisions are adequate; and that where participants are likely to be vulnerable to pressure or undue influence, extra safeguards are built in. Consent, separately, has to be sought in conditions that give a person real opportunity to consider and to decline, with the possibility of pressure minimised.
Read those together and the concern becomes concrete. A charge nurse asking their own team to fill in a survey at handover has done nothing wrong, but has created conditions in which declining is visible, and visible declining is not free declining. The board is not asking whether you would hold it against anyone; it is asking whether a reasonable person in that room would believe you might.
What does the second hat threaten, and what answers it?
Dual-role concerns are not one problem but four, and each has its own repair. Naming which one applies to your project is most of the work; a file that fixes the wrong one reads as though nobody understood the note.
| The pressure | How it shows up | The arrangement that answers it |
|---|---|---|
| Consent that is hard to refuse | You invite people who report to you, in person, where colleagues can see | A neutral inviter and a channel you do not watch: a posted link, a distributed message from someone without authority over the group, and no register of who responded |
| Refusal that is visible | Sign-up sheets, handover recruitment, a return route that runs through you | A route where taking part and declining look identical from the outside, described that way in both the application and the consent text |
| Small numbers that identify | "The night-shift educator" or a group so small that any quote names its speaker | Report at a level that cannot be resolved to a person, avoid role labels in quotes, and say in the file how you tested that |
| Access you have for another reason | Records or systems you can open as part of your job, used for the project | Permission for the project use specifically, with the fields named and the reason recorded, rather than reliance on your everyday access |
The right-hand column is the point. Each repair is an arrangement someone else could observe, which is exactly what a reviewer can approve. Sentences promising that participation is voluntary appear in every file ever returned; alone, they close nothing.
Who should be doing what?
The cleanest dual-role files read like a division of labour. Write it out before drafting, because every document then describes the same division.
- Who invites. Ideally not you, where you hold authority. Name the person and their relationship to the group.
- Who receives. A route that does not pass across your desk, and where responses arrive without names attached.
- Who consents. The process, the wording and the moment at which someone can stop reading and walk away without explanation.
- Who never knows. State plainly which facts about participation you will not learn — who took part, who did not, who withdrew.
- Who holds identifiers. Best answer: nobody, because none are recorded. If a linkage key exists at all, it is an identifier, and WGU's application asks how personal identifiers are handled and, where participation is not anonymous or confidential, why the design requires that.
- Who decides eligibility. A criterion applied by rule rather than by your judgement is fairer and easier to describe.
Where a dual role genuinely cannot be avoided, the honest move is to say so, explain why, and add compensating protection: an independent person to take consent, a gap between invitation and participation, or a design that never records who was involved. Boards accept unavoidable roles far more readily than unacknowledged ones.
What does the facility have to sign?
Permission from someone with authority to grant it, and it belongs in the file rather than in your memory of a corridor conversation. WGU's application asks you to describe how consent is obtained from every stakeholder, naming the organisation's administrator or supervisor alongside participants, and reminds applicants that where the work needs access to an organisation's people, the appropriate administrator or supervisor signs before the file is submitted. The precise instrument, and whether that signature sits on the consent document or on separate letterhead, is set by WGU's current handbook or IRB portal, so confirm the current form with your capstone instructor.
Start it first. It moves at the speed of somebody else's inbox, and the signer usually wants to read the instrument and the consent wording before signing, so those documents must be finished before the signature can even be requested.
Be careful about a quiet complication: the person who signs may also be the person you report to. If the signature comes from your own manager, the file should show that their permission grants access rather than supplying pressure — they authorise the project, they do not recruit for it, and they do not learn who took part.
Does a dual role move the project to a heavier branch?
It can, and it is better to know before drafting. WGU's policy describes the lightest handling for proposals that do not involve research purpose or atypical methods, that do not touch protected groups or sensitive topics, and that put nobody at risk — and its own application counts those you supervise or teach among the groups to be discussed as needing extra care. A project centred on your own reports may therefore be read more closely than an identical one run elsewhere.
That is an argument for planning rather than despair. Often a small, truthful change — a neutral inviter, no identifiers recorded, a wider group than your own team — walks the same question down a lighter branch without weakening the project. Sometimes it should not, and the work deserves the closer look. The decision tree on the homepage is built for exactly this fork, and the process end to end shows what each branch asks for.
Where does this desk fit?
Dual-role files are the ones we are asked for most often, because the repair is rarely obvious from inside the unit. We name the pressure, build the arrangement that answers it, write it consistently through application, consent text and recruitment wording, then carry the file through filing and every reply. This practice is independent of Western Governors University; the project stays yours, and the reading stays the board's. The eight stations describe the engagement; the questions page covers the usual first questions.
What to do next
Send the aim of the project, who the participants are and how you are related to them. The reading is free and will name the dual-role pressure a reviewer will see, the arrangement that answers it, and whether a lighter branch is honestly available. Request the free application review. Where a board has already raised this in writing, reading the return letter comes first.
Sources
- WGU policy handbook, Institutional Review Board (IRB) — review levels and the criteria described for the lightest handling, and the institutional route required before proposals seeking data about WGU's own population reach the board: cm.wgu.edu (WGU Knowledge Center, IRB policy)
- WGU Institutional Review Board, Request for Approval to Conduct Research (published copy of the university's capstone application form) — the relationship question, the protected-populations question naming those employed or taught by the applicant, the identifiers question, consent from each stakeholder, and the reminder to obtain the administrator's or supervisor's signature before submitting: request_for_approval_to_conduct_research (PDF)
- 45 CFR 46.111 — fair selection judged with the setting in mind, privacy provisions, confidentiality of data, and added safeguards for participants open to pressure: Cornell LII, section 46.111
- 45 CFR 46.116 — consent taken where there is genuine opportunity to consider and to decline, pressure minimised: Cornell LII, section 46.116
- CITI Program, the human-subjects course outline — standard modules covering privacy, populations needing additional protection, and conflicting interests: about.citiprogram.org, HSR course
Two hats are fine. Silence about them is not.
Tell us the aim, the participants and your relationship to them. The free reading names the pressure a reviewer will see and the arrangement that answers it.
Request the free application review- who invites, and who never sees the answers
- how declining stays invisible
- whose signature grants the access