Surveying staff or colleagues on your own unit: consent, coercion and what anonymity really claims

Yes, you can usually survey the people you work beside, and most WGU capstones that do are approved. What changes is not permission but scrutiny: the reviewer reads the relationship before the questionnaire. Who does the asking, through which channel, what visibly happens to a colleague who declines, and whether the anonymity your recruitment promises is still true once a small unit, a reminder list and a survey platform have all had their say.

Emma Sorensen, MSN, RN · 23 August 2026

The short answer

Surveying colleagues is routinely approvable. The board tests three things: that consent is sought where coercion and undue influence are minimised, that declining is genuinely invisible and costless, and that any anonymity you claim is true of the data as collected, not merely intended.

Why does the relationship matter more than the questions?

Because the rule says so before it mentions instruments at all. A request for consent, says 46.116(a)(2) of the Common Rule, is legitimate only where the person has a real chance to weigh it and where the possibility of "coercion or undue influence" is minimised. The approval criteria at 46.111(a)(3) direct the board's attention to anyone "vulnerable to coercion" of that kind, and 46.111(b) demands extra safeguards once such people are in the room.

Nothing in those sentences is about topic sensitivity. They are about the room. A charge nurse asking her own team to complete a questionnaire is asking people whose schedules she writes. That does not make the project improper; it makes the recruitment wording, the channel and the response record the three things a reviewer reads hardest.

What does coercion look like on a real unit?

Rarely like a threat. Almost always like ordinary workplace momentum that nobody thought about.

The link goes out from a manager's account, so declining means declining a manager.
The ask happens in a huddle, where everyone sees who reaches for a phone and who does not.
A completion list is posted to "help us reach the target", turning participation into attendance.
Reminders go to named non-responders, which announces that non-response is visible.
Participation is arranged on shift, so opting out reads as opting out of the unit.
A prize is offered on a unit where such things are scarce, which is influence rather than thanks.

Each of those is fixable, and the fix is usually structural rather than verbal. A neutral distributor rather than a supervisor. An open window rather than a huddle. Reminders to everyone, worded so they single nobody out. No completion tracking when the design does not need it.

Which exempt category does a staff survey usually sit in?

Survey procedures are named in exempt category 46.104(d)(2), which opens on any one of three conditions. Responses are recorded so that no one's identity can readily be worked out. Or disclosure outside the research would pose no reasonable risk of "criminal or civil liability", nor damage to a person's standing, employability, advancement or reputation. Or the responses are identifiable and the board runs a limited review making the privacy determination at 46.111(a)(7).

Read the middle condition closely, because a workplace survey is precisely where it stops helping. Ask colleagues how safe they feel raising concerns, how often a protocol gets skipped when the unit is short, or how badly handover fails, and disclosure outside the research plainly could damage employability or reputation. Two workable routes remain: record responses so that identity cannot be recovered, or accept identifiability and let the board make the judgement it exists to make. Picking the first and then collecting in a way that quietly breaks it is the commonest reason these files come back. Those routes are distinguished in the review levels piece.

How the relationship changes what the file has to carry
Who you are to them	What the reviewer worries about	What usually resolves it
Peer on the same unit, no authority over anyone	Social pressure and small-cell exposure	Open recruitment, no tracking, reporting rules for small groups
Charge nurse, educator or manager of the respondents	Declining is visible to whoever schedules them	A neutral distributor, no access to who responded, explicit no-penalty wording
Preceptor or evaluator of the respondents	Participation reading as part of assessment	Separated in time and channel from any evaluation, and said so plainly
Respondents are also your own patients	Care and consent becoming entangled	Plan before drafting; this often belongs on the talk-first branch
Respondents are WGU's own population	Institutional permission that precedes the board	Institutional Research approval first, then the IRB submission

Is your survey actually anonymous?

Anonymity is a claim about the dataset, and it is broken by four ordinary things far more often than by anything dramatic.

The platform. Many tools collect email addresses, IP addresses or respondent identifiers by default, and some stamp a time precise enough to match a shift roster. Switch those settings off deliberately, and say in the application that you did.
The demographics block. Role, shift, unit and service length look harmless one at a time. Together, on a unit with one night-shift educator, they are a name. Decide your smallest reportable group before writing the block.
The reminder list. If you can nudge only the people who have not responded, you hold a linkage between person and participation, which is the thing anonymity denies exists.
Free text. One vivid sentence about a particular night identifies its author to every colleague who reads the finished project. Say how quotes are handled, or that there will be none.

A plainer word is available when anonymity is not achievable: confidential. A confidential survey says you could link responses to people but will not, then describes what protects that promise. Reviewers accept confidential designs routinely. What they return is an anonymous claim the data plan contradicts.

What does the recruitment text have to say?

The elements at 46.116(b) apply to a questionnaire as much as to anything else, and four carry most of the weight in a workplace file: that this is research and what taking part involves (b)(1); the foreseeable risks, which here means the risk of being identified (b)(2); how far confidentiality will hold, stated as an extent rather than a promise (b)(5); and that taking part is voluntary, refusal costing nothing a person is otherwise owed (b)(8). Consent must also open with the "key information" someone needs in order to decide, under 46.116(a)(5)(i), which in a colleague survey puts the voluntariness sentence near the top rather than on the second page.

Two more provisions come up constantly. A signature may be waived under 46.117(c)(1) on either of two grounds. One: that signed form would be the single document connecting a named person to your project, its principal risk being a breach of confidentiality. Two: minimal-risk work involving nothing that would normally be signed for outside research. In an anonymous staff survey, collecting a signature manufactures the identifier you promised not to hold. Request the waiver explicitly, attach the information sheet people will see, and note that 46.116(a)(6) forbids exculpatory language anywhere.

What if the colleagues are at WGU?

Then a second gate opens before the board's. WGU's published policy requires that research collecting data from WGU's own population, or using extant university data about those groups, be approved by the Vice President of Institutional Research or a designee before the IRB submission, with the capstone instructor enabling that submission. Build the sequence into your plan from the start; the details of how that shapes the order of work are in the piece on gates and passes.

The checklist a staff-survey file has to pass

The recruitment text names who is asking, and makes clear they evaluate nobody.
The channel does not run through anyone with authority over respondents.
Declining is invisible: no tracking, no named reminders, no completion list.
The voluntariness sentence appears early, in plain language.
The instrument is attached as respondents will see it, demographics block included.
Every anonymity claim is matched by a platform setting that keeps it true.
A smallest-reportable-group rule exists and is written down.
Any incentive is small enough that nobody takes part mainly to receive it.
Site permission covers surveying its people, not merely being on the premises.

That last point catches people late. Permission to run a project at a facility and permission to approach its workforce are two different grants, and the letter must name the second explicitly. What it should say, and who can sign it, is covered in the site permission letter piece.

What to do next

Write the recruitment text and the anonymity claim before you write a single question. Those two paragraphs decide the branch; the questionnaire rarely does. Send them to the desk along with the platform you intend to use, and you will get back a plain reading of whether that anonymity claim survives your own unit, what the folder must carry, and whether the draft is clean enough to file without us. The free application review costs nothing; the coworkers question covers the ground more briefly.

Sources

WGU Knowledge Center: Institutional Review Board (IRB) — WGU's published IRB policy, including the Institutional Research approval that precedes IRB submission for data about WGU's own population.
45 CFR 46.104 — the exempt categories, (d)(2) among them.
45 CFR 46.111 — the approval criteria, fair selection, and extra safeguards where influence is a risk.
45 CFR 46.116 — consent: its elements, and the setting a request may be made in.
45 CFR 46.117 — putting consent on paper, plus the two grounds for waiving a signature.
WGU Leavitt School of Health requirements — WGU's statement that host-site interactions may include project development or conducting research.

Forms, portals and expectations change. Confirm them in WGU's current handbook or IRB portal rather than in any summary, this one included. We are independent consultants, not affiliated with Western Governors University.

Keep walking

Interviews and focus groupsguides, recording, confidentiality limits Existing records and secondary datawhen a dataset becomes human subjects How it worksthe process, off your desk Free application reviewsend the recruitment text

Write the recruitment text first. It decides the branch.

We build the consent file, align the instrument, file the submission and answer each reply until the letter arrives. What you research remains yours. What the board concludes remains its own.

Request the free application review

What the review answers

whether declining is genuinely costless
whether the anonymity claim survives your unit
whether you can file it yourself