Surveying employees or coworkers for a Purdue Global study: consent, coercion and anonymity
The easiest sample in the world is the one already sitting at the desks around you. It is also the one Purdue University Global's board looks hardest at, and the university's own criteria worksheet says so in plain language rather than leaving it to inference. A workplace survey is not forbidden. It has to be drawn so that a person can decline and nothing happens — and so that the file proves it, in the same words, in every document the reviewer opens.
Short answer
Yes, but under conditions. Purdue Global's criteria worksheet says convenience samples are generally not allowed, and a supervisor or anyone with authority over someone's employment must never recruit them. Justify the population, route the invitation elsewhere, and make anonymity a mechanism rather than a claim.
Why is a workplace survey treated as a harder problem?
Because voluntariness is a property of the situation, not of the sentence asserting it. A consent document can say participation is voluntary; whether that is true depends on who asked, who can see who replied, and what the person believes happens if they sit it out.
The federal criteria name this directly. Under 45 CFR 46.116, the circumstances of an ask must give somebody a real chance to weigh taking part, and must hold coercion or undue influence to a minimum. Under 45 CFR 46.111, selection must be equitable, and the board is told to watch for groups whose situation exposes them to pressure. Employees are not a listed vulnerable population — but the pressure those rules guard against is precisely what an employment relationship supplies.
Purdue Global does not leave the inference to you. Its criteria-for-approval worksheet carries an "Additional Considerations" block on recruitment that reads like a list of the findings a workplace study attracts.
What does the board's own worksheet say about employees?
- Nobody in authority may do the asking. Employees should never be recruited by a supervisor or anyone in authority over their employment. If you manage the unit you cannot send the invitation — and "I only forwarded it" is still the ask.
- Participation cannot be attached to employment. Employees should not be encouraged to take part in workshops or career training forming part of a research project as a condition of their employment.
- Results stay out of the employment file. Research activities or results should never be used by employers to aid performance evaluations or other employment decisions — a commitment the site letter should carry too.
- Convenience is not a rationale. Employees should not be selected solely on the basis of convenience.
Why is a convenience sample a finding rather than a shortcut?
The same worksheet states that recruiting samples of convenience is generally not allowed, and gives two examples: using departmental email lists, and recruiting a known group from a workplace or organisation familiar to the principal investigator. That describes most workplace studies as first drafted.
The worksheet also supplies the way through. Recruitment should align with the study's stated aims, so the recruited population is the pool the questions actually need; and where a question requires a specific pool, that must be justified. The file does not have to abandon the setting. It has to argue for it — here is why that pool is the right one rather than the near one.
Two further items in the same block belong in the protocol from the start. The total number of participants must be stated and justified — the worksheet suggests citing published work using similar methodology, or a sample size estimation. And the ages of all participants must be listed.
Is an anonymous survey exempt?
Often, though the exemption has conditions and the board decides. Purdue Global's exemption worksheet reproduces the survey and interview category, and its criteria are alternatives — you need one, and the file must name which.
| Criterion | What it means here | What the file must show |
|---|---|---|
| Not readily identifiable | Responses recorded so that nobody's identity is readily ascertainable, whether directly or via linked identifiers | Platform settings suppressing addresses and network identifiers; no name, no employee number |
| Disclosure would not harm | Disclosure outside the research would not risk liability or damage financial standing, employability, advancement or reputation | An honest read of the item bank — questions about a manager, an error, or an intention to leave rarely clear this bar |
| Identifiable, with review | Identity can readily be ascertained, and the board reviews the privacy and confidentiality provisions | Named storage, named access, a stated point at which identifiers come off, and a destruction plan |
| Straightforward procedures | Purdue Global's worksheet adds a criterion for straightforward survey, testing or interview procedures with no intervention or experimental design | A design with no manipulation — and a check against the current worksheet, since this is not phrased identically in 45 CFR 46.104 |
That last row is a real difference worth knowing about rather than tripping over. The Common Rule's survey exemption sets out three alternative criteria; Purdue Global's worksheet lists a fourth. Claim the criterion the university's current worksheet supports, and expect the reviewer to read your design against it.
The worksheet is also candid about misuse. Studies where people take part in a workshop or an intervention as part of the protocol do not qualify — even if the only measurements are surveys before and after. Pre- and post-surveys with a manipulation in between are named as the classic mistake, and many workplace projects are that shape.
Consent form or information sheet — which one?
Purdue Global runs two different participant-facing documents, and the pathway decides which you draw. The Investigator Guide pairs the exempt protocol with an Exempt Research Information Sheet template, and the expedited protocol with an Expedited Informed Consent Form template. The consent procedure confirms the logic: exempt protocols require a written statement about the research, such as an information sheet, in lieu of an informed consent document.
Whichever you are drawing, the exemption worksheet lists what an interaction-based study must disclose — in an information sheet, in the instrument's preamble, or verbally: that the activity involves research, the procedures, that participation is voluntary, and the investigator's name and contact details. For a non-exempt study the elements are longer: purpose, expected commitment, procedures, foreseeable risks, benefits, the extent of confidentiality, whom to contact, and that refusal carries no penalty.
Two mechanics catch people out. Exculpatory language — anything through which somebody appears to waive a legal right, or that appears to release the investigator or the university from liability for negligence — is not permitted. And after approval, the protocol number and expiration date go in the header or footer of every approved consent document; only approved forms may be used.
What has to be true for "anonymous" to survive review?
Anonymity is a set of mechanisms, and the reviewer reads it as one. Each item below is where a workplace survey quietly stops being anonymous.
- The platform. Response tracking, address capture and network logging are often on by default. Name the setting, not the vendor.
- The demographics block. Unit, role, tenure and shift re-identify one person in a small team faster than a name would. Collapse the categories, or drop them and say why.
- Free text. People identify themselves in comment boxes, and colleagues too. Say how comments are handled before analysis.
- Small cells. Decide the minimum group size you will report, and put that figure in the protocol and the participant document.
- Reminders. A reminder sent only to non-responders is a tracking mechanism. If the design is anonymous, reminders go to all.
- Distribution. A link posted where all can reach it beats an invitation from somebody who can see the list — and it answers the recruitment concern too.
The tolerance rule applies to every one of these. Whatever you decide must read identically in the protocol, the application, the participant document, the recruitment text and the site letter. A file promising anonymity in one document and describing a follow-up to non-responders in another has measured the same dimension twice and reported two answers.
What the site has to sign
If the survey reaches people at an organisation you do not control, Purdue Global's parameters for letters of collaboration apply: a letter on the site's letterhead from an authorised representative, naming the activities occurring there — the guidance gives distributing surveys to site employees as its own example — and the support the site will provide. The letter does not substitute for IRB review; it is part of the application. For a workplace study it is also the natural place to record that participation is voluntary, sits outside any employment process, and produces nothing that reaches a performance file.
What to do next
A workplace survey is usually approvable and rarely approvable as first drafted, because the drafting instinct — ask the people I can reach, promise anonymity, send it from my own address — collides with three separate lines in the board's own worksheet. Send us the item bank, the draft invitation and whatever the site has put in writing; a consultant reads it the way Purdue Global's reviewer will and returns a written findings list. Hand the process over and we carry it whole: pathway, drafting, filing, and each response reviewers ask for until the letter lands. Your question and your findings never move, and the verdict is not ours to give — only the board issues that. Start with the free application review, see what an end-to-end engagement covers, or browse the questions candidates ask most.
Sources
- PG IRB, IRB-314 Worksheet: Criteria for Approval — docs.google.com
- PG IRB, IRB-312 Worksheet: Exemption Determination — docs.google.com
- PG IRB, SOP 320: Informed Consent Requirements — docs.google.com
- PG IRB, 3_Parameters for IRB Letters of Collaboration — docs.google.com
- PG IRB, Investigator Guide — purdueglobalwriting.center
- PG IRB, IRB Toolkit — purdueglobalwriting.center
- 45 CFR 46.104, Exempt research — ecfr.gov
- Criteria for approval of research, 45 CFR 46.111 — ecfr.gov
- 45 CFR 46.116, Informed consent — ecfr.gov