Interviews and focus groups: the guide, the recording, and the promise you cannot make
Qualitative files fail at Grand Canyon University for an unglamorous reason. The design is sound, the population is willing, the chair is satisfied — and then the board reads the actual questions, the actual recording plan and the actual consent form, and finds three documents describing three slightly different conversations. This is what each of those three has to say, and where a focus group changes the rules.
The short answer
The board reads the interview guide itself, not a description of it. Recording is a procedure with its own consent, storage and destruction. And a focus group cannot be promised anonymity, because everyone in the room hears every answer.
Do interviews and focus groups always need board review?
They always need a determination. An interview is textbook interaction with a living individual, and 45 CFR 46.102 defines interaction as communication or interpersonal contact between investigator and subject. Combine that with information a person hands over expecting it to stay out of public view, and you are inside the human-subjects definition before the recorder is switched on. GCU's own position is unambiguous: no recruiting and no data collection of any kind until the approval notice exists.
Where the file can land is a different matter. 45 CFR 46.104(d)(2) reaches interview procedures on one of three conditions — identity cannot readily be ascertained; or a disclosure beyond the study would carry no reasonable risk of legal jeopardy and no reasonable harm to reputation, employment, advancement or personal finances; or identity is recorded and a limited board review makes the finding the rule requires. Most doctoral interview work records identity by necessity, so the third branch, or a level above it, is the usual destination. The assignment is the board's to make, and the criteria behind it are worth reading in full.
Why does the board want the guide itself?
Because questions are procedures. A protocol that says "semi-structured interviews exploring participant experience" tells the board nothing it can weigh: it cannot see whether an item asks about a diagnosis, an employer, an illegal act, a colleague's conduct or a trauma. The risk in a qualitative study lives in the wording, so the wording is the exhibit.
What a reviewable guide contains:
- Every planned question, in order, in the words you will actually use.
- The probes. "Tell me more about that" is harmless; "why did you not report it?" is not, and it belongs on the page rather than in the moment.
- The opening frame — what you tell the participant before the first question, which has to match the consent form word for word.
- The stopping rules. What you do if someone becomes distressed, discloses something reportable, or names a third party.
- The referral, where the topic warrants one. A support resource named in the consent, not improvised afterwards.
"Semi-structured" is not an exemption from writing the questions down. It is a statement about how faithfully you will follow them, and the board still needs the thing you intend to depart from. Where the study is genuinely emergent, say so and describe the boundaries within which follow-up questions will stay.
How should recording be handled in the file?
As a procedure in its own right, with four sentences attached. Who records; on what device; where the file travels once the conversation ends; and when it is destroyed. Recording is also where a consent form most often over-promises, because the equipment and the software carry facts the prose forgot.
- Transcription. If a service or a tool will hear the audio, that is a disclosure. The board expects to see who they are, what they are bound by, and — where required — a confidentiality agreement in the file.
- Cloud platforms. A video call that stores recordings to a vendor's servers is storage, and it belongs in the data-security plan alongside your own drive.
- Refusal. Consent to be interviewed and consent to be recorded are separate permissions. The file should say what happens when someone grants the first and withholds the second.
- Destruction. Name the point at which the audio ends, and make sure the consent form, the security plan and the application all name the same one.
Before it approves anything, a board has to satisfy itself under 45 CFR 46.111 that privacy is adequately protected and that the data will be held in confidence. A recording plan existing only in the candidate's intentions gives it nothing to make that finding on.
What can a focus group honestly promise?
Everything above applies to a focus group and one thing more: you are no longer the only person who hears the answer.
GCU's own IRB materials define anonymity strictly — information that can never be linked to identifiers. A focus group cannot meet that standard for a structural reason no procedure can fix: every participant learns who else attended and what each of them said. What you can promise is what you will do — that you will not attach names to quotations, that transcripts will be pseudonymized, that recordings will be held and destroyed as described. What you cannot promise is what the other people in the room will do afterwards.
Files that survive review say this openly. They tell participants that others will hear them, ask the group to agree not to repeat anything outside the session, and state that the researcher cannot guarantee the agreement holds. Files that come back copied confidentiality language written for one-to-one interviews and left it there.
Three further consequences follow, and each belongs in the procedures rather than in your head:
- Composition. A group containing both supervisors and the people they supervise is a coercion problem before it is a data problem — the same authority question that governs surveys of your own staff or learners.
- Withdrawal. A participant can leave the room, but their earlier contributions are already in the transcript and already in other people's memories. Say what withdrawal actually removes.
- Venue. A room chosen for convenience can identify the group. Anywhere colleagues can see who walks in is part of the confidentiality plan whether the file mentions it or not.
Where do the limits of confidentiality belong?
In the consent form, before the first question, in words a participant can act on. 46.116 requires that consent be sought in language understandable to the person and that it describe the extent to which confidentiality of records will be maintained — the extent, not the fact. That phrasing exists precisely because confidentiality has edges.
The edges that most often apply to doctoral work are professional and legal rather than academic: obligations to report suspected abuse or neglect of a child or a vulnerable adult; a disclosure indicating risk of serious harm to the participant or someone else; conduct a licensed practitioner is bound to escalate; and, rarely, a legal demand for records. A candidate who is also a licensed clinician or a mandated reporter carries those duties into the interview room, and the board expects the consent form to say so rather than leave the participant to discover it.
The practical test is simple. If a disclosure would make you pick up the phone, the participant must have known that before they spoke.
Does the consent form have to be signed?
Not always, and this is where interview files gain a little room. Under 45 CFR 46.117(c) a board may waive the signature where the only thing tying a participant to the study would be that page itself and the main hazard is a confidentiality breach; or where the work is minimal risk and uses nothing that would ordinarily be signed for outside research. Sensitive interview studies often qualify under the first branch: a signed page bearing a real name can be the most identifying object the study creates.
Waiving the signature does not waive consent. Spoken consent still delivers the substance 46.116 requires, still gets documented as having happened, and still matches the guide and the recruitment material exactly. And where consent is spoken onto the recording, the recording now holds a name — a fact that belongs in the security plan.
| Exhibit | The board reads it as | The common return |
|---|---|---|
| Interview or focus-group guide | The procedures themselves | A description instead of the questions; probes that ask more than the guide admits |
| Consent form | A set of promises with named limits | Anonymity promised to a focus group; no mention of reporting duties |
| Recruitment material | The first thing a participant sees | A topic framed more gently than the guide asks it |
| Recording and transcription plan | A chain of custody for a voice | A transcription service nobody declared |
| Data-security plan | How identifiers live and end | Audio, transcript and consent forms described inconsistently |
| Site authorization | Permission for these procedures with these people | A letter covering interviews but silent on recording |
A reading of how the criteria are applied. Which templates GCU currently expects, and in which portal they are filed, is set by GCU's current doctoral handbook or IRB portal.
What to do next
Put the guide, the consent form and the recruitment material side by side and read them as a stranger would. If the recruitment text is warmer than the questions, if the consent promises more privacy than the recording plan delivers, or if a limit you would honor in practice appears nowhere on the page, you have found what the board would have found. Send us the AQR-approved proposal with those three documents and the free application review will tell you where the file stands; the rest is set out in how the work is handled and the answers on the record.
Sources
- GCU Research — Institutional Review Board
- GCU — Everything You Need to Know About the IRB
- GCU — What Is the Role of the Institutional Review Board?
- GCU IRB Resource Center — IRB FAQs
- 45 CFR 46.102 — Definitions
- 45 CFR 46.104 — Exempt research
- 45 CFR 46.111 — approval criteria, and the safeguards clause
- 45 CFR 46.116 — informed consent, and when it may be altered
- 45 CFR 46.117 — documenting consent, and waiving the signature
Where this article and GCU's own materials differ, GCU's current doctoral handbook or IRB portal governs. gcuirb is an independent practice and is not affiliated with, endorsed by or connected to Grand Canyon University.