# Turn a voluntary practitioner interview into an accurate source story

Original constructed teaching cases with complete context. No real financial records, sponsorships, contracts, people, interviews, tests, preferences or payments.

Start with a genuine learning question, agree the participation, recording and publication scope, then ask neutral questions about an actual example. Keep the person’s words separate from your interpretation and review the proposed quote/context within that agreement. One interview can explain an experience; it does not establish audience preference, prevalence or a customer result.

## Choose one source question and a voluntary participant

A useful question might be how a practitioner checks whether two reporting columns mean the same thing. Seek someone with relevant actual experience, not a person selected because they will praise your product. GOV.UK’s interview guidance emphasizes real examples, careful planning and neutral questions; it is not a method for extracting a predetermined endorsement.

Before any real session, explain purpose, expected activities, observers, recording/transcription choices, intended use, access and retention, and how participation or consent can be withdrawn. Agree what can become public, separately from participation itself. Use the appropriate actual consent/data process; this guide’s exercise is not a universal legal form or permission to contact people.

## Use a complete neutral discussion guide

Constructed introduction: “I want to understand how you check a reporting handoff, not assess your performance. You can skip questions or stop. We will use notes only within the agreed purpose. Recording and public quotation need their separate agreed choices. Please avoid names or private customer material. Is that still suitable before we begin?” Adapt the actual terms honestly rather than promising a process you cannot provide.

Original question sequence: describe the last relevant handoff; what did you first inspect; where did a term or date become unclear; what source helped resolve it; what did you change or leave unresolved; what happened next that you actually observed? Follow up on the participant’s words and uncertainty. Avoid leading prompts such as “How much faster did our tool make you?” when no improvement was established.

## Separate source notes, quotes and interpretation

Keep a source reference and distinguish a verbatim transcript, your summary and your interpretation. Mark inaudible or uncertain wording instead of completing it from memory. A consent to participate does not automatically permit a public recording, named quotation, translation or unrelated marketing use.

Before a proposed public story, check the exact excerpt and the surrounding meaning within the agreed review scope. Preserve qualifications and the limits of the participant’s knowledge. Removing a name alone may leave a recognizable project; keep a story private when the required context or sharing basis is unresolved.

## Turn permitted notes into a traceable finding

GOV.UK’s analysis guidance separates observed/source notes, grouped patterns, interpreted findings and next actions. Keep the source/version with each note, identify your interpretation explicitly and check the proposed public wording within the existing agreement. Grouping is an analytical aid; it does not remove selection, memory or researcher bias by itself.

The original four-note exercise below is entirely constructed, not four interviews or four people. N1 and N2 can inform a tentative period-interpretation theme; N3 is a counterexample to a universal period problem and introduces a different denominator issue. N4 provides no first-hand file-reading observation. Keep those distinctions instead of deleting the inconvenient note or counting every code as another person.

A bounded synthesis is: “Two constructed notes raise different period-interpretation questions; one reports a clear range but another problem, and one does not describe opening the file.” A proposed next action is to clarify the range/timezone explanation and review it with appropriate future evidence. That is an interpretation and a plan, not an observed usability improvement, prevalence estimate or completed user test.

Retain a private note-to-theme-to-finding-to-action trace, competing explanations and the reason for any exclusion. Multiple notes or codes from one source are not independent respondents. Preserve the original quote/permission review before any public story, and do not turn these invented notes into the registered reader pilot, a testimonial or another research piece.

| Constructed note | Source wording | Bounded interpretation |
| --- | --- | --- |
| N1 | I saw October but no timezone. | Timezone question; outcome unknown |
| N2 | I found the timezone but not the exact date boundaries. | Different period-definition question |
| N3 | The date range was clear; I could not find the denominator. | Counterexample and separate issue |
| N4 | I did not open the file. | No direct reading observation |

## Work through three constructed source stories

Q1 — Constructed source excerpt: interviewer asks what changed; practitioner says, “I added a source column. We have not checked whether later handoffs improved.” Unsafe rewrite: “The new method made handoffs faster.” Bounded draft: “In this constructed interview, the practitioner added a source column; its later effect was not checked. The useful follow-up is to inspect whether the next handoff preserves the source.” Decision: keep the unmeasured effect and review the exact proposed excerpt before any real publication.

Q2 — Constructed incomplete record: notes contain “might have been the older file,” but the draft states the wrong file caused a lost client. Decision: remove the cause and client claim, retain the version uncertainty and ask a neutral clarification only within an agreed follow-up. Outcome: an unresolved source question, not a recovered story or a negative customer case.

Q3 — Constructed sharing limit: a participant agreed to private notes but not recording or named public quotation. Decision: keep the material within that agreement, hold the attributed story and use a separate original general example if useful. Outcome: no fabricated anonymous permission, testimonial, research participant count or claimed reader preference.

## Write one bounded story and account for missing evidence

A usable structure is the specific context, supported observation, interpretation labeled as yours, concrete change and an open next check. Do not make the source sound more certain to fit a dramatic opening. If the person reports a benefit, label it as their report unless the underlying measurement has actually been inspected and supports the claim.

One selected conversation does not establish what most practitioners do, how often a problem occurs or which guide readers prefer. Keep the sample, selection and consent scope explicit. This interview-writing workflow does not alter or substitute for the separate fixed comparative reader-pilot protocol.

## Keep the consent, source and publication decision together

The private worksheet separates learning question, participant basis, agreed recording/use/retention, neutral questions, source reference, exact excerpt, interpretation, proposed public wording, review decision and withdrawal handling. Minimize identifiers. If actual permission is withdrawn, use the agreed data process and record the outcome rather than promising that a published third-party copy can be recalled.

GOV.UK’s interview and informed-consent guidance and W3C’s writing guidance were read on 1 November 2026. The scripts, excerpts and cases are original constructed exercises. No recruitment, contact, interview, recording, public quotation or study outcome occurred. They do not count as another original research piece or real pilot participant.

## Primary sources

- [GOV.UK Service Manual: using in-depth interviews](https://www.gov.uk/service-manual/user-research/using-in-depth-interviews)
- [GOV.UK Service Manual: informed consent and data use](https://www.gov.uk/service-manual/user-research/getting-users-consent-for-research)
- [W3C WAI: clear writing and instructions](https://www.w3.org/WAI/tips/writing/)
- [GOV.UK: Analyse a research session](https://www.gov.uk/service-manual/user-research/analyse-a-research-session)