Join for free

Design a training-needs interview guide

Author: AILesson8 min setupTested with:ChatGPTReviewed: 2026-08-28

Quick answer

Investigate real work gaps, causes, conditions, and transfer needs before prescribing training. Provide: Decision and performance need, Participants and research context, Training and analysis constraints. Expected result: A non-leading interview guide with evidence probes, scenarios, privacy boundaries, coding fields, and synthesis plan.

1

Add your context

Your text stays in this browser. AILesson Prompts does not send it to a model or server.

2

Your prompt

Unfilled fields remain visible as placeholders, so you can still copy and edit the prompt

Design a non-leading training-needs interview guide for the supplied decision.

Decision and evidence:
[decision]

Participants and research context:
[participants]

Constraints and alternatives:
[constraints]

Do not assume the observed gap is caused by missing knowledge or that training is the solution. Separate required performance, current behavior, knowledge or skill, tools and information, incentives, workload, process, permissions, environment, and manager support. Start with consent and confidentiality limits, then ask for recent concrete work examples, decision points, errors and successful cases, evidence sources, variation, workarounds, consequences, existing learning, desired practice, transfer conditions, and non-training fixes. Use open questions before probes; avoid asking participants to diagnose themselves or disclose health and protected information. Tailor safe probes by participant group without creating inconsistent core questions. Include a short realistic scenario, closing prioritization, note-taking and coding fields, interviewer cautions, and a synthesis rule that preserves minority and contradictory evidence.
Private by defaultPrompt assembly happens locally in your browser. Avoid placing confidential information into any AI service unless your organization allows it.

From input to outcome

A worked example

See how concrete context turns this recipe into a usable result

Actual input

Decision and performance need
A regional retailer must decide whether to build training for store supervisors who approve online-return exceptions. In 48 audited exceptions over six weeks, 13 lacked the required reason code and 7 exceeded the supervisor limit; three had both issues. Audit logs show which record failed but not why. Correct behavior: verify receipt and item state, select an allowed reason, stay within CNY 500 authority, and escalate higher values. Errors can delay refunds and weaken controls. Some stores have no failures. The decision is whether training, a system or process change, or a combination is justified.
Participants and research context
Interview 6 supervisors from stores with errors, 4 from stores without errors, 3 service associates, 2 regional managers, and the system owner. Purposeful sample across tenure and store volume; not statistically representative. Thirty minutes by video in Chinese. Participation is voluntary; managers will receive aggregated themes, not names or identifiable quotes. Interviewer is from Learning and has no rating authority. Do not ask age, health, family, or disciplinary history. Participants may skip questions. No incentive.
Training and analysis constraints
Possible causes include unclear policy, confusing system labels, missing receipt data, customer pressure, queue volume, permissions, manager messages, or knowledge. Existing assets: four-page policy and a five-minute system demo. Possible learning formats: 20-minute live practice or short job aid; maximum one paid hour per supervisor. System changes require a quarterly release. Managers can run weekly case reviews. Need evidence table by cause category, critical scenarios, learning objectives only if supported, non-training fixes, confidence, and next validation.

Example output

Training-needs interview guide — return exceptions

Opening: “We are studying how exception approval works in practice and whether learning, tools, process, or another change would help. This is not an evaluation. Participation is voluntary; you may skip questions. We will report aggregated themes, but a serious control or safety concern may need confidential escalation.” Ask permission to take notes.

Core questions: Walk me through the most recent exception you handled, from receipt check to final record. What information was available at each decision? Where did you pause or seek help? What made the case easy or difficult? Describe a case that went well and one that needed correction. How did you choose the reason code? How did the CNY 500 limit appear in your workflow? What happened when value exceeded it? Which policy, demo, colleague, or workaround did you use? What differed during a busy queue? If you could change one thing in the system, process, manager guidance, or practice, what would it be?

Evidence probes: What did you see on screen or in the receipt? Can you describe the label without sharing customer data? Is that typical or one case? What record could confirm it? What is a counterexample? For associates ask how handoff information reaches supervisors; for managers ask what guidance and review they provide; for the system owner ask how labels, limits, missing receipt data, and permissions behave. Keep the core incident sequence the same.

Scenario: “Receipt verified; damaged item; requested refund CNY 680; queue is long. Show what you would check, record, and escalate.” Probe decision and information, not test performance.

Close: Rank the top two barriers and the most useful safe change. Ask what a pilot should measure. Coding fields: participant group, incident stage, evidence source, knowledge/skill, information, interface, permission, workload, incentive, process, manager support, successful practice, workaround risk, proposed fix, counterevidence, confidence.

Synthesis: require at least two independent examples for a recurring theme, while retaining every serious control signal separately. Compare error and no-error stores without identifying them. Do not convert frequency in this purposeful sample into prevalence. Recommend learning objectives only where evidence shows a practice gap; route label or permission problems to system work and inconsistent guidance to process ownership.

Why this works

  1. 1

    Cause categories keep training from being used to compensate for broken tools, priorities, or permissions

  2. 2

    Concrete incident probes produce evidence that can define practice and transfer conditions

Check the result

  • Can the guide distinguish a knowledge or skill gap from process, tool, workload, and management causes?

  • Are questions open, tied to recent work, privacy-safe, and comparable across participants?

  • Will synthesis retain successful cases, minority evidence, contradictions, and non-training options?

Use it with confidence

Frequently asked questions

Practical answers about when to use this recipe, what to provide, and where human review still matters

What should I prepare before using “Design a training-needs interview guide”?

For “Design a training-needs interview guide,” prepare Decision and performance need, Participants and research context, and Training and analysis constraints. Replace placeholders only with information you can verify. If a detail is unknown, preserve that uncertainty explicitly instead of asking the model to infer it.

When is the “Design a training-needs interview guide” result not ready to use?

The result is not ready if it does not yet deliver the stated outcome—A non-leading interview guide with evidence probes, scenarios, privacy boundaries, coding fields, and synthesis plan—from the supplied evidence, or if it relies on unresolved assumptions, missing approvals, or invented details. Use the checks as release gates: revise the source inputs or assign a named, authorized reviewer instead of polishing an unsupported output.

Which AI tools have recorded tests for “Design a training-needs interview guide”?

The published test record for “Design a training-needs interview guide” lists ChatGPT as of 2026-08-28. This confirms recorded runs, not guaranteed compatibility or identical results in later product versions. For another tool or version, keep every constraint visible and repeat the result checks before use.

More ways to explore

Where this recipe fits

Text promptsFeedback & reviewPeople, hiring & developmentPerformance reviewsTeaching & educationFor managersFor learners & educators

Keep the work moving