Research questions ≠ interview questions
- A research question is what the team needs to learn; an interview question is what you actually ask a participant. Confusing them produces useless data.
- Good research questions are specific, answerable and decision-linked.
- You never ask a research question out loud — you triangulate toward it.
The single most common beginner mistake in research is reading the research question to the participant.
A research question is written for the team: “What barriers prevent existing customers from using autopay?” It is abstract, analytical, and linked to a decision (should we invest in autopay UX?).
An interview question is written for a human being in a conversation: “Walk me through the last time you paid this bill — what did you do first?” It is concrete, behavioral, and answerable by memory rather than by speculation.
The transformation always runs the same way: research question → what observable behavior or story would answer it → concrete, past-tense prompts that surface those stories. If you ask a participant “what barriers prevent you from using autopay?”, you get a rationalization produced on the spot. If you ask them to reconstruct the last three times they paid, the barriers appear on their own — in the pauses, workarounds and apologies.
Rules of thumb for research questions: one decision per question (“what should we learn to make this call?”), specific enough to falsify, and few — a study that tries to answer ten questions answers none.
Try it: take the research question “Why don’t users trust our checkout?” and derive five interview prompts that never mention trust, checkout, or the product.
- Portigal — Interviewing Users, ch. 3
- NN/g — Formulating Research Questions