Customer research
Separate observations from interpretations in research notes
Write each note in two layers. The first records what the participant did and the words they used. The second holds your interpretation, labeled as yours, plus a question that could confirm or contradict it. A guess then stays visibly a guess when someone summarizes the notes.
A text to start with
Here are my raw notes from [number] sessions with [type of participant] about [topic]: [pasted notes]. Rewrite each note as four fields: what the participant did, their exact words if I recorded them, my interpretation, and a follow-up question that would test that interpretation. Mark any note where I did not record the words. Do not add explanations the notes do not support.
Example request. Change the details to fit your situation.Observation and interpretation note sheet
Session reference and date: [ ] Participant description: [ ] Task or topic being discussed: [ ] Observed action: [ ] Quotation captured verbatim: [yes / no] Exact quotation, if captured: [ ] Paraphrase, if no quotation exists: [ ] My interpretation: [ ] Why I think so: [ ] Other explanations that fit: [ ] Follow-up question to ask: [ ] Note status: [observed only / interpretation to test / interpretation supported]
Where interpretation hides in notes
Check verbs and adjectives first. Words like confused, frustrated, prefers or ignored usually describe the researcher's reading, not the participant's behavior. A paraphrase can carry an inference when the exact words were not captured, so see which notes have no quote. Be careful with motives: someone who skipped a step may have missed it, found it unnecessary or been interrupted. Watch for one participant's comment rewritten as what users want. GOV.UK recommends open, neutral questions grounded in real examples, so every note should point back to something that happened.
Ask another agent to check the result
Read my notes and list every interpretation that goes beyond what the participant said or did. For each, quote the note, name the observation it rests on, and give another explanation that fits the same evidence. Flag any statement presented as fact that has no recorded observation behind it.
Turn interpretations into questions
Take the interpretations you rely on most and read each beside its observation. Keep the ones the evidence supports and rewrite the rest as open questions for the next session. Where no words were recorded, say so in the note instead of reconstructing them from memory. Share the two-layer version with teammates so they can tell evidence from reading.