A leading interview question contains the researcher's preferred answer or proposed solution, so the repair removes that premise and asks for a specific past event the participant can describe from memory.
Find the answer embedded inside the question
Questions such as would this save you time, how useful would this feature be, or would you pay for this ask the participant to react to the researcher's frame. A friendly participant can supply a plausible answer without recalling any real behavior. The transcript may sound positive while containing no evidence about what happened before the call.
The review marks the premise, the type of answer it invited, and the missing fact. If the decision concerns a recurring workflow problem, the missing evidence may be the last occurrence, who was involved, what the participant tried, how long it took, what it cost, and what happened next.
Replace prediction with a recent event
A neutral replacement begins where memory is strongest: tell me about the last time this happened. Follow-up probes recover the sequence, tools, people, alternatives, constraints, and consequence without naming the proposed product. The GOV.UK Service Manual recommends open, neutral questions and real examples rather than general statements about how work should happen.
The interviewer should also remove introductory product detail that teaches the participant what matters. A short study explanation and consent process are necessary, while a feature tour before the behavioral section changes the evidence environment. The proposed solution can appear later only when evaluation of that solution is an explicit research objective.
Check the repair against the decision
Each revised section states which research question it serves and what evidence would change the team's view. The post-call review then separates observed behavior, participant interpretation, interviewer inference, and open question. If the repaired guide still produces only opinions, the next revision needs a narrower event or better-qualified participant rather than more persuasive wording.
Where the service stops
Reality Contact, LLC repairs research materials and interviewer practice, but does not recruit or contact participants, obtain consent, moderate the study, make product decisions, provide legal or ethics approval, or claim that qualitative interviews represent a population. The buyer approves the target segment, consent and data-handling process, recruits the next qualified wave, runs the repaired guide without presenting the proposed solution, and makes decisions from the resulting evidence. The work is research document preparation and interviewer training, and it does not replace legal, privacy, ethics, employment, or professional research advice. The buyer owns participant consent, recruiting, incentives, data handling, moderation, interpretation, and every product or segment decision.
Sources: GOV.UK in-depth interview guidance; Defra research methods and templates.