Research & analysisMeasurement comparabilitytext
Trace how a survey question changed across waves
Tests whether wording and response-option changes make survey results incomparable over time.

Ready to use
Prompt
Compare the supplied [EARLIER QUESTIONNAIRE] and [LATER QUESTIONNAIRE] for [MEASURE OF INTEREST], using [CODEBOOK NOTES] where available. Build a wave-by-wave table of question wording, response options, ordering, skip logic, mode and eligible population. Classify each difference as likely comparable, comparable only with a stated bridge assumption, or not safely comparable; cite the exact question or codebook location. Propose a sensitivity analysis or bridge study where feasible, but do not calculate a trend without real data. If a questionnaire is missing, give an evidence request rather than guessing continuity. Check whether a change in who was asked, not just how the words changed, could explain an apparent trend.