#questionnaire pretesting

Survey Cognitive Interview Pretest Protocol Writer: Think Aloud and Verbal Probes Mapped to Comprehension, Retrieval, Judgment, and Response, a Probe Matrix, Interviewer Script, Rounds, Problem Codes, and a Revision Log

Plan cognitive interviews to pretest a questionnaire before fielding: think aloud instructions, scripted probes for each item mapped to the four stage response model, room for spontaneous probes, a full interviewer script, sample size and iterative rounds, a problem coding scheme, and a findings to revision log.

Act as a survey methodologist who designs and conducts cognitive interviews to pretest questionnaires for public health, education, and government surveys, and who uses the four stage response model (comprehension, retrieval, judgment, response) to find out why an item fails, not just that it fails. Inputs: - Draft questionnaire items to test, numbered, with response options and skip logic: [DraftItems] - Target population and the subgroups most likely to struggle (language, literacy, age, region): [TargetPopulation] - Survey mode and length: web, phone, paper, or interviewer administered: [SurveyMode] - Known concerns about specific items from the team or past data: [ItemConcerns] - Resources: interviewers, session length, incentive, timeline, recording consent rules: [Resources] - Output format: [Format] Generate: 1. An approach note: concurrent think aloud, retrospective probing, or a mix, chosen for SurveyMode and TargetPopulation, with one sentence on why. 2. A probe matrix: for each item in DraftItems, the response stage most at risk, one or two scripted probes per stage worth testing (for example a paraphrase probe for comprehension, a recall strategy probe for retrieval, a confidence probe for judgment, a fit of options probe for response), and a cue for spontaneous probes when the participant hesitates or changes an answer. 3. The interviewer script: welcome, consent and recording, a think aloud practice exercise unrelated to the survey, item administration in the survey's own mode, the probes, and a closing debrief. 4. Interviewer rules: read items exactly as written, use neutral probes, do not teach the meaning of a term, note nonverbal hesitation, and how to probe without leading. 5. Sample and rounds: a recruitment plan by subgroup, participants per round within ranges commonly used in the cognitive interviewing literature, and at least two iterative rounds with revisions in between. Say that this is qualitative evidence and not a statistical sample. 6. A problem coding scheme: codes by stage (comprehension, retrieval, judgment, response, plus instructions, navigation, and sensitivity), with a definition and an example from these items. 7. A summary template per item: what participants said, codes applied, how many participants showed each problem, and whether it was seen across subgroups. 8. A findings to revision log: item, problem, evidence quote, proposed revision, round tested, outcome. Constraints: - Do not invent findings; use placeholders until real interviews happen. Probes must never suggest a correct answer. - Respect consent and privacy rules in Resources. No em dashes.