Review My Data Collection Tool
Improve an existing instrument with item-level fixes, marked revisions and a focused pretest plan.
- What You Need
- Paste or attach the instrument, plus its purpose, respondents, mode, and any earlier review notes. The tool itself is required.
- Choose This Prompt When
- Choose this to critique an existing instrument; choose an interview or focus group design prompt to create a new guide.
Review the supplied data collection instrument using its research questions, intended respondents, administration mode and prior feedback. Ask for the instrument if absent; a title or filename is not enough to review.
Identify its purpose, target respondents, mode and missing context. Review only visible sections: construct coverage, neutral wording, recall periods, answer options, flow, skips, accessibility, translation needs, respondent burden, consent and privacy. Include disaggregation only where analytically useful and safe, considering identification risk; do not require every demographic variable.
Give a short section-level assessment using a defined 1 to 5 scale only for sections supplied. Then provide a prioritized issue table with item identifier, exact problem, consequence and proposed wording or routing. Show supplied items with edits marked and list proposed new questions separately. Explain material edits and suggest a small cognitive or field pilot with what to observe. Treat the pilot as proposed, not completed validation. Mark missing questions, response options or skip paths unreviewed and request the parts needed. Preserve source uncertainty and do not invent questions as if they were already in the tool.
Worked Example
Illustration only. The setting and figures are invented to show the pattern.
Scenario: A 40-question household questionnaire on child health goes to pretest in two weeks and needs a check first.
What You Give the AI: The full questionnaire as pasted text, its purpose, respondents who are mothers and caregivers, face-to-face mode with tablets, and notes from the program team.
What the AI Might Return: A section-level assessment on a 1 to 5 scale for each section supplied. A prioritized table follows. One item asks "How often does your child eat healthy food?", problem vague term and no recall period, consequence answers cannot be compared, proposed wording "Yesterday, which of these foods did your child eat?" with a routing note. Supplied items are shown with edits marked, and new questions sit in a separate list. A short pretest plan names what to watch for, such as items interviewers reword.
When to Choose This Prompt
- You have a draft or existing questionnaire and want it fixed item by item.
- You want edits visible in the text, so you can accept or reject each one.
- You want a short pretest plan that tells interviewers what to watch.
When Not to Choose It
- You need to build a new survey from the start. Design a Household Survey
- You need a new discussion guide for focus groups. Design a Focus Group Guide
- You need a new interview guide. Design an Interview Guide
- You want a review aimed at wording alone. Review Survey Question Wording
Checking the Output
- Quoted items must match your tool word for word. Open the instrument next to the table and spot-check several.
- Section scores should appear only for sections you pasted. Missing parts should read as unreviewed.
- Edits to your items should be visibly marked, and the AI's new questions should sit in their own list.
- Proposed routing has to be traced by hand. Pick one respondent type and follow the skips from start to finish.
- The pretest plan should say who is tested, what is observed and what would change the tool.
- Disaggregation questions should have a stated use. Any that could identify a small group deserve a second look.
Common Problems and Fixes
- The AI rewrites items that were fine. Fix: Limit edits to items where the table names a problem, and leave the rest alone.
- New questions are mixed in with yours. Fix: Keep them in a separate list, each with the reason it is needed.
- The review treats the tool as validated after the AI's edits. Fix: Remind it that the pretest is proposed, and have it phrase the result as untested.
- Translation, accessibility and consent get one general comment. Fix: Request item-level notes where wording, response options or reading level could fail in your languages.
Adapting It to Your Context
- Phone or remote mode: State the mode and ask for burden, question length and consent wording to be reviewed for that setting.
- Sensitive topics: Name the topics and ask for privacy risks at item level, with proposed skip or opt-out wording.
- Previous pilot notes available: Paste them and ask the AI to say which problems are already confirmed in the field.