HomeCasa Prompts

Design a survey questionnaire without biased questions

Build a survey ready for piloting: questions mapped to your goals, balanced scales, an order that avoids priming and a question-by-question bias review with fixes. Designed for ChatGPT, Claude, Gemini and Microsoft Copilot. It includes 5 customizable variables and you can copy it for free in one click.

Intermediate · Text · Pending editorial review

Designed for:

Prompt

Act as a survey methodologist with experience in social and market research.

Context:
- Survey goal (which decision the results will inform): {{goal}}
- Target population: {{population}}
- Mode: {{mode}}
- Maximum length: {{length}} minutes
- Topics to cover: {{topics}}

Deliverables:
1. Information objectives: turn the goal into 3-6 research questions and map every questionnaire item to one of them (if an item serves no objective, drop it).
2. Full questionnaire in blocks: introduction and consent text, screening questions, core questions and demographics at the end. For each item state the type (closed, rating scale, open, ranking) and the response options.
3. Balanced scales: the same number of positive and negative options, a midpoint where it makes sense and "Don't know" or "Prefer not to say" where appropriate. Options must be mutually exclusive and exhaustive.
4. Order and flow: justify the order to avoid priming and anchoring effects and note any skip logic.
5. Bias review table: original item, problem found (leading question, double-barreled, loaded terms, absolutes like "always", social desirability, acquiescence, jargon) and corrected version.
6. Estimated completion time based on the item count, and what to cut if it runs over.
7. Pilot plan: 5-8 cognitive interviews, what to ask participants and which signals to check (drop-off, blank answers, time taken).

Constraints:
- Do not invent benchmarks or expected response rates. If you suggest a validated scale, name it only if it exists and remind me to check its license and validated version; otherwise use [FILL IN].
- Clear, neutral, inclusive language suited to the population.
- Collect only the personal data that is needed and include a privacy notice placeholder marked [FILL IN: data controller and privacy policy link].
- If key information is missing, ask me before drafting.

Format: the numbered questionnaire ready to paste into a form tool, followed by the bias table and the pilot plan.

How to use it

Start by naming the decision the results will drive; that is what lets you cut unnecessary questions. Paste the questionnaire into your form tool, pilot it with 5-8 people from the real population and come back with their feedback for a second draft.

Usage example

Goal: decide whether to extend public library weekend hours · Population: residents 16+ · Mode: online · Length: 7 minutes · Topics: current use, preferred hours, services.

Sample output

Information objectives

  1. Who uses the library today, and when?
  2. Which weekend hours would residents use?
  3. Which services are they missing?

Questionnaire (excerpt)

Block 0. Introduction This survey takes about 7 minutes and will help decide the library's weekend hours. Taking part is voluntary. [FILL IN: data controller and privacy policy link]

Block 1. Screening

  1. Do you live in [town]? (Yes / No → end)
  2. How old are you? (16-24 / 25-44 / 45-64 / 65+ / Prefer not to say)

Block 2. Current use (objective 1)

  1. In the past 12 months, how often have you visited the library?

(Never / Less than once a month / 1-3 times a month / Once a week or more)

Bias review

OriginalProblemCorrected version
"Don't you think the library should open on weekends?"Leading question"If the library opened on weekends, how likely would you be to visit?"
"Are you happy with the hours and services?"Double-barreledSplit into two questions
"Do you always use the study room?"Absolute "always"Frequency scale

Estimated time

14 items ≈ 6-7 minutes. If it runs over, cut the service ranking question first…

Key facts

Category
Research · Surveys
Output type
Text
Level
Intermediate
Designed for
ChatGPT, Claude, Gemini and Microsoft Copilot
Verified in
Pending editorial verification
Variables
Goal and decision, Target population, Mode, Maximum length (minutes) and Topics to cover
Language
English
Price
Free, no sign-up
Updated
Nov 7, 2026

Frequently asked questions

What is a double-barreled question?

A question that asks two things at once, such as "Are the staff friendly and quick?". Respondents cannot tell which part to answer, so split it into two items.

Should every scale have a midpoint?

Not always. It helps when neutrality is a real position. If you remove it to force a lean, offer "Don't know" so people without an opinion are not pushed to invent one.

What does a cognitive interview add?

You watch real people interpret each question while thinking aloud. It catches ambiguities that no desk review will find.

What is this prompt for?

Build a survey ready for piloting: questions mapped to your goals, balanced scales, an order that avoids priming and a question-by-question bias review with fixes.

Which AI tools does it work with?

It is designed for ChatGPT, Claude, Gemini and Microsoft Copilot. We have not verified it in these tools yet, so results may vary between them.

How do I customize the prompt?

Click “Customize” and fill in the variables (Goal and decision, Target population, Mode, Maximum length (minutes) and Topics to cover). The text updates instantly and you can then copy it in one click. Values are stored only in your browser.

Do I need experience to use it?

It is intermediate: it helps to know the topic so you can provide good context and review the result critically.

Is this prompt free?

Yes. Every HomeCasa Prompts prompt is free, with no sign-up or subscription. You can also use them in professional projects.

Can I trust the AI output?

Treat it as a high-quality draft: always check data, figures and important claims before publishing or making decisions.

Related prompts