Questionnaire Design: How to Write Good Survey Questions

Reviewed by Prof. Dr. Alexander Holtermann · Updated 16 October 2026 · 5 min read

A good questionnaire starts from your research question, uses validated scales, neutral wording and a logical order, and is pilot tested before launch.

Key takeaways

Every question must serve your research question or hypotheses. If you cannot say which analysis a question feeds, delete it.
Use validated scales for your key concepts wherever possible and write your own items only where none exist.
Avoid double-barreled, leading, loaded and double-negative questions; keep wording short, concrete and neutral.
Order matters: easy and engaging questions first, sensitive questions and demographics toward the end. Aim for 10–15 minutes.
Always pilot test, and include an information sheet and consent at the start, as your ethics committee will expect.

Start with your research question, not with questions

The most common mistake is to start writing questions straight away. Begin instead with a question map: list your research question, your hypotheses or sub-questions, the concepts in them, and how each concept will be measured (operationalization). Only then write or select items.

This map keeps your questionnaire short and makes the analysis straightforward, because you already know which variables go into which test. It is also exactly what your ethics committee and supervisor want to see.

Research question / hypothesis	Concept	Measure	Analysis
H1: Part-time work is related to higher stress	Part-time work	Hours per week (open number)	Correlation
	Stress	Validated 10-item stress scale (1–5)
H2: Stress is higher in final-year students	Year of study	Single choice: 1st, 2nd, 3rd, 4th+	t-test / ANOVA

Question types

Choose the type that matches the information you need and the analysis you plan.

Type	Example	Use for
Single choice	What is your year of study?	Categories, group comparisons
Multiple choice	Which of these resources do you use? (select all)	Behaviors, preferences
Rating scale (Likert-type)	“I feel overwhelmed by my coursework.” 1 = strongly disagree … 5 = strongly agree	Attitudes, perceptions, scales
Numeric open	How many hours per week do you work?	Ratio-level data
Ranking	Rank these five factors by importance.	Priorities (use sparingly)
Open-ended	What would help you most to reduce stress?	Exploring, explaining, quotes

Rules for wording

Small wording choices change answers. Follow these rules and check each item against them.

One idea per question. “How satisfied are you with the teaching and the facilities?” is double-barreled; split it.
Neutral, not leading. “Don’t you agree that tuition fees are too high?” pushes respondents. Better: “How do you rate the level of tuition fees?”
No loaded words. Terms like “bureaucratic” or “unfair” carry judgments.
No double negatives. “I do not disagree that …” confuses everyone.
Concrete time frames. “In the last four weeks” instead of “recently”.
Simple language. Avoid jargon and abbreviations; write for your least expert respondent.
Exhaustive and exclusive options. Categories must not overlap (e.g. 0–5, 6–10, 11–15 hours) and should include “other” or “prefer not to say” where appropriate.

Structure and order

A well-structured questionnaire feels like a conversation. A typical order:

Information sheet and consent: purpose, duration, voluntariness, anonymity, data storage, contact details, right to withdraw.
Screening questions (if needed): does the respondent belong to the target group?
Easy, engaging opening questions related to the topic.
Core sections grouped by topic, from general to specific.
Sensitive questions later, once trust is established.
Demographics at the end (age, gender, program), only what you will analyze.
Thank you page with debrief and support contacts if the topic is sensitive.

Length, layout and data quality

Long surveys lose respondents. For student projects, 10–15 minutes is a good upper limit; many respondents will complete it on a phone, so test the mobile layout. Show a progress indicator and avoid large grids of items, which encourage straight-lining (ticking the same answer down a column).

Plan data quality checks in advance: one or two attention checks (“Please select ‘agree’ for this item”), plausibility checks for numbers, and a minimum completion time to flag speeders. Decide the exclusion rules before you look at results and report them.

Ethics and data protection

Almost every university requires ethics approval or at least a self-assessment for surveys with human participants. Expect to submit the questionnaire, information sheet and consent wording. Under the UK GDPR and the EU GDPR you must tell respondents what data you collect, why, on which lawful basis, how long you keep it and who has access. In the US, your Institutional Review Board (IRB) decides whether the survey is exempt or needs review.

Avoid collecting identifying data you do not need. Even without names, combinations such as program, year, age and gender can identify people in small groups.

Pilot test before you launch

Test your questionnaire with 5–10 people from your target group, ideally while they think aloud, and then with a slightly larger group to check timing and data export. Fix unclear items, missing options and technical issues before the real launch. See our article on pilot studies for a full checklist.

Build it with Professor Digital

At /umfragen you can build your questionnaire for free, with single and multiple choice, rating scales, open questions, consent page and attention checks, and export the data for your statistics software. The sample size calculator tells you how many responses you need. Requirements for ethics review and questionnaire documentation differ by institution; your university’s own guidelines take precedence. Upload them to Professor Digital and every suggestion is tailored to them.

Frequently asked questions

How do you design a good questionnaire?

Start from your research question, map each concept to a measure, use validated scales where possible, write short neutral questions, order them logically, keep it to 10–15 minutes and pilot test it.

How many questions should a survey have?

As many as your research question needs and no more. For student projects, a completion time of 10–15 minutes, often 20–40 items, is a sensible upper limit.

What is a double-barreled question?

A question that asks about two things at once, e.g. “How satisfied are you with teaching and facilities?” Respondents cannot answer both with one rating, so split it into two questions.

Should demographic questions go first or last?

Usually last, because they are less engaging and can feel intrusive. Put them first only if you need them for screening or routing.

Do I need ethics approval for a student survey?

Most universities require at least an ethics self-assessment for any research with human participants. Check your department’s procedure early, as approval can take several weeks.

Sources and further reading

Clark, T., Foster, L., Sloan, L., & Bryman, A. (2021). Bryman’s Social Research Methods (6th ed.). Oxford University Press.
Creswell, J. W., & Creswell, J. D. (2023). Research Design: Qualitative, Quantitative, and Mixed Methods Approaches (6th ed.). Sage.

Your university’s examination regulations and style guide always take precedence.