Best practices for designing a research survey

Learnings from Quantitative-Empirical Social Research, taught by Susanne Sackl-Sharif

A well-designed survey is one of the most effective ways to gather meaningful insights. However, even small mistakes in question design can lead to unreliable or misleading results. By planning carefully and following a few best practices, you can create surveys that are easier to complete and produce higher-quality data.

Before You Start

Define What You Want to Learn

Before writing a single question, identify exactly what you want to find out. List every insight your research needs, then turn each insight into one—and only one—question.

Keeping each question focused makes your survey easier to answer and your results much easier to analyze.

Write a Clear Introduction

Your introduction sets expectations and builds trust. Explain the following:

• Why are you collecting the data

• Who will have access to the responses

• How will participant privacy and confidentiality be protected

• Approximately how long will the survey take

A transparent introduction reassures respondents and can improve completion rates.

Writing Effective Questions

Keep It to One Idea Per Question

Every question should ask about a single topic. If you find yourself using the word and, consider splitting it into two separate questions.

For example, instead of asking:

"How satisfied are you with our website and customer support?"

Ask:

• How satisfied are you with our website?

• How satisfied are you with our customer support?

This avoids double-barrelled questions that produce unclear results.

Use Neutral, Positive Wording

Questions should uncover the truth—not confirm your assumptions.

Avoid:

• Leading questions

• Double negatives

• Emotionally loaded wording

Neutral phrasing encourages honest and unbiased responses.

Be Realistic About Recall

People struggle to remember events over long periods. Instead of asking respondents to think back over the past year, focus on shorter, more specific timeframes, such as the past month or past few weeks.

The easier something is to remember, the more reliable the answers will be.

Define Technical Terms Up Front

If your survey includes industry-specific language, explain those terms in the section introduction rather than inside individual questions.

This keeps questions concise while ensuring every respondent shares the same understanding.

Mix Different Question Types

A balanced survey combines several question types, each serving a different purpose.

Behavior Questions

These explore what people do and how often they do it.

Example: "How often do you contact customer support?"

Attitude Questions

These reveal opinions, preferences, and motivations behind user behavior.

Filter Questions

Use filter questions to ensure respondents answer only the questions relevant to them, creating a shorter, more personalized survey experience.

Control Questions

In longer surveys, control questions can help check for response consistency by asking about similar concepts using opposite wording.

Open-Ended Questions

Open-ended questions give respondents the opportunity to share unexpected insights that predefined answers might miss.

They should never be mandatory, and it's best not to begin your survey with one. Starting with simple questions helps respondents engage before asking for more detailed responses.

Designing Better Answer Options

Label Both Ends of Rating Scales

Always make rating scales explicit.

For example:

Strongly disagree — Strongly agree

Clearly labeled endpoints help respondents interpret the scale consistently.

Keep Lists Manageable

When multiple items use the same rating scale, try to limit each group to no more than eight items.

Longer lists often encourage respondents to rush through by selecting the same answer repeatedly. When you have many related items, consider using matrix or checkbox grids to keep the survey organized and easier to complete.

Conclusion

Great surveys begin long before the first response is collected. By clearly defining your research goals, writing focused and unbiased questions, mixing appropriate question types, and designing clear answer options, you can gather data that is both reliable and actionable. A thoughtful survey not only improves the quality of your insights but also creates a better experience for every respondent.