What Are the Types of Survey Questions?
Survey questions fall into two groups. Closed-ended questions give respondents a set of answers to choose from, and open-ended questions let them answer in their own words. Within those groups there are about a dozen formats, each suited to a different kind of information.
The question type decides what kind of answer you get back. Choose the wrong one and no amount of analysis will recover what you needed to know.
This guide starts with a table for choosing, then covers each type with an example, when to use it, and the mistake people make with it.
Choose the Type From What You Want to Learn
You want to know | Use |
|---|---|
Which one option people prefer | Single choice |
All the options that apply to them | Multiple choice |
A plain fact with two outcomes | Yes or no |
How strongly they feel | Rating scale or Likert scale |
How several items compare on the same scale | Matrix |
What matters most among several things | Ranking |
How likely they are to recommend you | Recommendation scale, 0 to 10 |
A quantity, a price or an age | Number |
Their reasons, in their own words | Open text |
A reaction to something visual | Image choice |
Open-Ended vs. Closed-Ended Questions
Every question is one or the other.
Closed-ended questions offer fixed answers. They are quick to answer, easy to count, and comparable between people and over time. Use them for anything you want to measure. Their answers can also be split by group in a cross tabulation.
Open-ended questions let people write freely. They reveal reasons and ideas you did not think to offer as options. They take longer to answer and longer to analyze.
A good survey is mostly closed-ended, with one or two open questions placed after the ratings they explain. For example: a satisfaction rating, then "What is the main reason for your score?"
Key point: use closed-ended questions to measure and open-ended questions to understand why.
The 12 Types of Survey Questions
1. Single Choice
Respondents pick one answer from a list.
Example: "How often do you buy coffee outside your home?" Daily / A few times a week / A few times a month / Rarely / Never.
Use it when the answers exclude each other. Common mistake: options that overlap, such as "1 to 5" and "5 to 10".
2. Multiple Choice
Respondents pick every answer that applies.
Example: "Which of these brands have you heard of? Select all that apply."
Use it when more than one answer can be true. Common mistake: leaving out "None of these", which forces people to tick something false.
3. Dropdown
A single choice question shown as a menu.
Example: "Which country do you live in?"
Use it when the list is long and familiar. For short lists, show the options: people choose more carefully when they can see them all.
4. Yes or No
Two possible answers.
Example: "Have you bought from us in the last 12 months?"
Use it when the question is a fact. Common mistake: using it for opinions. "Do you like our new design?" hides everything between loving it and tolerating it.
5. Rating Scale
Respondents choose a number, usually from 1 to 5, with a label at each end.
Example: "How would you rate the quality of your last order?" 1 (very poor) to 5 (excellent).
Use it when you want a quick measure of strength. Keep the same scale across the whole survey.
6. Likert Scale
A rating where every point has a word label, most often for agreement.
Example: "The instructions were easy to follow." Strongly disagree / Disagree / Neither agree nor disagree / Agree / Strongly agree.
Use it when you want answers that mean the same thing to every respondent. See Likert scale examples for scales you can copy.
7. Matrix
Several items rated on the same scale, laid out as a grid.
Example: "How satisfied are you with each of the following?" with rows for delivery, price and support.
Use it when you rate related items. Common mistake: too many rows. Past five or six, people click the same column all the way down.
8. Ranking
Respondents put items in order.
Example: "Rank these factors from most to least important when choosing a bank: fees, branch access, app quality, customer service, interest rates."
Use it when you need to know priorities, because ratings let people call everything important. Keep the list to seven items or fewer.
9. Slider
Respondents drag a handle along a line to choose a value.
Example: "What share of your grocery shopping do you do online?" 0% to 100%.
Use it when the answer is a quantity on a continuous range. On a phone a slider is harder to set precisely than tapping an option.
10. Recommendation Scale (0 to 10)
One question on an eleven-point scale: "How likely are you to recommend us to a friend or colleague?"
Use it when you track loyalty over time. The answers are grouped into promoters (9 and 10), passives (7 and 8) and detractors (0 to 6). The NPS calculator turns the counts into a score.
11. Number
Respondents type a number.
Example: "At what price would this product start to feel expensive?"
Use it when you need a real quantity and do not want to steer the answer with ranges. State the unit or currency in the question.
12. Open Text
Respondents write an answer, short or long.
Example: "What is the one thing we could do to improve your experience?"
Use it when you want reasons, language and ideas. Make it optional, and ask no more than two or three in one survey.
Other Question Types Worth Knowing
Image choice: respondents pick from pictures. Good for packaging, logos and designs.
Screening questions: asked at the start to decide who continues. They are how you control who is in your sample.
Demographic questions: age, gender, income and similar. Usually single choice, placed at the end.
Attention checks: a question with one obviously correct answer that catches people who are not reading. See these attention check questions.
How Many of Each Should a Survey Have?
A survey that takes under five minutes gets better answers than one that takes fifteen. As a working guide for about 12 questions:
One or two screening questions at the start.
Six to eight closed-ended questions that carry the measurement.
One attention check in the middle.
One or two open questions near the end.
Put the questions that matter most early, while attention is highest. For wording you can copy, see 50 survey question examples.
Question Types in SegmentOS
SegmentOS has each of the 12 types above, plus image, video and text blocks for showing a concept or an ad before you ask about it. The rating scale uses numbers with a label at each end, and the Likert scale takes your own labels. You can describe what you want to learn in a sentence and get a first draft of the survey, or start from a template.
Key Takeaways
Two groups: closed-ended questions give fixed answers, open-ended questions let people write freely.
Measure with closed questions, then ask one open question to learn why.
Use ranking for priorities, because rating scales let people call everything important.
Keep grids short: five or six rows at most.
About 12 questions is plenty for most studies, including one attention check.



