How to write survey questions people answer honestly
Learn how to write survey questions that get honest answers: neutral wording, clean options, good order, and a checklist to edit with.
By Bhumi Avaiya. 8 min read. .
To write survey questions people answer honestly, ask about one thing at a time, use neutral wording, give answer options that cover every real answer, and keep the whole survey short. Most bad data comes from badly written questions, not from bad respondents.
This guide goes through each of those rules with weak and strong examples, then gives you a checklist to edit with. All examples are hypothetical. If you are ready to build, how to create a survey in WordPress covers the plugin side.
How to write survey questions: start with what you will do with the answers
Before you write a question, decide what it is for. Ask “what will I change based on this answer?” If you have no answer, delete the question.
This one habit cuts surveys in half. Most of the questions people add “just in case” are the ones that get skipped, and the ones that cause drop-off. Write your goal at the top of the page, and check each question against it.
One idea per question
A double-barreled question asks two things but allows one answer.
- Weak: “How satisfied are you with our prices and customer service?”
- Strong: Two questions, one about prices and one about service.
If someone loves your service but finds the prices high, they cannot answer the weak version honestly. They pick something in the middle, and your data says “fine” when the truth is “great and expensive.”
A quick test: if the question contains “and” or “or” joining two topics, split it.
Keep the wording neutral
Leading questions nudge people toward one answer. Respondents often go along with the nudge, especially when they want to be polite.
-
Weak: “How much did you enjoy our excellent new course?”
-
Strong: “How would you rate the new course?”
-
Weak: “Don’t you think the checkout is too slow?”
-
Strong: “How would you describe the speed of the checkout?”
Watch for loaded words like “excellent”, “amazing”, “unfortunately” and “simply”. Watch for questions that assume a fact, like “Why do you love our newsletter?” Not everyone does.
Another trap is asking for what someone would do in the future. “Would you buy a premium version?” gets optimistic answers, since saying yes costs nothing. Ask about what they do now: “What do you currently use for this?”
Use plain, specific language
- Avoid jargon. If your audience would not say the word, do not use it.
- Avoid vague words. “Often”, “regularly” and “recently” mean different things to different people.
- Name a time frame. “In the last month” is easier to answer honestly than “recently”.
Weak: “Do you use our blog regularly?” Strong: “In the last month, how many of our articles did you read?” with options like none, one or two, three to five, more than five.
The specific version is also easier to analyze, because people in the same option did roughly the same thing.
Design the answer options
Good options matter as much as the question. This is where many surveys quietly go wrong.
Cover every real answer
If someone cannot find their answer, they leave or pick something random. Add “Other” with a text field, “None of these” or “Not sure” where they fit.
Make options distinct and complete
Options should not overlap, and together they should cover the range.
- Weak: “0 to 5”, “5 to 10”, “10 or more”. Where does someone with 5 go?
- Strong: “0 to 4”, “5 to 9”, “10 or more”.
Keep them parallel and short
Write options in the same form and roughly the same length. One long option stands out and gets picked more.
Mind the order
People tend to pick options near the top. When the order carries no meaning, shuffle it if your tool supports it. Quizably can randomize answers, shuffled per visitor and stable within their session, so check the setting in your builder. Do not shuffle anything that has a natural order, like a scale from low to high.
Balance your scales
If you use a rating or agreement scale, make it even on both sides.
- Weak: “Excellent, very good, good, fair.” Three positive and one neutral, so there is no room for a negative.
- Strong: “Very poor, poor, fair, good, very good.” Two negative, a middle, two positive.
Keep the direction the same throughout the survey. If 5 means “very satisfied” in one question, it should not mean “very dissatisfied” in another. People do not always read the labels, so flipping the direction creates answers that are wrong and look right.
Label the ends at least, and ideally every point. A bare 1 to 5 leaves people guessing what a 3 means.
Put the questions in a sensible order
Order shapes how people answer.
- Start easy. A simple, interesting question gets people moving.
- Group related questions. Jumping between topics makes people work harder.
- Go general to specific. Ask about the overall experience before the details. Otherwise the details color the overall answer.
- Put sensitive questions later. People are more willing to answer once they are engaged.
- End with an open question. “Anything else?” catches what you missed.
Also remember that earlier questions can influence later ones. If you ask “How bad was the delay?” first, the next question about satisfaction will already be colored.
Keep it short
Every question costs you some respondents. A few guidelines:
- Five questions or fewer for people who do not know you.
- Six to ten for people who chose to help, like customers.
- More than ten usually means two surveys.
Tell people how long it will take, and keep that promise. A survey that says “a minute” and runs for ten loses trust fast.
Open or closed questions
Closed questions offer options. They are quick, easy to compare and simple to chart. Most of your survey should be closed.
Open questions let people write. They show what you did not think to ask, and the wording people use is worth reading. But fewer people answer them, and they take effort to review.
My approach is to use mostly closed questions, plus one or two open ones, placed late and marked optional. A good open question is specific: “What is the one thing we could change to make this more useful?” beats “Any comments?”
Handle sensitive questions with care
Questions about income, health, age or opinions on touchy topics need extra thought.
- Explain why you are asking. “We ask so we can tailor the examples” is enough.
- Make it optional. Forced answers to personal questions get false answers.
- Offer “prefer not to say.”
- Use ranges. Income or age bands feel less invasive than exact numbers.
- Put them late.
Tell people how their data is used, especially if you also collect names or emails. Privacy rules differ by country, so check the ones that apply to you.
Weak versus strong: a worked example
Imagine a hypothetical online shop surveying recent buyers. Here is a first draft.
- “How much did you love our amazing delivery and packaging?”
- “Do you shop with us often?”
- “Would you recommend us, and would you buy again?”
- “Rate us from 1 to 10.”
And a rewrite.
- “How would you rate the delivery speed?” (rating)
- “How would you rate the packaging?” (rating)
- “In the last six months, how many orders have you placed?” (single choice: one, two or three, four or more)
- “How likely are you to recommend us to a friend?” (rating, with both ends labeled)
- “What is one thing we could do better?” (short text, optional)
The rewrite splits the double questions, removes the loaded words, replaces “often” with a number, and labels the scale. It is also easy to analyze. For more ideas specific to customers, see customer feedback survey questions. For the whole process from planning to reading results, the WordPress surveys and polls guide puts these rules in order.
Using the six Quizably question types
Quizably has six question types. Here is how they map to survey jobs.
| Type | Good for |
|---|---|
| Single choice | Facts and preferences with one answer |
| Multiple choice | “Select all that apply” |
| Dropdown | Long lists people know, like country |
| Rating | Satisfaction, likelihood, usefulness, as stars, numbers or faces |
| Short text | The “why”, or a free comment |
| True or false | Better for trivia. In a survey, use single choice with Yes and No |
Set the quiz type to Survey in the builder, then add questions one at a time, and use the live preview to read each one as a respondent would. After it is live, the analytics show completion rate, per-question drop-off and answer distribution, so you can see which questions people skip or abandon at. To read individual answers, including short text replies, open a lead in the Leads screen or export the CSV. That only works for visitors who left an email, so add the lead form if you need to read every reply.
Editing checklist
Before you publish, go through every question.
- Does it ask about one thing only?
- Is the wording neutral, with no loaded words?
- Would everyone understand every word?
- Is there a time frame where one is needed?
- Do the options cover every real answer, without overlap?
- Is the scale balanced, labeled and in the same direction as the others?
- Is it in a sensible order, easiest first?
- Is it optional unless I truly need it?
- Do I know what I will change based on the answer?
- Have I taken the survey myself, on a phone?
Then ask one or two people to take it and tell you what they thought each question meant. Wherever their reading differs from yours, reword it.
Next step
Take the survey you are planning, cut it to the questions that map to a decision, and run each one through the checklist. Then build it in Quiz Builder and preview it on your phone. The full steps are in how to create a survey in WordPress, and if you are writing quiz questions instead, see how to write quiz questions. If you are still choosing a tool, how to choose a WordPress quiz plugin explains what to check for surveys, such as rating questions, open text and export.
Frequently asked questions
What is the most common mistake when writing survey questions?
Asking two things in one question, or wording a question so that one answer sounds better than the others. Both make the data hard to trust.
Should I use open or closed questions?
Mostly closed, with one or two open ones. Closed questions are quick to answer and easy to compare. Open questions catch what your options missed, but fewer people answer them.
How long should a survey be?
As short as your goal allows. Five to eight questions suits most audiences, and fewer is better for cold visitors. If you go past ten, ask whether it should be two surveys.
How do I ask sensitive questions?
Explain why you are asking, make the question optional, and offer a "prefer not to say" option. Put it late in the survey, after people are settled.
Which Quizably question types work for surveys?
Single choice, multiple choice, dropdown, short text and rating all work well. True or false is built for trivia, so a single choice with Yes and No is usually a better fit.