How to Write Great Survey Questions

There are a few points you need to follow to ensure you are writing a top-notch questionnaire.

We have listed nine of them below:

Keep survey questions neutral
Try to keep open-ended questions to a minimum
Keep a balanced answer set
Don't force an answer
Avoid using the same question type repeatedly
Pay attention to the order of your questions
Let questions be optional
Consider using supplementary questions
Run a test survey

Let's try to discuss these in detail, shall we?

Keep survey questions neutral
Try to keep open-ended questions to a minimum
Keep a balanced answer set
Don't force an answer
Pay attention to the order of your questions
Avoid using the same question type repeatedly
Let questions be optional
Consider using supplementary questions
Run a pilot survey

Keep survey questions neutral

A leading question has a hidden opinion prompt in it. This can redirect respondents towards thinking something they may not be feeling. A leading question is one which tries to prompt or encourage the answer which the surveyor wants from the respondent.

Let's consider you have asked a leading question:

We really love the Human Resources team at Zoho. Tell us why you love them as well:

Asking the question like this kind of sounds like you are trying to impose the opinion of the surveyors on the people. You can instead tweak the question to give it an unbiased tone—something along these lines:

How helpful or unhelpful did you find the human resources team at Zoho?

Try to keep open-ended questions to a minimum

An open-ended question requires the respondent to formulate a more complex answer, which can take a lot more time and effort when compared to a close-ended question, which often only needs something as simple as a Yes or No. This extra time might result in a higher dropout rate for your survey. So while writing a survey, try to keep the number of open-ended questions at a minimum.

Tip:

You can also add all the open-ended questions for your survey as a separate page at the end of the survey to ensure a smooth experience for people. This way, the open-ended questions won't get in the way and make people drop off the survey without completing it.

Keep a balanced answer set

Respondents need a way to provide honest and good feedback. Otherwise, you can't really trust the responses to be credible. Here lies a very important place where bias often comes in to surveys: the answer choices.

Let's assume the following were included as options in a question asking about the helpfulness of a customer service representative:

You can see there's not an option for respondents to say that they did not find the customer service helpful. One must assume an objective tone while writing survey questions. Therefore, in this case, a better answer set would be:

This set of answer choices has different degrees nicely laid out, thus giving space for all kinds of feedback. To read more about such answer choices, go through our article on Likert scales.

Don't force an answer

There's another version of questions that only require limited or definitive (Yes or No) responses, which can really limit the range of possible answers and are generally not valid or reliable. These types of questions include "Do you always use this product?” or “Do you use all the features this product has to offer?” where the respondent can only choose from categorical Yes or No answers. Because a respondent may not use the product all the time, or may use it often but not always, forcing them into a Yes or No means that you won't really understand the full context of the response.

Ultimately, questions should be framed clearly and concisely. Cut out unneeded words and make sure that you've defined all relevant terms that might not be as familiar to participants as they are to the people conducting the survey.

Examples of clear questions include:

"How would you rate this product?" (rather than “How would you describe your rating based on feelings on this product?”)
"How often (or not often) do you use electronic devices (tablets, laptops, smartphones, etc.) in the span of a day (i.e., 24 hours)?"

Pay attention to the order of your questions

While questions can go in any order you would like, you will do your respondents a favor by grouping similar questions together. If you have a bunch of questions that use a scale or are all on the same topic, it's better to combine those questions into one section of the survey rather than spreading them out. This helps your respondents answer more efficiently, since all of the items are related and relevant to each other.

Grouping like questions helps improve the flow of the survey and can help an otherwise disjointed set of topics come together. Question repetition with variation can also be used to ensure that the responses you collect are accurate. But be careful, because the way you order your questions can introduce a question-order bias (discussed later in this article).

Pro tip:

It can be useful to include similar questions in different parts of the survey to test the validity of answers. In other words, if you want to test whether the respondent is answering in a consistent way, spread like questions across the survey and use them to check.

Avoid using the same question type repeatedly

It's a common mistake for surveyors to use the same question type multiple times throughout a survey. This either prompts respondents to leave the survey or engage in something known as "straight lining” where they keep answering survey questions without giving much thought to it.

You can proactively avoid this by varying the question types you ask and changing the question type when it gets used too often in your survey.

Let questions be optional

Let's face it—respondents don't know the answer to or might not want to answer all your questions. There may even be questions they are not comfortable answering. While it might be tempting to make all you questions mandatory to answer, it's better to ensure that you retain your respondents. Forcing your respondents to answer all your questions might make people quit the survey altogether.

Consider using supplementary questions

When appropriate, it's a good idea to include questions that help collect some extra data about a respondent. During the analysis, a surveyor might want to consider factors such as age, gender, and race/ethnicity. While differences among these groups might not be a driving reason for why the survey is being given, this information might be useful down the line—for instance, if someone wants to analyze group differences or follow up with specific subsets.

Run a pilot survey

As a survey creator, there's nothing worse than finding out your important survey has gone out with some elemental mistakes.

You can prevent this situation from happening by doing a test run of the survey where you send it to your colleagues, friends, and others. This test run is called a pilot survey in the industry. A fresh pair of eyes can often spot mistakes in your survey better.

The art of taking a good survey lies in a balancing act—you have to ensure that your questions are answered truthfully while providing respondents with an easy survey-taking experience.

Remember, the better the surveys are, the better the responses that will pour in. So apply these practices while creating your next survey!

9 tips for writing great survey questions

Keep survey questions neutral

Try to keep open-ended questions to a minimum

Keep a balanced answer set

Don't force an answer

Pay attention to the order of your questions

Avoid using the same question type repeatedly

Let questions be optional

Consider using supplementary questions

Run a pilot survey

Related Articles

Quantitative vs. Qualitative Research Survey

Scaling your survey responses using a Likert scale

Using surveys to improve customer communication during COVID-19