Good survey questions will lead to more reliable and useful results from your questionnaire. Follow these tips for preparing your online survey.

One of the advantages of online surveys is that you can assemble them quickly and get results almost immediately. However, this does not mean you can approach them "head-on" and without thinking. If you take the time to prepare good survey questions, your questionnaire will yield more reliable and useful results.

Your first choice will be which type of question to use. You can ask open-ended questions, where the user answers in their own words, or closed-ended questions, where they choose from possible answers.

1. Prioritize Closed-Ended Questions

Open-ended questions require more time and effort from the respondent because they will have to formulate and write their answer. Additionally, it will take you more time to analyze the answers. This, of course, does not mean you should completely avoid open-ended questions, as they will often yield the most honest answers, which are not influenced by the suggestion of possible answers. However, you can limit them to, say, a maximum of 2 open-ended questions per survey. It is also good to ask them at the end. This way, you will get at least some answers from the respondent if they do not complete the survey.

2. Keep Questions Neutral

Avoid your own assumptions and biases when composing questions. If you hint to respondents what answers you expect, you will influence the results, which will thus not reflect the real situation. The question: “

We are thrilled with our new website. Do you also find it useful?

” will influence respondents to agree with you and give you positive answers. Instead, ask:

“Did you easily find the information you needed on our redesigned website?”.

3. Ensure Answer Choices are Balanced

Respondents need a way to give you their honest feedback. Only then will your survey results be credible. Just as it is important for your questions to be neutral, it is also important for the answer choices in closed-ended questions to cover all possible opinions of the respondents. If you ask: “

How do you rate the cafeteria's lunch offerings?

”, it will not be enough for the possible answers to be only:

good      very good      excellent

Your result might indeed indicate that the offering is good, but it will not be credible, as the answer choices did not allow for the expression of dissatisfaction with the offering. For credible results, you must also be neutral in your answer choices. For rating questions, the spectrum of answers should be:

very poor      poor      neither poor nor good      good      very good

4. Don't Ask Two Things at Once

Respondents who don't know exactly what you're asking will not give you an honest answer, whether knowingly or unknowingly. The most common cause of confusion is so-called double-barreled questions, which refer to two different things. Instead of asking: “

How do you rate the employees' professionalism and their willingness to help customers?

”, it is better to ask two separate questions. This way, respondents can rate professionalism as well as willingness to help, instead of giving you one rating that will not reflect the reality of either. Therefore:

“How do you rate the employees' professionalism?” and

“How do you rate the employees' willingness to help customers?”

5. Don't Ask the Same Thing Multiple Times

You are probably annoyed when you are asked the same thing over and over again. Respondents feel the same way when they have to express their opinion on the same matter multiple times. First, they will think that the survey designer did not put enough thought into it. They will give you the same answers, even if you ask something else in between, and sooner or later they will abandon the survey and close it. In either case, you will get unusable results, or no results at all. Take time to plan your survey, use different types of questions, and view the survey as a whole before opening it for responses. Eliminate questions that are too similar.

6. Not All Questions Should Be Mandatory

Respondents may not know the answer to all the questions you want to ask, or they may not want to answer them. Don't let them leave because of this and not answer the questions for which you need their honest answer. Choose a few questions that are truly crucial for your research, and leave the rest as optional. If you are unsure whether a question should be mandatory, it should not be. Respondents who are forced to give an answer they don't want to give will answer randomly or stop taking the survey.

7. Test Your Survey Before Launching It

Once you start getting answers to your survey, it's too late to fix it. Even if the tool allows you to make changes, you will have to discard the answers already obtained because they will distort the results if you change the survey. You can prevent this by putting yourself in the respondent's shoes and taking your survey. Send it to a few friends or colleagues and ask them to tell you about any uncertainties or concerns they have while taking it. Check if you have followed the above advice and if the survey is attractive enough for respondents to complete it. If the test is successful, open it and invite as many people as possible to participate. Well-designed questions will allow respondents to give honest and credible answers. It is also important to choose a tool for creating surveys that will allow for quick completion and a good user experience. The better your survey is, the more useful answers you will get, so follow these tips when preparing a new online survey.