Designing a survey or questionnaire can be a daunting task for the uninitiated. To write good survey questions and get reliable results and insights from your surveys you can follow the steps outlined here.
Like most things good in life, good surveys require good planning!
- Firstly grab a sheet of paper or open a document.
- Do not attempt to use survey software at this point! You are at the planning stage.
- Write your survey title
- Don’t get too caught up in the title at this stage, as you can change it later. Defining the title will give you focus on the topic of your survey.
- Write your objectives.
- What would you like to find out? What are the objectives of your survey, what do you want to find out? Write a list of these items, they will form the base of your surveys questioning.
- Group your objectives
- Have a good think about your objectives and what you would like to achieve. Can these be grouped together? You will often find questions are similar and can be grouped. Let’s say for example you have various questions about cars, you may like to ask which make, model, engine size, number of doors within 1 question rather than having several different questions.
- Survey logic
- Survey logic may initially sound daunting. what is survey logic? Here is an example of survey logic. You may ask the participant if they have a car. The answer to this question will then determine the flow of the survey and if the participant is asked further question son cars. Identify these questions as you will need to place these questions near the start of your survey.
- Survey software
- Open your survey software and begin to layout the survey in the software adding your questions, subquestions, and answers for each question. Once your questions have been added you can then add the logic to your survey and test it works!
When creating your survey in the software, one of the first choices you have to make is the type of question to use. Some questions will be both open-ended text input questions that ask participants to add comments or opinions. The second option is closed-ended or multiple choice style questions that allow the participants to choose from a fixed set of options. Closed questions range from
Likert rating scales, star rating and multiple choice options.
10 Tips for creating fantastic surveys
- Primarily ask closed-ended questions. Depending on the size of your survey and your objectives multiple choice and closed questions are usually the best question type. This is because open text questions require the participant more thought, effort and time to answer. To many of this type of questions will cause your participants to abandon the survey without answering. If possible, put your open questions on in a separate group at the end of your survey/form. If the participant drops out at this point, then you still have their responses from the rest of the survey.
- Ask neutral questions. Asking leading questions or putting opinion in your questions will influence respondents answers. Say you asked the question: “We think the best make of car is BMW, what do you think?” This is a leading question that would encourage people to choose BMW from the options. The question convey an opinion! You could better write the question as “What do you think is the best car manufacturer?” then have BMW in the answers, you can also use randomization to reduce bias.
- Keep the question answers balanced. Respondents need a way to provide a balanced and honest opinion, otherwise, the credibility of their responses is dubious. Answer choices included can be a source of bias. For example, the following answer options are used when asking respondents how good your customer service reps are: (a). Extremely helpful (b). Very helpful (c). Helpful. You’ll notice that there no choice for the participant to say that customer service is crap! We need to make the choices more balanced set of answer options by giving the participant choices d. Unhelpful e. Very unhelpful
- Only ask for one thing at the time Asking for more than one thing in a question will confuse the respondent. It is likely that they will choose an answer that doesn’t reflect their opinion. A common type of “double-barreled” question is asking respondents to assess two different things at the same time. For example: “How would you rate the brand and performance of your car?” Brand and performance of your car are two separate issues. Having both options in the same question will influence the participant to either evaluate one or to skip the question. The best way forward with this is to split the question into subquestions.
- Don’t keep asking the same question over and over. If someone believes that you are asking the same question or similar over and over they will not engage well with the survey and the results will become skewed and they will answer your questions without putting much thought into them. You can address this in the planning stage by planning to vary questions you ask, how you ask them, changing the order of the questions that are similar.
- Do not force participants to answer all of the questions. Some of the participants may find the answers difficult or may not have the answers. There may be some questions they can’t answer. Keep this in mind when deciding which questions require answers. And unsure make questions optional. Forcing respondents to answer questions they can’t is likely to make them abandon the survey or choose a random answer to get past the question.
- Test, test, and test again!!. Sending an untested survey can be fatal. It is far simpler to correct an error in your survey prior to it being sent to participants. To assist with your testing, ask friends family and colleagues to take the survey and give you feedback on it. Most importantly test it yourself several times.
- Use logic. Don’t keep the participant answering irrelevant questions. When there may be a selection of the questions that depend upon the answer to another question, use logic to skip over these questions.
- Use images. A picture is worth a thousand words. To get your message across better, you can use videos and images in your questions.
- Keep the survey short. Attempt to keep your survey as short as possible by combining questions and data points. Do your questions more than meet survey surveys objectives? If the participant feels that you are wasting their time or the survey is going on too long, they’re likely to abandon or choose random answers for the remainder of the survey. Use logic. Don’t keep the participant.
A good survey requires good planning, good questions