Best practices of survey design

1. Start from the end 

   Think about the goals of the survey. Starting your survey design process from the end will ensure your questionnaire is consistent and relates to your goal.

   So before drafting that first question, ask yourself:

   • What kind of information do you need from the respondents? 
   • What will the end data be used for? 
   • What format should it preferably be in? Numerical, qualitative, or geolocated data?

   Knowing this will help you a great deal in choosing the right question types and content.

2. Short and sweet is (almost) always better 

   You are more likely to get answers if your survey is not overly long. This is relevant both for the survey as a whole and for individual questions. Don't compromise on the quality and clarity of your survey, but keep the focus on what you really need to know.

   Sometimes it’s impossible to avoid longer chunks of texts, such as a privacy policy or further information on the research project. Remember that  you can always 'hide' these types of texts in a pop-up behind a click. In this way, the text is present and those who are interested can easily access it. At the same time, it won't create unnecessary clutter.

3. Avoid jargon and official language in your survey 

   Keep the language simple and easy to understand. If you are asking for public opinion, you should use the language the public is familiar with! 

   It can be hard to figure out how to discuss specific terms in plain language. What can help is checking discussion groups on Facebook or Reddit where you can pick up on how your future participants speak about certain topics.

   Also, check the readability of your questionnaire (which refers to the previous principle of a good survey). The readability level indicates how difficult it is to understand a text. For English, the Flesch-Kincaid readability test is the most optimal (here you can check your survey questions for free: https://goodcalculators.com/flesch-kincaid-calculator/). Aim for the 8th or 9th-grade readability level, but always have your audience in mind. 

4. Remember to make your survey accessible to all

   Maptionnaire's survey tool adheres to both the EU Accessibility Directive and the Web Content Accessibility Guidelines (WCAG 2.1), where we have AA-rating. 

   However, this does not mean that you shouldn't consider accessibility when creating your survey. Here are some guidelines for making your survey accessible: 

   • Since maps cannot be made accessible with screen reader technology, make sure to always combine a map question with an open question as an alternative way of responding. This way those who cannot access the map can still answer the question.
   • When you add images to Maptionnaire, remember to always provide an alternative description (alt-text) of the content for people using screen readers. You can do this by going to Materials and editing Alternative text setting in the image files.
   • Never use just color or just a position on the screen to indicate something. That is, “click on a green button in the right upper corner” is much better than “click on a green button” or “click on a button in the right upper corner”.
   • Remember to check that the contrast between for example text and its background is not too weak. This also goes for maps, e.g. for layering shapefiles on top of a basemap.
   • It’s also a good idea to provide alternatives to text-only answers and to provide information in multiple formats. In Maptionnaire, you can use images as answer options, allow respondents to submit their own images and voice messages, or include audio and video clips in your surveys.
   • Translate the survey. Urban areas in particular are more culturally diverse than ever, and thus it's always better if you can provide versions of the survey in at least all the major languages spoken in your area. This is not just about being able to understand the content of the survey, but about building trust between communities.

5. Use open questions sparingly 

   Typing a long-form answer to an open question takes a lot of time and can lead to survey fatigue. In turn, it can result in a respondent leaving the survey incomplete. The analysis of open questions also cannot be easily automated. So if you gather a great deal of written material, you also need to reserve the time to read and categorize it. 

   This is not to say you should not use open questions, but to encourage you to use them thoughtfully in places where no other question type can do the same job adequately. Why not ask a respondent to mark a place on the map instead of tediously describing which crossing is especially insecure for biking? With Maptionnaire, you can easily design these types of questions. Check this article for more on the use of GIS and geolocated data in questionnaires. 

   But it’s always good to have at least one open question per survey, where respondents can leave any comments and feedback they couldn’t communicate in other questions.

   You can use various types of questions to survey public experiences but also give a chance for leaving any other comments.

6. Visual communication

   Just as answering open questions can result in survey fatigue, and so can having to read a lot of text. Use visual communication whenever possible! 

   Use images of objects instead of describing them (but remember to add alt-text then, see the best practice of survey design no 4), schemes to visualize processes, and infographics to summarize research results. 

   Photobanks (e.g. Unsplash, Flaticon) have a wealth of free-to-use images, and tools like Canva help you create awesome illustrations without any graphic design skills.

   Maptionnaire offers many ways for making your survey more visual. In addition to using images as answer options, you can use them within the text, as static or interactive backgrounds. We also offer a variety of maps in different styles to add some visual punch!

   For more tips, check this article on creating visual questionnaires. 

7. Test, test, test! 

   Always test your survey: does it flow smoothly? Are the questions arranged in a logical way? Is the survey difficult to answer on a mobile phone?

   Also, take a look at the data you get in your results file. Is it categorized the way you imagined? 

   The best way to test a survey is to gather a group of voluntary testers who are in your target group. Remember, you can always make changes to a Maptionnaire survey at any stage!

When in doubt, contact us at support@maptionnaire.com. We are always happy to answer any queries you may have. Did you know that we at Maptionnaire also offer custom-made surveys?