Survey Design in the Wild: Lessons to be Learned from Bad Survey Questions
Have you ever taken a survey and seen a question that was so bad that you HAD to screenshot it and share it with people? That motivation is what this talk is founded on. On my Twitter account (@kohizeri) I have an ongoing series of tweets I call ‘Survey Design Lesson Time’ where I share screenshots taken from surveys and forms that are circulating in the wild and conduct mini case studies. I call out the weakness present in the questions and summarize what lessons we can all learn from these mistakes when we go about designing our own surveys. The approach is somewhat humorous, but provides concrete lessons in survey design that will be valuable for all levels of GUR, from juniors and students designing their first questionnaires to veterans who maybe just need a quick refresher. The presentation will be short and modeled on my Twitter posts (and probably reusing quite a few I’ve Tweeted already, tbh): show a screenshot of a survey/form question, point out its weaknesses, and provide advice on how to avoid or resolve the issues identified when writing your own surveys. Advice will cover everything at a fairly high level (no arguing the merits of 5-vs-7 points or the validity of NPS here) and will cover topics like question type and wording, response options, question organization, and the struggles of making sure your survey works on mobile.
- Category: Talk
Elizabeth Zelle is a User Researcher at the Bethesda Softworks Research Lab located at id Software, outside of Dallas, TX. She’s been in the industry over 10 years, and in her 7th doing GUR. Prior to starting at Bethesda in 2017 she spent 5 years conducting user researcher at Volition, so she has a breadth of experience working with both developer and publisher stakeholders on a wide variety of genres. She’s also an undercover marine biologist.