You're designing a survey and you really, really want people to answer all the questions, so you make them mandatory, right? Mmm, we don't think so. We don't recommend mandatory questions for these reasons:
- It's frustrating and annoying for people when they can't answer the question (maybe they're too new, or the question doesn't apply to their situation). You risk losing their responses altogether if they give up and exit the survey without submitting it. We prefer they have an enjoyable survey experience with the freedom to skip questions they can't or don't want to answer.
- The quality of answers is important. If people work out every question must be answered, they're more likely to rush the survey and choose the same answer for every question to get it finished quickly. We prefer to people to think about their answers.
- Participation in a survey should be voluntary, and forcing people to answer questions goes against this principle. People are more likely to give up and drop out when answers are required.
It's not just us worried about these affects - other researchers have shown that mandatory questions in online surveys will both increase the drop out rate and impact how people respond to the questions. Rest assured, when a person has skipped questions, they are prompted before submitting their survey. They can choose to continue and submit anyway.
When are mandatory questions useful?
If you're running a survey about a training course and knowing which course they are talking about is critical. You would likely have a demographic question built-in to the survey which you can make mandatory if you feel people may skip it.
Configuring mandatory questions
Survey and account administrators can mark individual questions as mandatory from the Survey Design configuration screens. See how to configure mandatory questions.
- Survey Design Guide
- Adding self-report demographics
- Study by Stieger, Reips & Voracek "Forced-Response in Online Surveys: Bias from Reactance and an Increase in Sex-Specific Dropout", JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE AND TECHNOLOGY, 58(11):1653–1660, 2007