If you’ve taken any kind of survey you’ve mostly likely answered a multiple choice survey question presented in one of several different ways. Why is the multiple choice survey question so popular? It’s because they:
- are easy to understand
- are adaptable to different situations
- are good at returning structured responses
- make survey completion easier.
But there are also a couple of things you need to know about using multiple choice questions. First, you inherently have to limit the answers. In the process of narrowing down the options for answers you could reflect personal bias. Bias in the available answers means bias in the results. So, be careful. Second, there are several types of multiple choice questions. Knowing when to use a particular type is important.
Multiple answers or just one?
Multiple choice questions can allow a single answer only. This works well for a number of question types:
- Yes or No
- Agree or Disagree
- Pick your favorite option from a list
- Pick your least favorite option from a list
Multiple Choice, Multiple Answer questions can also allow many responses, including a range of responses, a maximum number of responses, a fixed number of responses, a minimum number of responses, etc.
For example, rather than forcing respondents to choose a single favourite pizza topping, a multiple choice question could ask “Which pizza toppings do you like?” and allow respondents to choose all the topping options they like.
Or, possibly a question might allow a range of responses. For example, Please select the 5 most important strategies for us to work on. Or, Please select between 3 and 5 of the most important strategies for us to work on.
Include an ‘Other’ option
One drawback of having a list of multiple answers is that you can introduce bias into your results. What if the respondent wants to indicate a choice you haven’t provided?
Consider adding an answer option of ‘other’ with a “please specify” (or a comment field) at the bottom of the list of answers. This can allow a respondent to enter a custom response. This can help avoid bias, but it reduces your analytical flexibility, and means your analysis may take longer.