Blog
Bits and pieces from the Crowd Scientists
Questionnaire Design: Response Scales - Part I
Part I: Why use a scale?
Questionnaires, in market research anyway, are somewhat like the eyes – a window to the soul. Okay, maybe a bit too lofty, but we rely so heavily on questionnaires to access information on people’s attitudes and behaviour, so a well-designed questionnaire is vital to getting to the truth. Of course, a questionnaire that is poorly designed will get you answers – just not necessarily the right ones. It’s that old adage of GIGO (garbage in, garbage out)… after the initial stages of research and sampling design have been addressed, it’s the questionnaire that can make – or break – a research project.
There are so many aspects of questionnaire design I could talk about, so in the interest of starting somewhere, I thought I’d start with a topic that tends to incite a lot of discourse, and, sometimes, debate: scales – specifically, response scales. There is so much to say about response scales that I think this discussion will be best addressed in a series of shorter pieces, so stay tuned for more, after this first installment. And now, to begin….
Why use a response scale?
One key aspect of quantitative, closed-ended questionnaire design involves not only how the question is worded, but also how we design the response categories. Why only ask whether a respondent plans to buy a new car in the next 12 months, when we can ask the degree of likelihood to buy? Why only ask whether a respondent is satisfied or dissatisfied with a product or service, when we can ask how satisfied or dissatisfied they are? Scales that go beyond simple dichotomous response options open up a whole set of possibilities when it comes to understanding respondents.
Shades of grey
A response scale offers so much more detail than a simple ‘black or white’ type of response like yes or no, do you or don’t you, will you or won’t you. A scale is almost like having a dimmer switch – the respondent can dial up or dial down the intensity of their response. We are then provided with detail on the degree of activity, intent, or attitude, and also have more options with regard to the statistical analyses we can conduct with this type of detailed response. We can now make smarter decisions on how or where to allocate resources and efforts to maintain and grow customer bases/audiences/etc. Proper use of response scales can help us to understand where the risks are, and be in a better position to minimize them.
So next time you catch yourself asking “whether”, see if the question translates into a scale format that now allows you to ask “how much”. It’ll go a long way in providing a fuller picture, and better answers.