I recently did a week-long juice fast at a wellness center. Upon arrival, I was handed a questionnaire asking about my eating habits and how often I suffered from each of a long list of ailments.
“I’m really struggling here,” one of my fellow fasters confided one day. “It might be because I lied on my questionnaire,” she confessed. “They asked how often I had headaches, but the only options were never, often, very often, or constantly. I get a headache once or twice a month. That’s not often! So I just checked never.”
I immediately saw a ramification of poor questionnaire design: An unclear scale leading to confused respondents leading to inaccurate data leading to the wrong actions being taken.
Whether you’re doing an academic study or a simple intake questionnaire, scales are serious business! People take information from a scale just like they do from a question. A survey interested in measuring people’s drinking habits might use a scale like this:
How many alcoholic drinks do you consume in a typical week?
None * 1-2 * 3-4 * 5 or more
Someone who has 7 drinks a week might be hesitant to put themselves in the uppermost category, thinking it implicates them as a heavy drinker. This scale might lead to more comfort on the part of respondents and more accurate data:
Fewer than 5 * 5-8 * 9-12 * 13 or more
Whether and how scale points are labeled (is it 1-4 or “excellent, good, fair poor”?), the number of points, and even the spacing of scale points can all impact data dramatically— and making sure all respondents can accurately place themselves on a scale is critical!
Lesson learned: When it comes to respondents giving accurate, actionable data, scales are as important as the question wording!