To Label or Not to Label

Question tables and scales are a common element of most surveys. Both ends of the scale represent polar opposites (for example, "Very Dissatisfied" and "Very Satisfied"), and in between there are several spaces for less polarized opinions, filling out the rest of the scale so that respondents can easily provide their opinions. As researchers, you have to decide how you want to create this scale.

You have the option of simply leaving the scale as is so that respondents can choose how they want to interpret the remainder of the scale, or you can mark the definition of each scale point individually. For example "Satisfied, A Little Satisfied, Neither Satisfied Nor Dissatisfied" etc.

The primary benefit of leaving the scale blank is the visual stress. When every number is given its own scale, the individual is forced to read and take in all of the words on the scale in order to decide which answer is right for them. That much information in the face of the respondent can increase drop out rate or cause frustration if the scale looks too cluttered.

Part of this has to do with formatting, as well as how educated the respondents are and how large your question table is (if you have one). However, defining the scale may be more beneficial. Without a definition, respondents may define the scale incorrectly for themselves, such as choosing the middle answer because they deem it meaning "very slightly satisfied." If the individual sees the scale differently than some other respondent sees the scale, they may be providing identical answers but have those answers fall on different scale points, throwing off your data.

There's no science on which of these choices is better, so the decision to use one or the other is ultimately up to your research. Take special care to look at formatting. If the way you have formatted the scale looks too cluttered, it may be advantageous to leave the scale labels off.

If, on the other hand, your scale is well formatted and each individual scale name is easy to read and understand, it may instead be more useful to label each one. Ideally a label would work best to avoid confusion overall, but only if the way you have built your survey doesn't cause respondents to drop out from clutter.