Question scales come in two types: seesaws and stairs.  The main point of differentiation of the two ideal-types is the presence or absence of a tipping point. See saws have them and stairs do not.

When you are thinking about a scale, you need to ask yourself is there a point on the scale that the scale tips to the other side because this will help ensure you have the right points on the scale.

Some questions naturally have a tipping point

Consider an agree-disagree scale. Regardless of the number of points on the scale, at some point the respondent must be on one side or the other (or precariously balancing in the middle). As you move from strongly agree to strongly disagree, at some point you move from on-balance agreeing to on-balance disagreeing. A yes-no question also has a natural tipping point. Thus a seesaw

But not all question scales have tipping points because they are asking, how much of something do you have? Just like there is no tipping point (a single place in the scale where it breaks into two halves) for income, there isn’t one for many questions that capture the amount of something like concern, interest, or priority.

How concerned are you about the safety of the food you buy?

The question above would be best captured with a scale that went from “Not at all concerned” to “Very or extremely concerned”. As you can see, there is no place where you specifically tip from one side to the other. The scales act more like stairs or a ladder with artificial stepping points on an underlying dimension.

Of course, you can turn this into a seesaw scale by polarizing the poles of the scale (e.g. very unconcerned to very concerned).

Why is matters

To some extent using a seesaw versus a stair scale is a matter of preference but there is one place where it matters. When you have a stair type scale that moves up from nothing to a lot of some thing (e.g. importance, concern), then there is little place for a neither category in the middle.

Neither categories fit nicely into the middle of seesaw questions because a respondent can refuse to take sides — balancing in the middle. They do not work the same way for a stair-type scale. Respondents who are unable or unwilling to answer a question must choose a category outside of the scale for these questions such as “Don’t know”, “Not applicable”, or “No opinion”. These non-answers can’t be in the middle now but rather must be outside of the main scale.

So do you prefer scales that force a tipping point or those that allow people to place themselves on a continuum from nothing to a lot of something?