Research has shown that respondents are typically less comfortable revealing their
exact income than they are revealing a grouping that it falls within (see Dillman, 2000).
For this reason, an ordinal-level measure was developed instead of a continuous-level
one.