Many data that appears ordinal at first glance, isn't. For example, consider the rating scale:
- Strongly agree
- Somewhat agree
- Neither agree nor disagree
- Somewhat disagree
- Strongly disagree
- Don't know
This data really consists of an ordinal scale nested within the first category of a two-category nominal scale containing labels of Know and Don't Know:
- Know
- Strongly agree
- Somewhat agree
- Neither agree nor disagree
- Somewhat disagree
- Strongly disagree
- Don't know
- Don't know
There are a variety of ways of constructing analyses for such data, but outside of some specialist areas of modeling, normal practice is to either:
- Set the unordered categories (e.g., don't know) as missing values and treat the data as ordinal or numeric.
- Treat the data as nominal.
The term nominal-ordinal is not a standard term. Often such data is (incorrectly) described as being ordinal.
See also
Comments
0 comments
Please sign in to leave a comment.