2
$\begingroup$

Can a variable that contains 'don't know' as a response be ordinal?

For example, say the variable corresponds to responses to the question 'Do you agree with ...?' where the responses are:

  1. Agree Strongly
  2. Agree Somewhat
  3. Agree Slightly
  4. Disagree
  5. Don't Know

The first four answers would form an ordinal variable but adding 'don't know' seems to violate this. Would the variable now be categorical?

$\endgroup$
  • $\begingroup$ This comment preempts the waggish response "Good question, but I don't know". $\endgroup$ – Nick Cox Dec 18 '14 at 18:46
3
$\begingroup$

"Don't know"s are necessarily problematic here. If "Don't know" was explicitly presented as an intermediate category in the questionnaire, there is a case for treating it that way.

That would be, in my experience, highly unusual; and it doesn't appear to have been the case here.

I'd argue that the only defensible actions otherwise are

  1. to regard "Don't know"s as missing and to exclude them from the model fitting

  2. to retreat to a multinomial model that doesn't regard the response or outcome as ordinal (but don't be surprised if the "don't know"s don't fit easily into any model).

(I'd assert that "categorical" (wide sense) as including ordinal is more common in statistically flavoured literature than "categorical" (very narrow sense) as meaning nominal only. It is, I think, uncontroversial to underline that narrow and wide senses are both encountered.)

$\endgroup$
2
$\begingroup$

You could put don't know intermediate:

  1. Agree Strongly
  2. Agree Somewhat
  3. Agree Slightly 4 Don't Know
  4. Disagree

But I suggest making "Don't know" = missing and leave the other 4 as ordinal. This would be a stronger recommendation if you already had "neither agree nor disagree".

Also note that your scale is tilted toward agree; this may be a bad idea or a good one.

$\endgroup$
  • 1
    $\begingroup$ There is a little typo - missing the dot after 4 in your bullet list means it comes out in a way you clearly did not intend! $\endgroup$ – Silverfish Dec 19 '14 at 0:48
  • $\begingroup$ In what situations might having a surplus of favourable options be a good thing by the way? I am struggling to see an advantage; such a survey would generally look less convincing to a third party because of the possibility of bias from the framing of the question. $\endgroup$ – Silverfish Dec 19 '14 at 0:52
0
$\begingroup$

If anything "don't know" could be in the middle:

  1. Agree Strongly
  2. Agree
  3. Don't Know
  4. Disagree
  5. Disagree Strongly

Here's a similar Likert scale discussion.

Also, as @PeterFlom noted, you have more "agree" answers than disagree. Likert himself recommended using equal number of right and left answers. The paper is rather fascinating, btw, I suggest reading it in full.

$\endgroup$
  • $\begingroup$ What about if there was already a 'neither agree nor disagree' response in the middle and a 'don't know'? $\endgroup$ – mattdevlin Dec 18 '14 at 18:37
  • $\begingroup$ What's the meaning of "dunno"? That's where you should start. What information does it convey? How different it is from "neither..."? Do you think people "know that they neither agree or disagree" as opposed "dunno"? $\endgroup$ – Aksakal Dec 18 '14 at 18:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.