I have a list of 50+ ordinal variables which have been rated by two researchers. The ratings go between zero and three. I am unsure how to combine these two ratings. It feels natural to take an average, however I end up with scores like 2.5 - which (one could argue) are does not make sense in the ordinal context.

I'm not sure how to proceed! If you have any ideas it would be great to know.

Thank you.

  • $\begingroup$ Since this is highly relevant to psychology research especially with Likert-like items, I think it is on-topic here, but nonetheless you may get better answers at stats.SE. Searching for your title there returns many results you may find useful. In summary, there is no strict principled way to do this, but there are norms by field which include treating ordinal data as if it were continuous; the more items there are, the more it may make sense to do this, but ultimately however you do it is "wrong" from a very strict sense. $\endgroup$ – Bryan Krause Jul 10 '19 at 15:39

Your Answer

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

Browse other questions tagged or ask your own question.