I am looking to analyze nonparametric, related sample data using a test similar to the McNemar. My independent and dependent variables are not dichotomous (7x5). For example, if my IV was level of anxiety (five levels) and my DV was rate of criminal offending (seven levels), how could I determine the association between anxiety level and membership in criminal offending groups? Any suggestions on what test would be appropriate?

  • 2
    $\begingroup$ McNemar test needs two variables with identical categories which is not your case. Use chi-square if you treat your variables as nominal or correlation (such as Kendall tau-c or Spearman) if you treat them as ordinal or polyserial correlation if anxiety is ordinal and criminal rate is scale. $\endgroup$ – ttnphns Oct 12 '12 at 5:55
  • 1
    $\begingroup$ Please, specify types of variables or paste the table. $\endgroup$ – Miroslav Sabo Oct 12 '12 at 6:06
  • $\begingroup$ I don't know offhand, I will try to look later, but the book to look for on this subject is Agresti on ordinal data. $\endgroup$ – Peter Flom - Reinstate Monica Oct 12 '12 at 10:24

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.