Subscribe Bookmark RSS Feed

Non-parametric tests and ordinal data

abmu

Community Member

Joined:

Aug 8, 2012

My first day with JMP, hope someone can help..

I work with data on an ordinal scale and need to make a Kruskal-Wallis and Wilcoxon signed rank test. I have entered the data in JMP, but when i set data as "ordinal" in "Fit Y by X" I get a "Contingency analysis" and from here I cannot choose the above mentioned non-parametric tests. However if I set my data as "continous" (which they are not) in "Fit Y by X" I get a "Oneway analysis" and from here I can do my tests.

Why can´t I choose Kruskal-Wallis/Wilcoxon when my data i set as "ordinal", these are the test that should be used when working with "ordinal" data?

Should I just set my data as "continous" (even though they are not), and perform the tests?