BookmarkSubscribeRSS Feed

What test do I use for two variables that are ordinal and seem to have a non monotonic relationship?


Community Trekker


Apr 23, 2017


I'm trying to find the right test for association/relationship between two likert scale, ordinal variables from a questionnaire I did:


  1. "How often do you recycle" - (1) Almost Daily, (2)2-3 times a week, (3)Once a week, (4) Less than once a week, (5)I don't recycle


  2. "Recycling can conserve the environment" - Ranges from (1)Strongly agree to (5)Strongly disagree, with (3)Neutral in the middle and a (6)I don't know option.

I want to see if there is a relationship between attitude of respondents and their actual reported behavior.


After graphing the data, it seems the variables have a non-monotonic relationship. So this seems to exclude some ordinal-ordinal tests which require a monotonic relationship. 

From what I've searched, Kruskal-Wallis test might suit my situation but I can't seem to find the option that allows me to either do the test or see the result (if JMP automatically does it for me). Is Kruskal-Wallis test a good test to use for my case? Help would be much appreciated.





Super User


Jun 22, 2012

For ordinal data, you can get several different measures of association if you analyze you data using Fit Y by X and it provides a nice mosaic plotl;

There are also several methods in the under the 

     Analyze==>Consumer Research

The Consumer Research area is very nicely documented in the "Consumer Research" book

     Help==>Books==>Consumer Research




Community Trekker


Apr 23, 2017

Ah, so JMP in fact automatically does a number of measures of association automatically then. All I have to do is click the "Measures of Association" option in the output window and it'll display the different measures of association. Just saying it out loud so that if I'm wrong you can correct me. Thanks!


measures of association.jpg