Examining correlation

Contingency Tables are good visualization method, with counts, percentiles in your case a 5 x 5 mosaic plot and table of counts, etc. Chi Sq tests use likelihood ratio and Pearson tests for example, but there are numerous options in stat software for analysis of those mosaic plots and their contingency table data. And of course the Nominal Logistic Regression Modeling tools have effects tests (Wald, Likelihood Ratio) for the main effects and interactions of your model. JMP.com or most other stat software tools support this type of data. Pasted below are list of OPTIONS for the Mosaic Plot and its Contingency Table from JMP help file (no detail, just names or tests and analysis options for your consideration). This list is property of JMP.com