How to statistically test for differences in post-build failure patterns among build months?

[Ferra]

I have monthly repair rate Nevada chart and I want to know how can I statistically test the difference between them (χ2-test?). Nevada chart is a kind of table which shows the repair numbers based on the build month and failure month. (we have other data for the other months but I just entered the data for the blocks that I want to compare). So, I have two questions. First, how can I compare any two same-size blocks together? second, how about they are not the same size? Also is there any way that I can do analysis on the whole table and show the part (block) that shows significant difference with the rest of the table?


http://stats.stackexchange.com/ques...rences-in-post-build-failure-patterns-among-b

I know that Poisson rate equivalence test might be useful for this case, but I have problem with applying that due to the format of the data..