How do I run chi-square post-hoc analysis?

XH — Fri, 09 Jun 2023 00:27:00 GMT

Hello,

I am trying to run post-hoc analyses following a chi-square test. I have a binomial dependent variable (yes/no) and a categorical independent variable with 3 levels (A, B, C). When I use this same categorical independent variable to run post-hoc analyses against continuous dependent variables, I have used Dunnett's post-hoc with A as the referent.

I am trying to run post-hoc following a chi-square and have struggled to find a clear solution online. My understanding is to split my independent variable into 2 columns (A, B; A, C) and then to use Analyze>Modelling>Response Screening to compare A vs B and A vs C (effectively manually running post-hoc analyses while controlling for the FDR). Is this an appropriate way to do this, or is there a better method on JMP?

Thank you in advance :)

Re: How do I run chi-square post-hoc analysis?

Mark_Bailey — Mon, 25 Jan 2021 17:34:57 GMT

You can analyze such data with the Contingency platform and the Nominal Logistic platform, but neither one provides such post-hoc tests as you describe. There is the Odds Ratio for Independent report that might be useful. Here is a mock up of your study:

Here is the platform:

It is generally better not to divide the data up for separate tests. It reduces the power of the tests.

I do not recommend the Response Screening platform. It is generally intended to screen many response. The FDR in particular is an adjustment for very large tests and should not be used otherwise.

topic Re: How do I run chi-square post-hoc analysis? in Discussions

How do I run chi-square post-hoc analysis?

Re: How do I run chi-square post-hoc analysis?