post hoc test for Fisher's Exact test

monsterpie — Tue, 16 Feb 2021 15:28:32 GMT

Hi everyone,

I have 2 categorical variables: Var1 - 3 levels, Var2- 5 levels (i.e., table is bigger than 2x2). Since I have multiple cells <5, I have run a Fisher's exact test with Monte Carlo estimation for the p value (code and output is below). I understand that if I want to determine which specific cells are not associated, I need to perform a post hoc test, but I'm not sure which test to perform, or how to do this in SAS? From doing some research online, it seems like a Bonferroni correction might be a suitable choice? Any help would be very much appreciated.

CODE:

proc freq data=winter_t1;
title "Fisher's exact with exact option";
tables var1*var2;
exact fisher/mc;
run;

OUTPUT:

Statistics for Table of Var1 by Var2

Statistic	DF	Value	Prob
Chi-Square	8	434.4255	<.0001
Likelihood Ratio Chi-Square	8	186.6401	<.0001
Mantel-Haenszel Chi-Square	1	92.1028	<.0001
Phi Coefficient		0.7704
Contingency Coefficient		0.6103
Cramer's V		0.5447
WARNING: 27% of the cells have expected counts less than 5. Chi-Square may not be a valid test.

Fisher's Exact Test
Table Probability (P)	<.0001

Monte Carlo Estimate for the Exact Test
Pr <= P	<.0001
99% Lower Conf Limit	<.0001
99% Upper Conf Limit	0.0005

Number of Samples	10000
Initial Seed	814770001

Sample Size = 732

Re: post hoc test for Fisher's Exact test

SAS_Rob — Tue, 16 Feb 2021 15:36:20 GMT

You might try one of the suggested approaches in the usage note linked below.

https://support.sas.com/kb/22/565.html

topic post hoc test for Fisher's Exact test in Statistical Procedures

post hoc test for Fisher's Exact test

Re: post hoc test for Fisher's Exact test