average p values???

[crisisinmedicine]

I have to compare demographics of a study that was s .
In the "pre" group, I have 385 patients, in the post group, I have 317 patients. No patients are the same; I also took all of their ethnicities and compiled it into my MS excel spread via 0, 1, 2, 3, 4 as white, black, asian, native american, and other, respectively. How can I calculate a SINGLE P VALUE FOR THIS? HELP! I am struggling a great deal beacuse the two groups are not of the same size (385 vs 317). I simply need ONE p value, so what do I do? Should I run it was 0 vs. 0, 1 vs. 1, etc. etc., calculate the p values and take the average? Is that allowed? In group 1 I have 220 white patients vs group 2 has 213, again their is a difference in n, how do I compensate for this when conducting the chi sq? Thanks!!!!

Sincerely,
Crisisinmedicine

 

[noetsi]

One p value for what method? P values are the probability of getting a result given what is expected for a specific distribution(commonly). You can not talk generically of getting a p value - you have to explain what your hypothesis is and what method you are running to begin to address this.

 

[crisisinmedicine]

I want 1 p value to compare all of the ethnicity is what I'm looking for. So amongst the 385 patients vs. 317 I have 5 different ethnicities, but I'd like to calculate an overall pvalue to show that hopefully ethnicity was not a significant factor in my study.

 

[Dason]

It sounds like you just want to do a chi-square test of independence.