which test to use

#1
Hi guys, I am not a statistician, but need to do a small project. (sorry if this sounds too easy for you). I've done my homework, just want to check if I am doing this correctly. I have two independent groups of patients. One group has 180 patients, second group 100 patients. I am measuring a level of certain medication in blood. I divided patients into two groups those that have the level within 10mg (low), and those who have the level above 10mg (high). I want to compare the proportion of low/high between the two groups. Am I correct to use Fisher's test?. I am planning to use this online calculator http://graphpad.com/quickcalcs/contingency1/ and enter the following

Outcome1 Outcome2
Group 1 100 (low) 80 (high)
Group 2 77 (low) 23 (high)

which gives me a p value of 0.0004. Am I doing it correctly, also, under what circumstances would I use a chi-square test instead?

Thanks a lot!!!
 

hlsmith

Omega Contributor
#2
Yes, you appear to be doing everything correctly. You could use the chi-square calcuulation here, however the Fisher's is an Exact Test which I always use in lieu of the Chi (asymptotic), when the groups or sample sizes are not too huge (requiring alot of processing power).

Not, you just have low and high with no middle group, no sure if this is your typical nomenclature, but perhaps Lower/Higher may see less mysterious : )