I have a treatment group (n=42) and a control group (N=12). From this this, 13 used in the treatment HTN drugs and 4 people in control group used HTN drugs. How should the contingency table be set up? Is the chi-squared / Fisher's exact possible in this case as the 13 and 4 do not equal the total?
I assume that there are only two possibilities for HTN drug use (namely, "yes" and "no"). Thus, there would have to be $42-13=29$ non-drug users in the treatment group, for example. It is straightforward to set up a contingency table with your data in this way (this example uses
c.table = rbind(c(13, 42-13), c( 4, 12- 4) ) colnames(c.table) = c("yes", "no") rownames(c.table) = c("treatment", "control") c.table # yes no # treatment 13 29 # control 4 8 chisq.test(c.table, simulate.p.value=T) # # Pearson's Chi-squared test with simulated p-value (based on 2000 # replicates) # # data: c.table # X-squared = 0.0245, df = NA, p-value = 1
Because the odds ratio is very close to $1$ (viz. $0.9$) and you have relatively few data, the p-value is high, but this is certainly doable.