Hello,

I need help determining which statistical test to use when comparing discrete variables pre and post. The question I am looking at is to statistically compare the number of hospitalizations (DV) before and after an intervention on the same sample. Since the DV is not continuous and the data in this instance is not normally distributed I cannot use the paired-t- test. I also don’t believe I can use the non-parametric version of the paired t-test (Wilcoxon Sign Test) because the DV is not a continuous or ordinal level variable. Any suggestions on what the appropriate test would be to compare pre and post?

Thank you,