Solved – How to Compare Mortality Rates Among 5 Groups

You want to see whether the distribution in frequencies among the five groups is consistent with a discrete uniform distribution (null hypothesis) or they are instead different enough to reject the null. To do so, you can use a goodness-of-fit test chi squared test.

If you are using R, you can do something like this:

group_mortality  <- c(15, 21, 11, 09, 22) expected_freq    <- rep(1/5,5) (ch_sq <- chisq.test(group_mortality, p= expected_freq))  Chi-squared test for given probabilities  data:  group_mortality X-squared = 8.6667, df = 4, p-value = 0.06999 

In a case like this the $chi^2$ value wouldn't be extreme enough to reject the null with a risk alpha of $<5%$.