Fisher’s Exact Test is used to determine whether there is a difference in the population proportions between two small groups. The definition of “small” is variably defined. Many users would define small as when the total number of patients is each group is less than 30 or when the absolute number of successes in each group is less than 5 and when the overall incidence is less than 20%.
- H0: π1 = π2 (there is no difference between the two portions)
- H1: π1 ≠ π2 (there is a difference between the two portions)
Example: You might use Fisher’s exact test to compare the proportion of patients that had surgical site infection between two small groups of patients when the incidence of infection was less than 20% in each group.