Inferential Statisticshard
0:00.0
In a clinical trial with binary outcomes, a Wald-type interval for the log-odds ratio is constructed as . If the sample size is small and the true probability of success is near , why does this Wald interval often fail to achieve its nominal coverage probability?
In a clinical trial with binary outcomes, a Wald-type interval for the log-odds ratio is constructed as . If the sample size is small and the true probability of success is near , why does this Wald interval often fail to achieve its nominal coverage probability?