A Study of the Classical Asymptotic Noninferiority Test for Two Binomial Proportions

Abstract

In a recent article by Li and Chuang-Stein (1), an evaluation was made of the performance of two frequently used methods—the classical asymptotic normal approximation and the same method with the Hauck-Anderson continuity correction—to test noninferiority between two proportions. The evaluation, using simulation, estimated type I errors and power. Continuing the research of Li and Chuang-Stein, this study evaluates the performance of these two methods. However, here type I errors are not estimated from simulation, and levels of significance are computed by enumerating all possible cases rather than through simulation. And finally, instead of the confidence intervals approach adopted in Li and Chuang-Stein, our work uses the hypothesis tests approach. The results show a comparison of the two methods that is different from that obtained by Li and Chuang-Stein.