|
Group-Sequential Tests for Two Proportions using Simulation
This procedure can be used to determine power, sample size and/or boundaries for group sequential tests comparing the proportions of two groups. The tests that can be simulated in this procedure are the common two-sample Z-test with or without pooled standard error and with or without continuity correction, the Mantel-Haenszel test, the T-test, and Fisher’s Exact test. For two-sided tests, significance (efficacy) boundaries can be generated. For one-sided tests, significance and futility boundaries can be produced. The spacing of the looks can be equal or custom specified. Boundaries can be computed based on popular alpha- and beta-spending functions (O’Brien-Fleming, Pocock, Hwang-Shih-DeCani Gamma family, linear) or custom spending functions. Boundaries can also be input directly to verify alpha- and/or beta-spending properties. Futility boundaries can be binding or non-binding. Maximum and average (expected) sample sizes are reported as well as the alpha and/or beta spent and incremental power at each look. Corresponding P-Value boundaries are also given for each boundary statistic. Plots of boundaries are also produced.
There are four procedures for group-sequential tests for two proportions using simulation. These procedures are identical except for the type of parameterization. The parameterization can be in terms of proportions, differences in proportions, ratios of proportions, and odds ratios. Each of these options is listed separately on the menus. Whether the parameterization is in terms of proportions, differences, risk ratios, or odds ratios, the hypothesis tests concerning the differences are those that are simulated.
A second but similar group of four procedures is available for tests where the null hypothesis difference is a value other than zero.
|
