FIGURE 1.

Power and expected sample sizes (ESS) under $\mathit{\theta }=(0.5,{\theta }^{(2)})$ and ${\theta }^{(2)}\in \{0,0.1,0.2,0.3,0.4,0.5\}$ for the FSD, 3‐arm 1‐stage ORD (FSD(h)), 3‐arm 2‐stage ORD, and MAMS(m) designs when all designs are powered at 80% to reject both hypotheses under $\mathit{\theta }=(0.5,0.5)$. 3‐arm 2‐stage ORD and MAMS(m) use triangular bounds. Results are provided using $10^6$ replications