TABLE 5.

Properties of the optimal adaptive enrichment (AE) design for different timings of the interim analysis $\tau$ when $\mathit{\theta }$ has the prior distribution given by (13) with ${\mu }_1=12$, ${\mu }_2=4$, ${\sigma }_1^2={\sigma }_2^2=25,$ and $\rho =0.75$. The interim analysis takes place after a fraction $\tau$ of the total sample has been observed

$\tau$	$P({ℛ}_1)$	$P({ℛ}_3)$	P(Enrich)	$E\{G(\mathit{\theta },X)\}$
0.1	0.14	0.58	0.13	6.84
0.2	0.17	0.56	0.23	6.87
0.3	0.19	0.55	0.28	6.89
0.4	0.19	0.55	0.31	6.91
0.5	0.18	0.55	0.33	6.91
0.6	0.17	0.56	0.34	6.89
0.7	0.15	0.57	0.34	6.88
0.8	0.13	0.58	0.32	6.87
0.9	0.11	0.59	0.27	6.85