Table 1.

Glossary of Terms

Notation	Definition
θ0	Null response rate
θa	Alternative response rate
K	Total number of baskets
A	Number of truly active baskets
ε	Target family wise error rate when A = 0
(1 − β)	Target marginal power when A = 2 (or 3 depending on K)
(1 − β)min	Minimum acceptable power when A = 1
n1k	Stage 1 sample size for basket k
N1	Total stage 1 sample size
n2k	Stage 2 sample size for basket k, given heterogeneous design path
N2	Total stage 2 sample size, given homogeneous design path
γ	Assessment of heterogeneity tuning parameter
rS	Minimum required number of responses in stage 1 for an individual basket to continue to stage 2, given heterogeneous design path
rC	Minimum required number of responses in stage 1 across all baskets to continue all baskets to stage 2, given homogeneous design path
αS	Significance level for final separate analyses (before correction for multiple comparisons), given heterogeneous design path
αC	Significance level for final combined analysis, given homogeneous design path
FWER	Empirical family wise error rate
Pk	Empirical marginal power (%) for basket k = 1, …, K
EN	Expected trial sample size
ET	Expected trial duration (months)