Table 2 Target values, maximal possible deviations from the target values, and variances for the conditional performance score. CP: conditional power; N: total recalculated sample size; \(1-\beta\): anticipated conditional power value; \(\alpha\): significance level; \(n_{fix}(\lambda )\): sample size for \(\lambda\) in fixed sample size design; \(n_1\): first stage sample size; \(n_{max}\): maximal sample size
Performance measure | target value for \(\varvec{n}_{\varvec{fix}}(\varvec{\lambda }) \varvec{\le } \varvec{n}_{\varvec{max}}\) and \(\varvec{\lambda >0}\) (target value for \(\varvec{n}_{\varvec{fix}}(\varvec{\lambda }) \varvec{>} \varvec{n}_{\varvec{max}}\)) and \(\varvec{\lambda \le 0}\) | maximal possible deviation from target value | maximal possible variance |
|---|---|---|---|
CP | \(1-\beta\) (\(\alpha\)) | \(1-\alpha\) | \({\left( (1-0)/2 \right) }^2\) |
N | \(n_{fix}(\lambda ) (n_1)\) | \(n_{max}-n_1\) | \({\left( (n_{max}-n_1)/2 \right) }^2\) |