Skip to main content

Table 1 Parameter estimates for the three potential designs for the motivating example. A simulation was performed to calculate expected sample size, Type I error and power

From: Generalizing boundaries for triangular designs, and efficacy estimation at extended follow-ups

Estimate Design 1 Design 2 Design 3
  (3 analyses) (4 analyses) (7 analyses)
Maximum sample size 62 62 66
Expected sample size under H0 36 31 30
Expected sample size under H1 40 40 39
Type I error 0.0479 0.0482 0.0484
Power 0.893 0.889 0.894