Fig. 5

The statistical bias based on the fixed Design and design with adaptations. The statistical bias is based on each planned dose group (placebo, 0.03, 3, 10, 20, and 30 mg/kg) under four scenarios of design setting as fixed design, no adaptive (Scenario 2), half adaptive (Scenario 3) and fully adaptive (Scenario 4). The true dose responses follow dose profiles of a: Emax curve; b: flat curve; c: log linear curve and d: U-shaped curve