February, 2003
© 2000, 2001, 2003 Scott S. Emerson, M.D., Ph.D.
Session 1- Fixed Sample Design: 122
S+SeqTrial
Re-designing the fixed sample study
Sample size 1700 subjects (850 / arm)
PROBABILITY MODEL and HYPOTHESES:
 Two arm study of binary response variable
 Theta is difference in probabilities (Treatment - Comparison)
 One-sided hypothesis test of a lesser alternative:
                                              
           Null : Theta >=  0           (size  =  0.025 )
    Alternative : Theta <= -0.06019     (power =  0.8   )
 (Fixed sample test)
STOPPING BOUNDARIES: Sample Mean scale
                           a       d
    Time 1 (N= 1700) -0.0421 -0.0421