powered by
Consider 2 by 2 crossover design. \(\zeta = \delta^2+sT^2+sR^2-thetaBE*max(\sigma_0^2,sR^2)\) . \(sT^2=\sigma_{BT}^2+\sigma_{WT}^2\), \(sR^2=\sigma_{BR}^2+\sigma_{WR}^2\)
Ho: \(\zeta \ge 0 \)
Ha: \(\zeta < 0 \)
Vitro.BE(alpha, beta, delta, sigmaBT, sigmaBR, sigmaWT, sigmaWR, thetaBE)
significance level
power = 1-beta
delta is the mean difference
\(\sigma_{BT}^2\) is the between-subjects variance in test formulation
\(\sigma_{BR}^2\) is the between-subjects variance in reference formulation
\(\sigma_{WT}^2\) is the within-subjects variance in test formulation
\(\sigma_{WR}^2\) is the within-subjects variance in reference formulation
here thetaBE=1
Chow SC, Shao J, Wang H. Sample Size Calculation in Clinical Research. New York: Marcel Dekker, 2003
# NOT RUN { Example.10.5<-Vitro.BE(0.05,0.2,0,0.5,0.5,0.5,0.5,1) Example.10.5 # n=43 Vitro.BE reach 0 # }
Run the code above in your browser using DataLab