Info

Applying the equations in Section 16, we obtain the log hazard ratio:

Log hazard ratio = 0-486.

From parts 1(a) and 1(b) of this example, already given in the text, we have survt — 0-20 and surv2 = 0-30, giving log (/u) = log(log(0-20)/log(0-30)) = 0-290. Hence asccp = log(/i1)/l-6445 = 0-176, and Np = 4/0-1762 = 129 patients.

Priors: From the summary of equations, we have the following priors, data likelihood and posteriors:

Uninformative prior

= N(0, 4/0)

0 0

Post a comment