Group Sequential Design

Dear newsgroup,

recently i wanted to design an group sequential trial with survival as
outcome.
I used PASS to calculate everything i needed. But then i wanted to
really understand things.
First of all i computed O'Brien Fleming stopping boundaries, which i
think is quite easy.
Then i wanted to calculate the needed sample size for the given alpha,
beta and the stopping boundaries. To do so i need the drift parameter
of the brownian motion.
Here lies the problem: I have no idea to calcute this dirft parameter,
however i do know how it is connected to the Hazard Rate and the needed
number of events.

Hope somebody can help me on this topic

Greeting
Christoph int-Veen

.

Relevant Pages

  • Re: Real world modeling.
    ... And these boundaries are typically where we may ... stating that actual domain based high level design means creating system ... readily apparent point. ... When Facing Major High Level Design Problems ...
    (comp.object)
  • Re: Please help!
    ... employee's salary range and corresponding base tax amount and tax ... base tax for that salary range and the product of percentage of excess ... to check the boundaries: one just below each boundary, ... Not run each test case through your design, again by hand, ...
    (comp.programming)
  • Re: Context Diagram?
    ... A context diagram isn't part of UML, but it is intended to graphically ... everything else that interacts with it. ... boundaries between a system being design introduced by structured ... design techniques in the 80's. ...
    (comp.object)
  • Re: Real world modeling.
    ... collaboration from the real DOMAIN, ... And these boundaries are typically where we may ... system design upon ACTUAL entities and dynamics is generally superior to ... Elliott ...
    (comp.object)