ASM 17.1, 17.2

[Old_Timer]

A couple of questions:

ASM is asking for "fitted" survival function. Does this (always) refer to the Note's baseline distribution and not the Cox hazard fcn?

Will the fitted survival function depend on how you set up z? I am coming up with a different answer since I set z = 0 for regular light bulbs in 17.1, S(600) = exp(-1).

For 17.2, if the Q asked for an estimate of S(7) from the Cox model, would the answer be exp(14/11)?

Thanks for your help!

[Abraham Weishaus]

"Baseline" is not well defined--you are free to define z as you please; you can certainly reverse my definition. Your final answer to the question is independent of how you defined z.

I'm not sure what your second question means, since the question does ask for an estimate from a Cox model. If you chose not to assume an exponential baseline but instead used the Breslow approximation of the baseline, you would get a different answer (and would need a computer to solve the problem).

[Old_Timer]

Thanks!