List,
My model:
survfit(Surv(sur, state) ~ trt+stk+tend, data=miasurv)
My output is appropriate, I can successfully run the analysis with any
combination of the three factors. I can also successfully compare the
resultant survival curves with:
survdiff(Surv(sur, state) ~ trt+stk+tend, data=miasurv)
And I realise that I can compare combinations of individual curves with the
'subset' argument to survdiff. However, I am wondering:
1) Is there a multiple-comparisons approach for survival curves to indicate
which factor combinations are different and/or similar?
2) Is it safe and/or appropriate to partition the data by heuristically
logical factor combinations and run survdiff as many times as required to
find the dissimilar groupings of survival responses?
3) Does the nesting of the factors, tend within stk and stk within trt,
impose restrictions on how I must approach the survival analysis?
Thank you all, DaveT.
**********************************************************
nuthin but now.
there it is again... there it is again... there's another one
**********************************************************
David J. Thompson
Silviculture Data Analyst
Ontario Forest Research Institute
Ontario Ministry of Natural Resources
1235 Queen Street East
Sault Ste. Marie, Ontario, P6A 2E5
(705) 946-7433
(705) 946-2030 Fax
david.thompson@mnr.gov.on.ca
**********************************************************
|