deltamethod SE for nest survival estimate

[mcallen]

Hello,
I have a pretty easy question:
Is this the correct way to calculate SE for (daily nest survival rate)^23 in RMark? ...

Code: Select all

deltamethod(~x1^23,mean=dsr.null.model$results$real$estimate,cov=dsr.null.model$results$real$se^2,ses=T)

...here be the model it is based on:

Code: Select all

dsr.null.model=mark(nest.data,nocc=90,model="Nest",model.parameters=list(S=list(formula=~1)))

Thanks!

Mike
NJ

[jlaake]

That looks correct. For a simple example like this the delta method variance is quite simple and you should be able to check your result by hand. See the Appendix in C&W book. If my head cold isn't preventing me from thinking straight, the se for what you want is (23*dsr^22)*se(dsr). The delta method variance for a single variable is the derivative squared times the variance. The derivative is 23*dsr^22. Since you want the std error it is simply the derivative * std error of dsr.

--jeff

[mcallen]

That yields the same result. Thanks!

Mike