mean value of survival using variance components approach

[rems]

Dear list:

The final example of chapter 6 Cooch and White manual states that the adequate approach to estimate a mean value of survival when you have a model with phi(t) is to employ a variance components approach based on a random intercept model, rather than using a phi(.) model or modifying the design matrix, since variance component approach incorporates the sampling and process variance.

I was wondering if any of you would be so kind to explain how to do the procedure in MARK since apparently the specification of this method has not been incorporated yet in the manual but promises to be (e.g. much more on this later on).

Kind Regards,

Roberto

--------------------------------------
Dr. Roberto Munguía-Steyer
University of Sao Paulo, SP, Brazil.

[cooch]

Dear list:

The final example of chapter 6 Cooch and White manual states that the adequate approach to estimate a mean value of survival when you have a model with phi(t) is to employ a variance components approach based on a random intercept model, rather than using a phi(.) model or modifying the design matrix, since variance component approach incorporates the sampling and process variance.

I was wondering if any of you would be so kind to explain how to do the procedure in MARK since apparently the specification of this method has not been incorporated yet in the manual but promises to be (e.g. much more on this later on).

Kind Regards,

Roberto

--------------------------------------
Dr. Roberto Munguía-Steyer
University of Sao Paulo, SP, Brazil.

Actually, an 'intercept only' model, but it amounts to much the same thing. See same section of current version of Chapter 6.

[rems]

Hi Evan:

Thanks a lot for your replay and your continuous updating of the manual, it is a great source of information. I was reading the chapter 6 version that appeared a month ago (29/07/09) and now I find that today's version (31/08/09) incorpotates how to do the estimation with the variance component approach.


Regards,

Roberto

[cooch]

Hi Evan:

Thanks a lot for your replay and your continuous updating of the manual, it is a great source of information. I was reading the chapter 6 version that appeared a month ago (29/07/09) and now I find that today's version (31/08/09) incorpotates how to do the estimation with the variance component approach.


Regards,

Roberto

I just uploaded a slightly revised version. I've been hesitant putting this material online until a more complete treatment of random effects (and related things) is finished. But, the question of 'estimating the mean and process variance' comes up sufficiently often as a special case of the larger issue, that I decided in the end to add the material.