Estimation of Standard error for annual survival estimate

[skohlmann]

I'm analyzing monthly survival data of radiocollared individuals via the known fate model. My data are collected in monthly intervals, but for modeling (PVA) I'm interested in the annual survival. I suspect the annual survival figure thus is the monthly (Phi)^12 - correct? How do I obtain the standard error and confidence intervals for this annual estimate?

Thanks!
Steve

[gwhite]

Steve:
Check the derived parameter output for the annual estimate (assuming that you structured the encounter history for 12 months). For the known fate model, the derived parameter set is the probability of surviving the entire encounther history period.
Gary

[skohlmann]

Unfortunately, my study went on for 35 months. The total survival probability for the entire study is thus not very interesting. How can I derive the precision of survival probablities for any time period that is not the encounter interval and not the entire study length?

[gwhite]

Steve:
Restructure your encounter histories into 3 intervals (2 of 12 months, 1 of 11 months with no animals in the last interva -- all entries in the encounter history for the last month will be 00, so that no survival rate will be estimated for this interval). Now treat these years as different groups in MARK. With this structure, you will be able to easily examine differences between years, and still get the annual survival rate for each year as a derived parameter. For the 11-month group, you will have to fix the unestimated survival as 1 to get an estimate.
Gary