real estimates for closed capture model

[annaren]

I have thoroughly read chapter 15 but I am still having problems with closed capture models.
I was wondering if anybody could explain to me how to interpret the real estimates from a closed capture with heterogeneity model. I have two group and 5 sample occassions. The output I get is :

Parameter Estimate Standard Error
------------------------- -------------- --------------
1:pi 0.2161283 0.0181923
2:pi 0.7567365 0.0396966
3:p 0.7665796 0.0228468
4:p 0.2134030 0.0159723
5:p 0.2283322 0.0348782
6:p 0.7904683 0.0444091
7:N 1732.9404 53.684175
8:N 382.93028 22.408477

I understand that parameter 7: N is pop size for group 1 and parameter 8:N is pop size group 2 but I am not sure on parameters 1-6

Many thanks

[jlaake]

You can't interpret real parameters correctly without knowing the model that was built and how it was built. I can only guess from the output that you fit p(g*het) so you have a pi value for each group and 2 p's for each group. With 2 het classes, pi is the proportion in class 1 and 1-pi is the prorportion in class 2. Each of the p values is a probability for one of the 2 classes in the 2 groups. Which is which is unclear without looking at the PIMS. Likewise you don't know which pi relates to which group without the PIMS.