Page 1 of 1

adjusting beta parameter SE estimates for c-hat

PostPosted: Wed Aug 22, 2012 11:30 am
by roche042
Hi Jeff,

I saw that the adjust c-hat value function does not automatically change the SEs associated with the beta estimates for each adjusted model. Is there anyway to do that within RMark?

Thanks in advance!

Erin

Re: adjusting beta parameter SE estimates for c-hat

PostPosted: Wed Aug 22, 2012 11:53 am
by jlaake
Erin and I discussed this offlist and I wanted to have a record of it on phidot. The original se values in the model object are not modified when you use adjust.chat. That only changes the value of chat in the model object. The extractor functions like summary, get.real, etc use the value of c-hat to adjust std errors. If you use summary(model,se=T) you'll get the adjusted values but if you grab the std errors from the results directly then that is what was extracted from MARK and will not be adjusted by chat unless you specified chat in the call to MARK. That is where things could get confusing. Don't use a chat in the call to MARK and then expect the adjust.chat to work as you would expect.

--jeff