s-news
[Top] [All Lists]

frailties

To: s-news@wubios.wustl.edu
Subject: frailties
From: "Gordon A. Fox" <gfox@mail.cas.usf.edu>
Date: Fri, 28 Jan 2005 13:41:33 -0500
Organization: University of South Florida
User-agent: Mozilla Thunderbird 1.0 (X11/20041206)
Two related questions on use of frailty terms in survreg and coxph.

1. How do I interpret the values that are stored in a survreg or coxph object under the heading "frail?" They'd appear to be individual estimates of log frailties, but their sample variance doesn't appear to be closely related to the value reported as the variance of the random effect.

2. I want to run a large number of model estimates in a loop, varying the length of the data set. The output for both survreg and coxph includes an estimate of the variance of the random effect, but I don't see (that is, I'm probably missing it) what this value is actually called, so that I can store an entire vector of them from a large series of models. Any help?

Thanks.

Gordon

--
Dr. Gordon A. Fox       Voice: (813)974-7352       Fax: (813)974-3263
Dept. of Biology ((for US mail:)SCA 110) ((for FedEx etc:)BSF 156)
Univ. of South Florida                 4202 E. Fowler Ave.
Tampa, FL 33620, USA                   http://boojum.cas.usf.edu/index.pl/home

"Trying is the first step towards failure." -- Homer Simpson


<Prev in Thread] Current Thread [Next in Thread>
  • frailties, Gordon A. Fox <=