Hello,
I am conducting organizational survival analyses with a data set containing
repeated annual observations for each organization during its existence. In
the models I also include external factors measured at the state and county
level. Given the repeated observations it makes sense to share frailty by id,
but the organizations likely also share unmeasured geographic environmental
conditions contributing to their differing survival probabilities. Are there
Stata survival analysis techniques to handle the multi-level nature of the
analyses so that I can effectively address the shared organizational level
frailty and, say, shared county level frailty?
Additionally, in preliminary examinations, significant unobserved heterogeneity
appears to exist when running shared frailty models either at the level of the
org id or the county. However, for one type of organization the best fitting
model (via AIC) appears to be the one using the id shared frailty, while for
another it is the one with county level shared frailty. Are these apparent
differences informative such that organizational frailty is more dominant in
one organizational population than the other? Or am I reading too much into
the log likelihoods?
Many thanks,
David
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
