Hi, listers
When fitting a parametric hazard model with frailty, the output from
stata includes a modified LR-test of the nullhypothesis that the frailty
component theta equals zero. This as suggested by Gutierrez et
al.(2001).
Example:
Likelihood-ratio test of theta=0: chibar2(01) = 17.62 Prob>=chibar2 =
0.000
However, when adding a cluster component to the model, this output is
lacking. An alternative to clustering would be fitting a shared frailty
model, but this does not seem to be supported with gamma regression in
Stata yet.
My question is then, is there a way to get Stata to supply the modified
LR-test for clustered models so I can evaluate the significance of the
theta parameter when the subjects in the dataset are not independent?
Hope you can help.
Steinar Fossedal
*
* 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/
