Dear All,
Please forgive my ignorance if this question is nonsense.
I'm trying to fit a GEE model to a panel data collected at 3 points in time. My outcome variable is continuous and I'm considering the family of distribution as gamma with log link and unstructured correlation structure.
Although the help menu for postestimation for xtgee tells me that the predict command can be used to predict residuals, stata gives me an error:
.predict r, residuals
option residuals not allowed
Does stata allow residuals for xtgee?
Also, is there any goodness of fit statistics for xtgee: AIC or BIC ?. How would one check whether the model fits the data well? (other than comparing the estimated correlation matrix to the actual and looking at the predicted values )
I'm using version 8.2
Any comments or suggestions would be appreciated.
Leny
*
* 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/
