Dear stata-listers,
I have a large panel data of censored observations in which residuals
heteroskedasticity is an issue. I have three questions:
1) Is the CLAD estimator the right choice for panel data? In a recent post:
http://datasets.org/statalist/archive/2003-05/msg00033.html it is explained
that CLAD is not consistent for the fixed effect (FC). However, I have no
interest in the FC coefficients and I simply would like to pool the data
together. Could CLAD be a solution in this case?
2) Which is the right procedure to obtain prediction and
confidence/prediction intervals? Presumably, since CLAD do not assume any
distribution for the error term, a bootstrap procedure might be necessary.
Can anybody shed some light on this?
3) Is there any multivariate version of CLAD? Or has anybody come across a
solution proposed in the literature for multivariate censored models, robust
to heteroskedasticity and non-normality?
Thank you very much in advance for your help,
Carlo
**************************************************
Carlo Fezzi
Senior Research Associate
Centre for Social Research on the Global Environment (CSERGE)
Department of Environmental Sciences
University of East Anglia
Norwich (UK) NR2 7TJ
Telephone: +44(0)1603 591385
Fax: +44(0)1603 593739
*************************************************
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
