st: weighted estimation of the Hopit model in gllamm

[Danilo Cavapozzi <danilo.cavapozzi@unive.it>]

Dear all,

I have a question on the use of weights in gllamm. I am estimating the 
standard Hopit model with anchoring vignettes (as in the example on the 
gllamm website, http://www.gllamm.org/chopit.pdf). Each individual in 
the sample contributes with the self-evalutation and the evalutation of 
vignettes. Then, there are 2 levels: the first one consists of the 
single evalutations (self-evaluations OR vignette evaluations) and the 
second consists of ALL the evaluations provided by a given individual. I 
am not estimating any random effect. I use frequency weights.

Should I apply the same set of frequency weights to both levels by 
specifying two variables (say weight1 and weight2) that are equal? Or 
should I apply the weights to one level only? If the latter is true, 
which level should I consider?

Thanks in advance for any suggestion.

Best,
Danilo

--
Danilo Cavapozzi
Department of Economics, Ca' Foscari University of Venice
Fondamenta San Giobbe, Cannaregio 873
30121 Venezia, Italy
Phone: +39 041 234 9180
Fax: +39 041 234 9176

*
*   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/