I am trying to estimate a two-level nested logit model on schooling choice using the nlogitrum-command.
Problem: Stata refuses to estimate the model and says "too many values r(134);". Strange error message, normally it applies to the tabulate command..
The statement I use:
. nlogitrum chosen ctrav ctrav2 f1ff f1fm f1mm f1mf ACAD VOCL VOCS ACADage VOCLage VOCSage ACADmale VOCLmale VOCSmale,group(indiv) nests (option PROG ATTEND) ivc(OUT=1, ACAD=1, VOCL=1, VOCS=1,NONE=1) clogit;
Remarks:
- the clogit, which is estimated first to obtain the initial values, works fine
- removing constraints on the IV-parameters makes no difference
- !! nlogit (instead of nlogitrum) with the same model works !!
- the choice set contains 563 alternatives: 148 in the ACAD-branch, 92 in the VOCL-branch, 322 in the VOCS-branch and 1 in the NONE-branch.
My guess is that perhaps there is some kind of Stata-limitation on the number of alternatives in the choice set? But then it's still strange that nlogit works...
Any suggestions greatly appreciated...
Stijn Kelchtermans
Doctoral Researcher, K.U.Leuven, Belgium
+32 (0)16 32 69 43
[email protected]
*
* 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/
