st: rologit

[apereza@uvigo.es]

Dear Statalist users,

Following Allison, P.D. & Christakis, N.A. (1994), "Logit Models for  
Sets of Ranked Items", Sociological Methodology, 24, 199-228, we model  
the case where "J", the number of alternatives, can differ across  
respondents. From the theoretical point of view, this is not  
problematic, but to implement it on STATA is not straightforward.

We ask each respondant to name the 6 most dangerous alternatives from  
an initial list of 20 (we clarify that it was possible to add  
alternatives to the list). By doing this, we do not have ties in our  
sample. However, we ended up with 32 different alternatives. When  
using the rologit command of STATA 11 I got the message:

there are 32 alternatives in invasive; the maximum number of  
alternatives is 20
r(149);

My question is twofold:
1) Is there a way to increase the maximum number of alternatives from  
20 to 32?
2) Is it possible to accommodate "rologit" to the case where "J" is  
not fixed among respondants?

Thanks in advance for your help,
Alicia

*
*   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/