[Limdep Nlogit List] Attribute-specific scale?

[Gonzalez, Juan]

Dear Limdepers,

 

I am using NLOGIT to estimate an RPLOGIT model for a conjoint survey and
have a question about the ability to fix the "scale parameter" in this
model. Does NLOGIT assume the scale is equal to 1 even when we estimate
random parameters? Or does it use this random parameter as an
attribute-specific scale? From the text in the guide it seems the latter
is true. 

 

Thanks,

 

Juan Marcos Gonzalez