[Limdep Nlogit List] Error components model

[Thijs Internet]

Dear all,

 

In estimating error components models NLOGIT imposes a
normal distribution with mean 0 and variance to be
estimated. This process can be mimicked by a standard rpl specification using
beta(n,*,0). I’m interested in an error components model based on a uniform
distribution with mean and domain [-h,h]. However, using beta(u,*,0) sets the
mean equal to the spread such that the domain actually becomes [-2h,0]. Is there a way
to fix the mean of the uniform distribution at 0 in NLOGIT?

 

Thanks for your support!


With kind regards,

 

Thijs Dekker

PhD researcher

 

Institute
for Environmental Studies 

Faculty of Earth and
Life Sciences 

VU
University AmsterdamThe Netherlands





 		 	   		  
_________________________________________________________________
Een netbook met Windows 7? Hier vind je alles dat je moet weten.
www.windows.nl/netbook