[Limdep Nlogit List] simultaneous equations
lionel fugon
lionelfugon at hotmail.com
Fri Oct 2 18:43:25 EST 2009
Hello everybody,
That's my first message on Limdep list, so I'm sorry if I don't send it to the good address.
Well, I have just installed Limdep on my computer and I'm a bit lost. I would like to evaluate some simultaneous equations
with Limdep because I haven't succeed with Stata. So, the problem can be written as follows :
Y1 = Y2 + L1
Y2 = Y1 + L2
Y1 is the number of side effects of an HIV treatment (so it's a continu variable which can be considered as a Poisson variable)
Y2 is the cannabis use so it's binary (use or not)
L1 and L2 are a list of explanatory variables.
So, I have few problems with this simultaneous equations : I would like to know if there is a mutual effect between Y1 and Y2 and that's for this reason I use simultaneous equation. The problem is : can I do this model with a Poisson variable or I have to normalize the variable and I use a classic regression with a normal distribution (in Stata I can't use a Poisson variable, I have to normalize my outcome and use the cdsimeq command to realize simultaneous equations with a continu and a binary outcome). And if I can, what is the syntax ?
The second problem is that for the moment it's a cross-sectionnal study but soon, I will have others observations for the same patients because it's a longitudinal study. I read that Limdep can take account of the dependence of incomes with a mixed model in simultaneous equations. Is it right ? Syntax is complicated ?
I am sorry, I have many questions and I have no manual of Limdep, just and old version (8.0) and I haven't used simultaneous equations before so if somebody can help me it would be very nice.
Thank you very much
_________________________________________________________________
Découvrez toutes les possibilités de communication avec vos proches
http://www.microsoft.com/windows/windowslive/default.aspx
More information about the Limdep
mailing list