[Limdep Nlogit List] different procedures to include interaction terms

[carlesi at agrecon.univpm.it]

Dear Prof. Greene and LIMDEP/NLOGIT users,

I collected data from an online survey. It included a discrete choice
experiment.
I wish to employ a RPL model taking into account interactions between
choice attributes (A,B,C) and socio-economic variables (var1,var2).

I am wondering which is the difference between:

1)
CREATE; A_var1 = A*var1$
CREATE; A_var2 = A*var2$
CREATE; B_var1 = B*var1$
CREATE; B_var2 = B*var2$

RPLOGIT
; Lhs = choice
; CHOICES = car,bus,train
; Rhs = A,B,C,A_var1,A_var2,B_var1,B_var2
; Rh2 = one
; Fcn = A(n),B(n)
; Halton
; Pts = 100$

and

2)
RPLOGIT
; Lhs = choice
; CHOICES = car,bus,train
; Rhs = A,B,C
; Rh2 = one
; RPL = var1,var2
; Fcn = A(n),B(n)
; Halton
; Pts = 100$

Moreover, is it possible to compute the WTP in the second case? Which is
the right procedure to follow?

Thank you in advance for your help,

Lorenzo