[Limdep Nlogit List] continuous or categorical price parameter should be used to estimate WTP

[Rady Talaat]

Hello,
I am using choice experiments to estimate the willingness to pay to access reef sites in a marine park. Unlabeled experiments were used for choice treatment combinations. Each choice set containing the current situation and an alternative option. My first question is: is the constant important for such binary unlabeled experiments? How do we interpret the sign of the constant?
In addition models with both continuous (fees attribute) and categorically recoded attributes were investigated to allow for predictions over the range of attribute-level labels for values not observed within the data. This recoding does not impact upon the overall model performance, however, the constant and fees parameter have new values. Which fees parameter should be used to estimate WTP (the continuous or the categorical one)?
Cheers,
Rady
_________________________________________________________________
Drag n’ drop—Get easy photo sharing with Windows Live™ Photos.

http://www.microsoft.com/windows/windowslive/photos.aspx