What information do the error terms hide in Choice Modelling environmental valuation studies?

Choice modelling (CM) is a developing non-market valuation technique that provides a rich data set to the analyst. In addition to parameter estimates for the influence of attributes, labels and respondent characteristics, models also provide information about, (and are sensitive to), the error terms implicit in model parameters. The error terms provide information about the appropriate model form to be analysed in at least three important areas. First, specifications of nested logit models, which are useful for improving model validity, are based on allowing correlation to occur within (but not between) the error terms for choices grouped together in nests. Second, the error terms for separate, but similar CM experiments can be used to generate ratios of the scale parameters that are confounded with individual model parameters. This enables the results from different models to be compared, and identifies choice models that have larger amounts of inherent variability. Third, by identifying the error terms associated with successive choices in CM experiments, it is possible to search for learning and fatigue effects that might be displayed by respondents. A series of these disaggregation exercises have been performed on CM experiments dealing with rainforest valuation to determine where learning and fatigue effects might be present.


Issue Date:
1999
Publication Type:
Conference Paper/ Presentation
PURL Identifier:
http://purl.umn.edu/124543
Total Pages:
22




 Record created 2017-04-01, last modified 2017-08-26

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)