Skip to main content
The text for this item is currently unavailable.
Testing Among Classes of Decision Models by Evaluating Critical Properties
No data is associated with this publication.
Abstract
Sometimes it is useful to assess models by comparing their accuracy in fitting data. This approach can, however, be misleading when models are specified with assumptions that are not required by the model. I will describe case histories of testing models such as Cumulative Prospect Theory, Regret Theory, and Lexicographic Semiorder models (including the priority heuristic) by testing critical properties that can be deduced as theorems of one model and which should be violated according to another model. Model fitting is used to calculate where to search for predicted violations under one model that refute another model.
The text for this item is currently unavailable.