Random utility models with binary tree decomposable rank orders satisfy Tversky's elimination-by-aspects model

From MaRDI portal
(Redirected from Publication:1111491)





\textit{A. Tversky} [ibid. 9, 341-367 (1972; Zbl 0249.92010)] showed that his elimination-by-aspects model is a random utility model. He constructed the random utility representation by recursively defining a set of ranking probabilities on various sets. In this paper, we demonstrate the converse of this result, namely, that if a random utility model has ranking probabilities that satisfy Tversky's recursive formula, then the associated choice probabilities satisfy Tversky's elimination- by-aspects model. We also illustrate structures of preference and aversion probabilities, each of which satisfy the elimination-by-aspects model and which are compatible with a common random utility model.











This page was built for publication: Random utility models with binary tree decomposable rank orders satisfy Tversky's elimination-by-aspects model

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1111491)