Multi-attribute search with general value function and jointly distributed attributes

Abstract

A usual assumption in the multi-attribute search literature is independent distribution of attributes. This assumption is necessary for the tractability of the model. However, in reality many economic search problems show some degree of correlation among the attributes, i.e., by searching an attribute the decision maker not only resolves the uncertainty with regards to that attribute but also updates her beliefs about the distribution of other values. Furthermore, unlike it is usually assumed the decision maker's utility function does not need to be linear in all attributes, in fact, in most cases it is sensible to consider other functional forms. The search order is thus decided based on the marginal utility and informativeness of each attribute. In this thesis, I introduce a simpli ed search model where the decision maker seeks to choose between two objects which are described by their two attributes. These attributes are assumed to be jointly distributed according to a distribution function known by the decision maker. Moreover, I will try to analyze the optimal search and stopping rule.