Expected utility theory with probability grids and preference formation (Q2206005)

The author deals with expected utility theory from the viewpoints of bounded rationality and preference modeling. Permissible probabilities are restricted to decimal fractions up to a given cognitive bound \(\rho\). The presented theory is constructive from the viewpoint of decision makers. According to author's opinion, bound \(\rho = 2\) is sufficient for the needs of ordinary decision makers. The main part of the paper is a study of preferences including of incomparabilities in cases of finite cognitive bounds. Representation theorems are given in terms of two-dimensional vector-valued utility functions. Possible extensions or generalizations of the presented results are discussed in the conclusions of the paper.