Old and New Roles for Expected and Generalized Utility Theories

Multiattribute Decision Making in Context: A Dynamic Neural Network Methodology

“…Hence, the type of networks we use differ from the back propagation networks, which are primarily a tool for prediction rather than for underlying description of the phenomena. Our theory is descriptive rather than normative or prescriptive (see, e.g., Eppel, Matheson, Miyamoto, Wu, & Eriksen, 1992, for a discussion of this distinction), but we hope that it will ultimately lead to better predictions due to greater understanding of the dynamics of real-life decision makers. We will return to this point in Section 3.4 of the Discussion.…”

“…In the standard distinction between descriptive, prescriptive, and normative decision theories (e.g., Eppel et al 1992), ours is a descriptive theory. We also intend it, though, to lead to prescriptive theories for particular cases.…”

Can we assume that research participants are utility maximisers?

“…In 1989, leading academics working in the field of normative, descriptive and prescriptive utility theory participated in the ''Utility: Theories, Measurement and Applications'' conference [1]. The delegates unanimously agreed that expected utility and subjective expected utility were still the best normative theories for decision making under risk or uncertainty [1,2] but that utility maximisation (the primary assumption/ axiom and the one constant in all utility theories) is currently indefensible as a descriptive decision theory [1]. In the meantime, utility theory has been assumed to be the best approximation available and stated preference techniques have been used in economic evaluation.…”

“…Eppel et al [2], Sarin [3] and Schoemaker [4] have suggested that many of the violations of utility theories when experimentally tested were the result of psychological attributes that were not accounted for in the experiments. Generalised utility theories such as regret theory [5,6] and prospect theory [7] highlighted the issues of regret, subjective distortion of probabilities and utilities, choices and options being judged from a reference point, and framing or context effects leading to completely different choices being made for mathematically equivalent choices.…”