Prospects elimination by risk attributes: A simple decision-aid procedure under uncertainty

Abstract

The traditional paradigm for decision-making under risk and uncertainty is based on a concept of rationality which has often been criticized for its neglect of the human decision processes and its limited descriptive ability. The present discussion notes the need for prescriptively oriented and descriptively justifiable decision-aid methodologies, and it offers a simple elimination procedure based on a multiattribute concept of risk. This procedure aims at narrowing a set of stochastically non dominated decision alternatives first through a partial ordering induced by risk-attribute dominance, and second, by a stepwise elimination which utilizes an importance ordering of risk attributes.