Utility Invariance Despite Labile Preferences

This chapter discusses the robustness of preference reversals as well as the research in the domain of risky decision making that goes beyond simple demonstrations of preference reversals based on a few pairs of gambles. It explains how the entire preference order over a large set of gambles changes with the response mode. It examines preference reversals in a riskless domain in which subjects state their preferences for apartments using choices and ratings of attractiveness. The chapter proposes a two-pronged theory of preference reversals. The attribute judged more important has a greater effect in choices than in ratings. There is also an explanation as to how two pronged theory, which assume that subjects change either strategies or weights across tasks, give a coherent account of many important properties of the data while allowing the elicitation of utilities or psychological values to remain constant across tasks. The chapter illustrates that, with the appropriate models, utilities are stable and have meaning over and beyond the task from which they are derived.