Paradoxes and mechanisms for choice under risk

Abstract

Experiments on choice under risk typically involve multiple decisions by individual subjects. The choice of mechanism for selecting decision(s) for payoff is an essential design feature unless subjects isolate each one of the multiple decisions. We report treatments with different payoff mechanisms but the same decision tasks. The data show large differences across mechanisms in subjects' revealed risk preferences, a clear violation of isolation. We illustrate the importance of these mechanism effects by identifying their implications for classical tests of theories of decision under risk. We discuss theoretical properties of commonly used mechanisms, and new mechanisms introduced herein, in order to clarify which mechanisms are theoretically incentive compatible for which theories. We identify behavioral properties of some mechanisms that can introduce bias in elicited risk preferences-from cross-task contaminationeven when the mechanism used is theoretically incentive compatible. We explain that selection of a payoff mechanism is an important component of experimental design in many topic areas including social preferences, public goods, bargaining, and choice under uncertainty and ambiguity as well as experiments on decisions under risk.

Temporal Stability, Cross-Validity, and External Validity of Risk Preferences Measures: Experimental Evidence from a UK Representative Sample

“…Moreover, we faced a trade-off between the simultaneous needs of making the monetary outcomes salient, and diversifying the range of monetary stakes in the attempt to capture possible cases of risk preference depending on the absolute value of the monetary stakes (for instance, being risk averse for low amounts, but risk loving for high amounts, or the other way around: Hey and Orme, 1994;Harrison et al, 2005a;Holt and Laury, 2005;Cox and Sadiraj, 2006;Cox et al, 2015). Due to the binding time constraint, we opted for two MPL series of HL questions with the following prize sets: [A1: £40 and £32; B1: £77 and £2], [A2: £100 and £40; B2: £180 and £2].…”

“…These choices were thus increasing in the variance of the outcomes and in the risk they represented, with 7 See for instance Conlisk (1989);Starmer and Sugden (1991);Wilcox (1993);Bernasconi (1994); Beattie and Loomes (1997); Cubitt et al (1998); Harrison et al (2002); Drehmann et al (2005); Harrison et al (2007b); Harrison and Rutström (2008). See Baltussen et al (2012), Harrison and Swarthout (2014), Cox et al (2015), Harrison et al (2015b), andMarch et al (2016) for some recent studies looking more closely at the random lottery incentive mechanism.…”

An Empirical Review of the Connection Between Model Viewer Characteristics and the Comprehension of Conceptual Process Models

“…These results call for more research to investigate these diverging patterns. Potentially additional factors will have to be included, such as cognitive abilities (Recker et al 2014), limitations of sight (Permvattana et al 2013) or risk aversion (Cox et al 2014). What becomes clear though is the fact that the discussion of the student versus practitioner dichotomy only touches the problem at the surface.…”

Small stakes risk aversion in the laboratory: A reconsideration

“…9 We opted for using the Random Lottery Incentive Method (RLIM), where one of the 100 choices was chosen at random for playing out and payment. We did so for two reasons, recognizing that this is not as innocent a procedure as some maintain (see Harrison and Swarthout [2014], Cox, Sadiraj and Schmidt [2015] and Brown and Healy [2016] for evidence and extensive literature discussion). The first reason was to ensure that we collected choices over a wide enough array of lotteries to be able to test the premiss using withinsubject data.…”