When payoffs look like probabilities: Separating form and content in risky choice

Area of Science:

• Decision science
• Cognitive psychology
• Behavioral economics

Background:

• Behavioral models of risky choice often assume specific psychophysical transformations for outcomes and probabilities.
• Prospect theory, a prominent model, utilizes a value function and a probability weighting function.
• Traditional experiments often confound the numerical form with the content of outcomes and probabilities.

Purpose of the Study:

• To investigate the impact of unconfounding numerical representations of outcomes and probabilities on psychophysical functions in risky choice.
• To explore how altering the representation of probability and outcomes affects decision-making models.
• To challenge the assumptions of traditional prospect theory by examining the role of numerical form.

Main Methods:

• Conducted experiments that interchanged the numerical representations of outcomes and probabilities.
• Utilized a probability-like representation for outcomes and an outcome-like representation for probability.
• Analyzed the resulting psychophysical functions and observed framing effects.

Main Results:

• A proportional representation of outcomes led to an inverse S-shaped value function for gains, differing from traditional models.
• This alternative value function produced novel framing effects, common ratio effects for bounded gains, and a reduced impact of gain-loss framing.
• An absolute representation of probability decreased sensitivity to high probabilities.

Conclusions:

• The psychophysics of risky choice are significantly constructive and influenced by numerical representation.
• Traditional models may overemphasize subjective reactions to numerical form rather than substantive content.
• Findings suggest a need to reconsider how numerical representations impact behavioral models of decision-making.