Comparison of probabilistic choice models in humans

Area of Science:

• Neuroeconomics
• Psychopharmacology
• Decision Science

Background:

• Probabilistic choice is a key area of study in neuroeconomics and psychopharmacology.
• Existing parametric models include the entropy model, Prelec's probability weight function, and hyperbola-like functions.

Purpose of the Study:

• To evaluate the fit of different probabilistic choice models to behavioral data.
• To investigate the relationship between model parameters and psychological processes like aversion to potential loss and unpredictability.

Main Methods:

• Estimated parameters and AICc (Akaike Information Criterion with small sample correction) for probabilistic choice models.
• Assessed points of subjective equality across seven probability values (95%-5%).
• Examined model fit and the relationship between parameters and aversion to non-gain.

Main Results:

• The entropy model showed the best fit for group data, while Prelec's function best fit individual data.
• Aversion to potential non-gain and aversion to unpredictability were identified as distinct psychological processes.
• Model parameters were successfully estimated and analyzed in relation to psychological factors.

Conclusions:

• The entropy and Prelec models are suitable for studying risky decision-making in psychopharmacological and neuroeconomic research.
• Findings contribute to understanding the psychological underpinnings of probabilistic choice.
• The study validates the utility of specific parametric models in decision science.