Sequential sampling and paradoxes of risky choice

Area of Science:

• Cognitive Psychology
• Decision Science
• Behavioral Economics

Background:

• Key findings in decision-making research, including common-ratio, common-consequence, reflection, and event-splitting effects, challenge expected utility theory.
• These effects serve as a benchmark for evaluating descriptive theories of risky choice.
• Current sequential sampling models, like decision field theory, do not predict these robust violations.

Purpose of the Study:

• To demonstrate that a modified sequential sampling model can explain established decision-making effects.
• To provide a parsimonious and cognitively plausible account for violations of expected utility theory.
• To show that these effects emerge under specific model conditions.

Main Methods:

• Extending the decision field theory (a sequential sampling model) by incorporating stochastic error in event sampling.
• Analyzing the model's predictions against known decision-making effects.
• Evaluating parameter values, including those derived from existing choice data.

Main Results:

• The extended decision field theory successfully predicts common-ratio, common-consequence, reflection, and event-splitting effects.
• The inclusion of stochastic error in event sampling provides a parsimonious explanation for these violations.
• These effects are shown to emerge across a wide range of model parameter values.

Conclusions:

• A minor extension to decision field theory, incorporating stochastic error, offers a cognitively plausible explanation for key decision-making phenomena.
• The modified model provides a benchmark for understanding risky choice behavior beyond expected utility theory.
• This approach enhances the predictive power of sequential sampling models in decision science.