To predict human choice, consider the context.

Area of Science:

• Cognitive Science
• Decision Science
• Behavioral Economics

Background:

• Established models of choice, such as prospect theory, often exhibit low predictive accuracy in real-world scenarios.
• Predictive accuracy in choice modeling is a critical metric for understanding decision-making processes.
• Previous research has primarily focused on refining existing models rather than re-evaluating core assumptions.

Purpose of the Study:

• To identify the fundamental reasons behind the low predictive accuracy of current choice models.
• To investigate the role of contextual evaluation in human choice behavior.
• To propose a new perspective on understanding cognitive processes underlying decision-making.

Main Methods:

• Analysis of data from choice prediction competitions.
• Comparative evaluation of existing choice models against empirical data.
• Theoretical examination of the assumptions underlying independent vs. contextual evaluation of alternatives.

Main Results:

• Popular choice models, including prospect theory, demonstrate significantly low predictive accuracy.
• The primary limitation identified is the assumption of independent evaluation of alternatives.
• Contextual factors critically influence choice prediction, challenging isolated-evaluation paradigms.

Conclusions:

• Re-evaluating the assumption of independent choice evaluation is crucial for improving predictive models.
• Understanding cognitive processes requires acknowledging the contextual nature of decision-making.
• Future research should focus on context-dependent models to enhance the accuracy of choice prediction.