Rational adaptation under task and processing constraints: implications for testing theories of cognition and action

Area of Science:

• Cognitive Science
• Computational Neuroscience
• Psychology

Background:

• Existing cognitive theories offer computational explanations for flexible human behavior but possess high degrees of freedom.
• Modeling cognition often involves fitting parameters to data, which can limit predictive power.

Purpose of the Study:

• To introduce a new approach, cognitively bounded rational analysis, for modeling and understanding cognition.
• To enhance the predictive acuity of integrated theories of cognition and action.
• To narrow the space of predicted behaviors by analyzing strategy payoffs.

Main Methods:

• Developed cognitively bounded rational analysis, focusing on rational adaptation to utility functions under cognitive and environmental constraints.
• Reanalyzed existing data on psychological refractory period (PRP) dual-task performance.
• Developed and analyzed a new theory of ordered dual-task responses.

Main Results:

• Demonstrated that analyzing strategy payoffs, rather than fitting parameters, sharpens predictive accuracy.
• Provided a novel understanding of strategic variation in PRP dual-task accounts.
• Developed the first predictive, quantitative model of ordered dual-task phenomena from rational cognitive control.

Conclusions:

• Cognitively bounded rational analysis offers a powerful framework for advancing computational cognitive theories.
• This approach effectively models complex behaviors like dual-task performance by considering system constraints.
• The findings highlight the importance of rational control in emergent cognitive phenomena.