Incentive theory: II. Models for choice

Area of Science:

• Behavioral psychology
• Reinforcement learning

Background:

• Concurrent chained schedules of reinforcement are complex behavioral paradigms.
• Existing models may not fully capture the nuances of choice behavior under these schedules.

Purpose of the Study:

• To extend incentive theory to account for concurrent chained schedules of reinforcement.
• To develop a quantitative model integrating primary and secondary reinforcer effects.

Main Methods:

• Formulated a basic model with additive contributions from primary and secondary reinforcer effects.
• Incorporated parameters for the rate of reinforcement, delay of reinforcement gradient (q), and bias (b).
• Tested the model against empirical data from studies varying terminal-link and initial-link schedules.

Main Results:

• The model accurately described results from studies manipulating terminal-link schedules.
• The model explained a significant portion of results from studies manipulating initial-link schedules.
• The extended model successfully accounted for modifications in terminal links, including amount, signaling, and segmentation.

Conclusions:

• Incentive theory provides an accurate and integrated framework for understanding choice behavior.
• The extended model successfully predicts behavior under concurrent chained schedules and their variations.