Conditioned reinforcement and information theory reconsidered.

Area of Science:

• Behavioral Psychology
• Learning Theory
• Information Theory

Background:

• Conditioned reinforcers historically linked to information about primary reinforcers.
• Previous information theory applications failed to adequately explain observing behavior.
• Modern learning theories often overlook information-based explanations for conditioned reinforcement.

Purpose of the Study:

• To re-evaluate information theory's role in conditioned reinforcement using recent advances.
• To demonstrate how information theory can explain previously problematic 'observing behavior'.
• To integrate Pavlovian conditioning principles with information-theoretic approaches to reinforcement.

Main Methods:

• Applied information theory to Pavlovian conditioning, focusing on uncertainty reduction.
• Calculated information based on reductions in uncertainty about the time to primary reinforcement.
• Compared predictions with Delay Reduction Theory in observing-response and concurrent-chains procedures.

Main Results:

• The novel information-theoretic approach successfully remedies failures in explaining observing behavior.
• The framework yields predictions consistent with Delay Reduction Theory.
• Demonstrates a formal link between information conveyed by conditioned reinforcers and their reinforcing power.

Conclusions:

• Recent advances in information theory provide a viable framework for understanding conditioned reinforcement.
• The approach offers an analytical explanation for the success of Delay Reduction Theory.
• Suggests conditioned reinforcement is fundamentally rooted in Pavlovian conditioning processes.