Contingency spaces and measures in classical and instrumental conditioning

Area of Science:

• Behavioral psychology
• Learning theory
• Quantitative analysis

Background:

• Classical conditioning involves stimulus-response associations.
• Instrumental conditioning involves response-consequence associations.
• Existing measures of association may not fully capture these relationships.

Purpose of the Study:

• To analyze contingency in classical and instrumental conditioning paradigms.
• To propose a new measure for classical conditioning effects.
• To re-evaluate measures of association in instrumental conditioning.

Main Methods:

• Analysis of contingency between stimuli and responses (classical conditioning).
• Analysis of contingency between responses and consequences (instrumental conditioning).
• Application of statistical measures, including root mean square contingency (Ø).

Main Results:

• Root mean square contingency (Ø) is proposed for classical conditioning.
• Traditional measures are inappropriate for instrumental conditioning due to response probability.
• The "Matching Law" and "Probability Matching" are shown to be equivalent to zero contingency.

Conclusions:

• A unified understanding of contingency across conditioning types is advanced.
• The proposed measure (Ø) offers a novel approach for classical conditioning.
• The equivalence of "Matching Law" and "Probability Matching" to zero contingency has significant implications for learning theory.