Choice as a function of local versus molar reinforcement contingencies

Area of Science:

• Behavioral psychology
• Animal learning
• Operant conditioning

Background:

• Organisms learn based on reinforcement schedules.
• Reinforcement can be defined by overall rates (molar) or immediate outcomes (local).
• Understanding how these different schedules interact is crucial for behavioral analysis.

Purpose of the Study:

• To investigate how rats learn under conditions where molar and local reinforcement contingencies oppose each other.
• To determine the relative influence of molar and local contingencies on response patterns.
• To examine the effect of intertrial intervals on the discrimination of reinforcement contingencies.

Main Methods:

• Rats were trained on a discrete-trial probability learning task.
• Experiment 1: Equal molar reinforcement, differential local reinforcement (win-stay, lose-shift).
• Experiment 2: Asymmetrical local reinforcement favoring win-shift/lose-shift, leading to differing molar rates; varied intertrial intervals.

Main Results:

• The win-stay pattern was learned rapidly, suggesting it relies on simple reinforcement effects, not contingency discrimination.
• Control by local contingencies diminished with longer intertrial intervals.
• With long intertrial intervals, molar contingencies determined choice proportion, as evidenced by the generalized matching law.
• When molar and local contingencies opposed each other, both exerted independent control.

Conclusions:

• Immediate reinforcement (local contingencies) is learned more readily than overall reinforcement rates (molar contingencies).
• Both molar and local reinforcement contingencies can independently influence behavior.
• Molar contingency control is not reducible to local contingency control.