Context reinforcement degrades discriminative control: a memory approach

Area of Science:

• Animal behavior
• Cognitive psychology
• Behavioral neuroscience

Background:

• Understanding how reinforcement schedules impact learning is crucial in behavioral science.
• Context reinforcement, where reinforcement is tied to specific environmental cues, plays a role in modulating behavior.
• Previous research has explored various factors influencing discriminative learning in animals.

Purpose of the Study:

• To investigate the impact of varying reinforcement densities on visual discrimination in pigeons.
• To examine how reinforcement of sample stimulus pecks and noncontingent reinforcement affect discriminative sensitivity and stimulus control.
• To compare the effectiveness of different reinforcement conditions on learning and accuracy.

Main Methods:

• Five pigeons were trained on a visual discrimination task involving different light wavelengths.
• Psychometric functions were used to quantify discriminative sensitivity (d') and stimulus control.
• Experimental conditions included varying percentages of reinforcement for correct choices (5% vs. 20%), reinforcement of sample stimulus pecks, and noncontingent reinforcement during intertrial intervals.

Main Results:

• Higher reinforcement density (20%) led to significantly better discriminative sensitivity and stimulus control compared to lower density (5%).
• Reinforcement of sample stimulus pecks substantially reduced stimulus control and slightly decreased sensitivity.
• Noncontingent reinforcement, whether signaled or not, also reduced stimulus control.

Conclusions:

• Reinforcement density is a critical factor in visual discrimination learning.
• Contingent reinforcement schedules are more effective than noncontingent or sample-based reinforcement for maintaining high levels of discriminative performance.
• The findings support memory-based models of discrimination, suggesting that reinforcement history influences current stimulus control.