Operant conditioning of behavioral variability using a percentile reinforcement schedule

Area of Science:

• Behavioral psychology
• Operant conditioning
• Animal behavior research

Background:

• Understanding the mechanisms that control behavioral variability is crucial in operant conditioning.
• Previous schedules often confound variability requirements with reinforcement probability, limiting analysis.
• A need exists for schedules that independently manipulate these factors.

Purpose of the Study:

• To develop and validate a new percentile reinforcement schedule for studying pattern variability.
• To dissociate the variability requirement from the overall probability of reinforcement.
• To investigate the relationship between reinforcement schedules and the emergence of behavioral variability.

Main Methods:

• A discrete-trials procedure was used with pigeons producing four-peck patterns on two response keys.
• A novel percentile reinforcement schedule adjusted the criterion (N) for response variability.
• Experiment 1 manipulated the percentile requirement, while Experiment 2 manipulated reinforcement probability.

Main Results:

• Increased percentile requirements led to a direct increase in behavioral variability.
• A higher variability requirement correlated with decreased within-group variance.
• Reinforcement probability did not show a consistent relationship with behavioral variability.

Conclusions:

• The developed percentile schedule effectively controls and studies behavioral variability independently.
• Findings support a tentative hypothesis of probability-dependent selection in operant conditioning of variability.
• This schedule offers a valuable tool for future research on the evolution and control of behavioral variation.