Computational model of selection by consequences: patterns of preference change on concurrent schedules

Area of Science:

• Behavioral science
• Computational modeling
• Evolutionary psychology

Background:

• The selection by consequences principle explains adaptive behavior.
• Computational models offer a method to test behavioral theories.

Purpose of the Study:

• To assess a computational model of selection by consequences.
• To investigate preference change phenomena in dynamic reinforcement environments.

Main Methods:

• Utilized a virtual organism based on a genetic algorithm.
• Tested the model with single alternative and concurrent schedules.
• Analyzed molar and molecular effects of behavioral adjustment.

Main Results:

• The computational model generated behavior quantitatively equivalent to live organisms.
• Molar and molecular effects of behavioral adjustment were consistent with observed phenomena.
• The model exhibited preference change phenomena in rapidly changing reinforcement rate ratios.

Conclusions:

• The computational model provides a robust account of adaptive behavior.
• Results strongly support the selectionist account of behavior.
• The model serves as a valuable tool for studying behavioral adaptation.