Extending a misallocation model to children's choice behavior

Area of Science:

• Behavioral Psychology
• Cognitive Science
• Developmental Psychology

Background:

• Reinforcers guide behavior by signaling future event likelihoods, though this control is often imperfect.
• Imperfect control may stem from misdetecting event relations, suggesting complex cognitive processes in learning.

Purpose of the Study:

• To test a quantitative model explaining imperfect operant control in children.
• To investigate if reinforcer control mechanisms in children mirror those in nonhuman animals.

Main Methods:

• Children participated in a game involving two responses, one yielding a reinforcer.
• Reinforcer probabilities (.1, .5, .9) for consistent responses varied across conditions.
• Behavioral choices were analyzed against a quantitative model.

Main Results:

• Children's choices closely, but not perfectly, matched signaled probabilities, indicating strong reinforcer control.
• A quantitative model accurately described children's choice behavior.
• The model also fit nonhuman animal data, suggesting cross-species behavioral mechanisms.

Conclusions:

• Reinforcer control is proportional to the detectability of event relations.
• Simple operant learning may involve more complex psychological processes than previously assumed.
• Evidence suggests fundamental similarities in operant control mechanisms between children and nonhuman animals.