Infants and Mobiles: Developing an Understanding of Cause and Effect

Area of Science:

• Cognitive Science
• Developmental Psychology
• Computational Neuroscience

Background:

• Infants in the mobile conjugate reinforcement paradigm learn to control mobiles by moving their legs.
• This behavior is often seen as evidence of infants' understanding of causality.
• Alternative explanations suggest simple reinforcement learning models may suffice.

Purpose of the Study:

• To propose and evaluate computational models that capture infant behavior in the mobile paradigm.
• To investigate whether a causal model or a reinforcement model better explains infant learning and the extinction burst phenomenon.
• To develop an active-learning causal model capable of discovering cause-effect relationships dynamically.

Main Methods:

• Development of an active-learning causal model that identifies cause-effect relationships without prior specification.
• Integration of an active-learning mechanism based on expectation violation with causal and reinforcement models.
• Simulation of infant behavior under different model conditions, including the mobile paradigm and extinction burst scenarios.

Main Results:

• The proposed causality-driven model successfully simulates infant behavior in the mobile paradigm.
• Alternative models, such as a naive reinforcement model, failed to fully account for the observed behaviors, particularly the extinction burst.
• The active-learning mechanism based on expectation violation effectively modeled the extinction burst and demonstrated robustness.

Conclusions:

• A causality-driven active-learning model provides a robust explanation for infant behavior in the mobile paradigm.
• The expectation violation mechanism unifies causal learning and hypothesis testing, offering a coherent framework for understanding infant cognitive development.
• This research sheds light on the mechanisms underlying infants' developing understanding of cause and effect relationships.