Toward Dynamical Modeling of Infants' Looking Times

Area of Science:

• Developmental Psychology
• Cognitive Science
• Computational Neuroscience

Background:

• Analyzing infant looking times is crucial for understanding cognitive development.
• Existing methods lack process-based dynamical models for infant gaze behavior.
• Microstructure of gaze (saccades, fixations) is often overlooked in dynamic analyses.

Purpose of the Study:

• Propose a novel dynamical framework for modeling infant gaze behavior.
• Incorporate the microstructure of gaze (saccades and fixations) into a process-based model.
• Provide a method to map model parameters to specific cognitive processes.

Main Methods:

• Developed a new dynamical framework for infant gaze analysis.
• Integrated saccades and fixations into a process-based model.
• Utilized Bayesian data assimilation for robust model inference.

Main Results:

• The proposed model successfully reproduced inter-individual differences in a developmental study of noun comprehension.
• Model parameters were shown to map onto cognitive processes like attention and working memory.
• Demonstrated the framework's ability to capture the dynamics of infant looking times.

Conclusions:

• The dynamical framework offers a novel approach to studying infant gaze behavior.
• This model provides insights into cognitive processes underlying gaze control in infants.
• The framework has potential applications across developmental and cognitive sciences.