# Toward Dynamical Modeling of Infants' Looking Times

## TL;DR

The paper introduces a new dynamical model to study infants' gaze behavior, linking gaze patterns to cognitive processes like attention and memory.

## Contribution

A novel process-based dynamical framework for modeling infant gaze behavior with cognitive process interpretations.

## Key findings

- The model reproduces inter-individual differences in infant gaze behavior.
- Model parameters map onto cognitive processes such as attention and working memory.
- The framework can generate hypotheses about mechanisms underlying developmental changes.

## Abstract

Analyzing looking times is among the most important behavioral approaches to studying problems such as infant cognition, perception, or language development. However, process‐based approaches to the dynamics of infants' looking times are lacking. Here, we propose a new dynamical framework for modeling infant gaze behavior with full account of the microstructure (i.e., saccades and fixations). Our process‐based model is illustrated by reproducing inter‐individual differences in a developmental study of noun comprehension (Garrison et al. 2020). In our modeling framework, numerical values of model parameters map onto specific cognitive processes (e.g., attention or working memory) involved in gaze control. Because of the general architecture of the mathematical model and our robust procedures in model inference via Bayesian data assimilation, our framework may find applications in other fields of developmental and cognitive sciences.

A dynamical model reproduces gaze trajectories in infants during looking time experiments, explains inter‐individual differences via parameter variations, and generates hypotheses on mechanisms underlying developmental change.