Simulating Storytime – Neural Dynamics of Parent–Child Brain Synchrony

Abstract

“Simulating Storytime” aims to develop a computational model of parent–child brain synchrony during shared reading. Empirical studies using EEG hyperscanning suggest that reading printed books fosters stronger inter-brain alignment than screen-based reading, likely due to sustained joint attention and interaction. However, collecting such dual-brain data in real-life settings is methodologically complex and resource-intensive. This project proposes an *in silico* approach: simulating two coupled neural systems (“parent” and “child”) exposed to a shared narrative input. By systematically manipulating interaction strength and distraction, the model will visualize and quantify how different reading environments influence brain-to-brain connectivity. All code and outputs will be open-source, making developmental neuroscience accessible through interactive simulation and data visualization.

Hypothesis

We hypothesize that simulated brain signals driven by a shared rhythmic stimulus (representing a common story) and mutual feedback (representing interaction) will exhibit significantly higher synchrony than signals driven by independent or noisy inputs (representing distractions). Synchrony will be operationalized using Pearson correlation and phase-locking values. Specifically, we expect a clear “synchrony index” to emerge in the joint reading condition, which will decrease when high-frequency noise or uncorrelated inputs are introduced to simulate screen-based or distracted reading.

Method and Workflow

The simulation follows a three-stage pipeline.
First, we generate two independent oscillatory signals representing baseline parent and child brain activity, using EEG-like frequency components (e.g. theta and alpha bands).
Second, we introduce a shared external driver: the amplitude envelope of an audiobook recording, which acts as a common rhythmic stimulus modulating both signals. In the interactive condition, additional bidirectional coupling is added to simulate conversational feedback.
Third, we compute synchrony metrics across different experimental conditions, including joint reading, passive co-presence, and distracted reading.

Tools and Implementation

All simulations will be implemented in Python using standard scientific libraries. NumPy and SciPy will be used for signal generation and filtering, while Matplotlib will be used for visualizations such as time-series plots and correlation heatmaps. The shared story input will be derived from a real audiobook. For validation, the analysis pipeline can be compared conceptually to open dual-EEG datasets, such as the Leong & Wass (2018) mother–infant hyperscanning dataset.

Significance

This project provides a conceptual sandbox for exploring why shared reading is neurobiologically different from solitary or screen-based media use. While the model is necessarily simplified, it allows users to test mechanistic hypotheses about joint attention, interaction, and distraction in a transparent and reproducible way. The simulation translates abstract findings from developmental neuroscience into an intuitive, interactive tool aligned with the open-science and interdisciplinary ethos of Brainhack. Ultimately, it demonstrates how computational modeling can complement empirical research by offering insight into complex social brain dynamics that are otherwise difficult to observe directly.