A Guide to Parent-Child fNIRS Hyperscanning Data Analysis

Abstract

The use of functional near-infrared spectroscopy (fNIRS) hyperscanning during naturalistic interactions in parent-child dyads has substantially advanced our understanding of the neurobiological underpinnings of human social interaction. However, despite the rise of developmental hyperscanning studies over the last years, analysis procedures have not yet been standardized and are often individually developed by each research team. This article offers a guide on parent-child fNIRS hyperscanning data analysis in MATLAB and R. We provide an exemplary dataset of 20 dyads assessed during a cooperative versus individual problem-solving task, with brain activity measured using 16 channels located over bilateral frontal and temporo-parietal areas. We use MATLAB toolboxes Homer2 and SPM for fNIRS to preprocess the acquired data, and suggest a standardized procedure previously employed in several publications. Next, we calculate interpersonal neural synchrony between dyads using Wavelet Transform Coherence (WTC) and illustrate how to run a random pair analysis to control for spurious correlations in the signal. We then use RStudio to estimate Generalized Linear Mixed Models (GLMM) to account for the bounded distribution of coherence values for interpersonal neural synchrony analyses. With this guide, we hope to offer advice for future parent-child fNIRS hyperscanning investigations and to enhance replicability within the field.