Beyond Dyadic Coupling: The Method of Multivariate Surrogate Synchrony (mv-SUSY)

Area of Science:

• Social Neuroscience
• Computational Social Science
• Behavioral Data Analysis

Background:

• Interpersonal synchrony analysis is key to understanding social interaction complexity.
• Current methods are largely limited to dyadic (two-person) interactions.
• A need exists for methods analyzing synchrony in larger groups and complex data.

Purpose of the Study:

• To introduce and evaluate multivariate Surrogate Synchrony (mv-SUSY) as a novel computational method.
• To extend the measurement of interpersonal synchrony beyond dyads to multivariate group settings.
• To demonstrate the feasibility and validity of mv-SUSY for analyzing complex social dynamics.

Main Methods:

• mv-SUSY was applied to eight datasets, including simulated time series (white noise, non-stationarity, autocorrelation, oscillation, multivariate correlation) and empirical movement data.
• Simulated datasets allowed for controlled testing of mv-SUSY's sensitivity to different data characteristics.
• Empirical datasets included dyadic interactions and group discussions to assess real-world applicability.

Main Results:

• mv-SUSY correctly identified the absence of synchrony in simulated datasets 1-4 and its presence in dataset 5 (multivariate correlation).
• Significant movement synchrony was detected in empirical datasets involving dyadic and group interactions.
• Results from mv-SUSY were largely consistent with established dyadic synchrony methods (SUSY, SUCO).

Conclusions:

• The novel mv-SUSY method is feasible and valid for estimating multivariate nonverbal synchrony.
• mv-SUSY expands the toolkit for analyzing complex social interactions in groups.
• This approach offers new possibilities for quantifying synchrony within and between individuals in diverse settings.