Abstract
Physiological synchrony is the phenomenon of correspondence of physiological processes in interacting individuals. Research into physiological synchrony has the potential to offer a deeper insight into interactive processes across organisms. It is often investigated in the autonomic nervous system. Regarding cardiac synchrony, explicit recommendations for which method to use when analyzing this type of synchrony are currently missing. To address this, we here compare five distinct methods suggested for analyzing physiological synchrony applied to electrocardiogram data. We use simulated and real data with specific synchronized properties and evaluate which method can detect which type(s) of synchrony. For each method, we provide an introduction, specify its implementation in the R analytical environment, and discuss how well synchrony could be detected with it. Specifically, cross-correlation, dynamical correlation, cross-wavelet power, cross-recurrence quantification analysis, and time series distance are being compared against each other. We discuss strengths and limitations of each method, and conclude by recommending specific methods for specific types of data and expected physiological synchrony in a given data sample.