Tang et al. reveal correlated simple contagions mimic complex ones

A team of researchers led by Katerina Tang and including Daniel Kaiser, William Thompson, Jean-Gabriel Young, Laurent Hébert-Dufresne, and Nicholas W. Landry has published a new paper on arXiv titled 'Emergent contagion complexity: Disentangling mechanistic complexity from correlated heterogeneity.' The paper addresses a fundamental challenge in network science: distinguishing between true complex contagions (where multiple exposures act synergistically) and simple contagions (where exposures act independently) that appear complex due to correlated heterogeneity in the underlying population. The authors introduce a formal measure of contagion complexity and develop an inferential framework capable of estimating mixtures of nonparametric contagion rules from time-series data.

This framework allows researchers to test whether observed complex spreading patterns are genuinely mechanistic or merely emergent from correlated simple processes. The paper includes 15 pages and 7 figures, and its findings have significant implications for fields ranging from epidemiology to social media analytics. By showing that past studies claiming evidence of complex contagion may instead reflect heterogeneous mixtures of simple rules, the work provides a more parsimonious lens for understanding how behaviors, information, and diseases spread through networks. The authors' framework offers a practical tool for model selection and hypothesis testing in contagion research.