B0718
Title: Finding influential subjects in a network using a causal framework
Authors: Youjin Lee - Brown University (United States) [presenting]
Ashley Buchanan - University of Rhode Island (United States)
Elizabeth Ogburn - Johns Hopkins University (United States)
Samuel Friedman - NYU Grossman School of Medicine (United States)
Elizabeth Halloran - Fred Hutchinson Cancer Center (United States)
Natallia Katenka - University of Rhode Island (United States)
Jing Wu - University of Rhode Island (United States)
Georgios Nikolopoulos - University of Cyprus (Cyprus)
Abstract: Researchers across a wide array of disciplines are interested in finding the most influential subjects in a network. In a network setting, intervention effects and health outcomes can spill over from one node to another through network ties, and influential subjects are expected to have a greater impact than others. For this reason, network research in public health has attempted to maximize health and behavioural changes by intervening in a subset of influential subjects. Although influence is often defined only implicitly in most literature, the operative notion of influence is inherently causal in many cases: influential subjects should be intervened on to achieve the greatest overall effect across the entire network. A causal notion of influence is defined using potential outcomes. Existing influence measures are reviewed, such as node centrality, which largely relies on the particular features of the network structure and/or on certain diffusion models that predict the pattern of information or disease spreads through network ties. Simulation studies are provided to demonstrate when popular centrality measures can agree with the causal measure of influence. As an illustrative example, several popular centrality measures are applied to the HIV risk network in the transmission reduction intervention project and demonstrate the assumptions under which each centrality can represent the causal influence of each participant.
