Best performing models for predicting victims of gun violence based on social contagion, demographics
TUESDAY, Jan. 3, 2017 (HealthDay News) — Social contagion accounts for a considerable proportion of gunshot violence episodes, according to a study published online Jan. 3 in JAMA Internal Medicine.
Ben Green, from Harvard University in Boston, and colleagues examined the extent to which the people who will become victims of gun violence can be predicted by modeling gun violence as an epidemic during an eight-year period. The authors assessed the spread of gunshot violence over the network of individuals arrested using a probabilistic contagion model that assumed that individuals were subject to risks associated with being arrested together, as well as demographic variables. Participants represented a network of 138,163 individuals, of whom 9,773 were subjects of gun violence.
The researchers found that social contagion accounted for 63.1 percent of the 11,123 episodes of gunshot violence. The best performing models were those based on both social contagion and demographics. When identifying the 1.0 percent of people considered to be at highest risk of being shot each day, the combined model identified 6.5 percent of subjects, compared with 4.3 and 5.3 for the demographics model and the social contagion model, respectively.
“Violence prevention efforts that account for social contagion, in addition to demographics, have the potential to prevent more shootings than efforts that focus on only demographics,” the authors write.
Copyright © 2017 HealthDay. All rights reserved.
