The Mathematics and Geography of Mass Shootings in the U.S., 1982-2017

Abstract

The authors use simple bilinear regression to assess changes in the geographical movement (latitude and longitude) of mass shootings in the United States between 1982 and 2017. The path taken by the location of the ninety-five mass shootings over the 36-year period has shifted south. An analysis of differences by census region and blue/red state distinctions within each census region reveals disproportionately many mass shootings in Midwestern states between 2000 and 2008, and disproportionately many in red Southern states over the past three-plus decades.

Analyzing the risk of mass shootings in the United States

“…Spatial risk analysis of rare events have used a modified logistic regression for the occurrence of landslides (Guns and Vanacker, 2012;Bai et al, 2011), a generalized linear model for nation state failure (Calabrese and Elkink, 2016), and a hidden Markov model for the spread of a rare disease (Forbes et al, 2013). A simple bi-linear regression finds that the latitude and longitude of mass shootings linearly dependent on the year (Andersen et al, 2018). A kernel density estimation and K-means clustering algorithm are used to identify different levels of mass shootings in the United States (D'anna, 2020).…”