In this study, school shootings based in the U.S. were probed between 1990 through 2017.
From the findings: “In this paper, we attempted to detect and estimate copycat effects using spatio‐temporal panel count models.”
“This class of models allowed us to determine whether an initial school shooting spawned subsequent attacks in neighboring states and the same state in the following years. Estimating the temporal dependence helped us to answer whether the frequency of school shootings in one state increases if the same state reports a higher number of victims in the past year.”
The results showed how there are drastic spatio-temporal effects in American school shooting data.
“This information may help authorities plan a response after school shooting incidents. In addition, we found that the victim counts cluster temporally after an initial shooting which suggests that the media should alter their coverage of school shootings and mass killings in general. Ideally, the media should find a way to cover these events while minimizing the risk of provoking additional shootings,” the study’s authors concluded in their findings.