Researchers gathered records of school shootings and mass killings from several data sets and fit them into a mathematical "contagion model." The spread they found was not dependent on location, leading researchers to believe that national media coverage of a mass shooting might play a role. On average, mass shootings occur about once every two weeks in the United States and school shootings happen about once a month, the study said.

"What we believe may be happening is national news media attention is like a 'vector' that reaches people who are vulnerable," said Sherry Towers, a research professor at Arizona State University and lead author of the study.

Those vulnerable people are those who have regular access to weapons and are perhaps mentally ill, Towers said. Once "infected" with knowledge of a shooting from national media coverage, data shows that a person is more likely to commit a similar crime. "When at least three people are shot, but less than four people are killed, the media reports tended be local," Towers said. These shootings that received local news coverage, but no national news coverage, did not have the same contagious effect, according to Towers.

Collecting information about school shootings and mass killings wasn't easy, Towers said, noting "right now there is no federal database on these tragedies."

Towers said there are still a number of important questions left unanswered, and creating consistent data about these incidents is the first step. "An official database needs to be compiled," Towers said. "The dynamic in the society needs to be addressed so we can fix this public health crisis."

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0117259#sec014