Research sits at the core of our mission to educate for justice,” says President Karol V. Mason. The remarkable research that our esteemed faculty, students, and alumni perform directly impacts our society and helps move the needle forward toward creating a more just world. Their empirical data reveals inconsistencies, pushes boundaries, and opens minds to a whole new way of thinking. Our “Research on the Rise” series aims to showcase the ongoing, upcoming, and noteworthy research happening right here at John Jay.

Every year, as we watch yet another school shooting event unfold on the news, the public responds with a familiar set of questions. Why is this happening? What can be done to stop it? How can we keep our students safe? After being awarded $797,000 by the National Institute of Justice (NIJ), John Jay College Professor Joshua Freilich, Ph.D., along with Steven Chermak, Ph.D. from the Michigan State University; Brent Klein, Ph.D., from the University of South Carolina; and a host of talented students are setting out to answer questions like these by expanding their previous NIJ school shooter database. “We’re quantitatively exploring the root causes of school violence using a case-control comparative analysis,” says Freilich. The team hopes to compare and contrast the differences in offender backgrounds—including their social networks, family relationships, criminal and personal histories—that could lead up to an adolescent’s involvement in school shooting violence. “We want to identify the potential root causes and related factors that contribute to school violence,” says Freilich, “then use our findings to devise strategies and programs to promote student and school safety.”

The 2016 NIJ Grant
The motivation for creating the original national school shooter database—which in 2016 was also funded by the NIJ—was the lack of national-level data on the issue itself. “In talking to my colleagues, I realized that a lot of the strategies we used in terrorism research—even though they’re different substantive areas—could be applied to school shootings,” says Freilich. “Terrorism and school shootings are similar in that they’re very high profile and they get a lot of attention for obvious reasons. Thankfully, they’re really not as frequent as regular homicide. We had over 20,000 homicides last year. School shootings are much lower, even when you debate the criteria to label it as such.”

He goes on to explain the challenges when collecting data on school shootings. Obtaining “regular criminology reports” from sources like the Federal Bureau of Investigation (FBI) is standard practice, but there’s a higher level of difficulty encountered when researching school shootings or terrorism—with many offenders not detailed at all or at length in law enforcement databases—leading researchers to use “open sources” or publicly available information. “The open sources can run the gamut, from police reports, court records, newspaper articles, or even social media information.” Throughout their research in 2017, Freilich and his team harnessed public information to create their first school shooter database. To be included in this database, the shooting had to occur on school grounds and at least one person had to be injured or killed. With this as the criteria, not all of the shooters accounted for were students. Anyone committing gun violence on school grounds—be it staff members, visitors, or someone simply walking into the school—could be put into the database. Freilich’s current research narrows down the criteria even further.

The 2021 NIJ Grant
The research being conducted now is limited to adolescents (ages six to 19) who are enrolled at the school at the time of the shooting. “We’re really honing in on the adolescent population. We wanted to isolate the differences between these offenders and non-school shooters,” says Freilich. “So, we came up with two other categories, the first one being ‘regular violence’ where an adolescent might get angry, go to a bodega, and shoot someone there, but it’s not on school grounds.” The idea is to see if there’s something unique about those adolescents who target individuals specifically affiliated with the school and on the school grounds, as opposed to regular youth violence. The second category added in this year’s research is the “non-offender.” “We wanted to isolate part of the population that could have been a shooter, but weren’t. That would be non-offending students at the same school. We came up with a strategy that with each shooter we had to identify a non-offending student that was enrolled at the school at the same time, at a comparable age to the school shooter.” Thinking about the data in three rows, there’s: row one, the school shooters, attending the school, committing the shooting on school property where at least one person was injured or killed; row two, youths who committed a shooting in the community where somebody was injured or killed; and row three, non-offending students who were enrolled at the same school as the shooters.

To identify the root causes of school violence, the team wants to hone in on what’s unique about the school shooters compared to the other student categories. “Being trained criminologists, we have all these different theories of crime. We wanted to pull out some of the key predictors used to explain youth crime and apply them to these groups,” says Freilich. “A lot of these theories highlight ‘risk factors’ which put someone more at risk for anti-social or criminal behavior. Some of these risk factors would be suffering a trauma in the past, being a victim of abuse, being impoverished, suffering thoughts of suicide or self-harm, parents getting divorced, recently moving, being in a gang, or even breaking up with a girlfriend or boyfriend 90 days before the event.”

The team is also looking into “protective factors” that the subjects may or may not have had in their lives. “A protective factor might be that you were in a mentee-mentor relationship, maybe a teacher or a coach. Or, if you were really involved in school clubs.” All of these variables, both risk factors and protective factors, are being delved into for all three categories of adolescents, with the end goal of analyzing the differences between the three groups. “We really want to see if the school shooters have more or less of the risk factors or protective factors compared to the two other adolescent categories.”

The Research Methodology
The main idea behind the research methodology for the project is to leverage public information. For the first category, the school shooters, the team already has information from the first NIJ grant project, where they’ll be pulling out the students aged six to 19 enrolled in a school. For the second group, youth committing violence in the community, they’re using official data from uniform crime reports in the same area, location, and year. “The most interesting new development is the non-offending student category. From the first research project, we have some yearbooks, neighboring students’ information, and classmates giving quotes to media outlets. So, we have lists of students, some of whom were the same age as the offender, available in our files,” says Freilich.

To gather the data around the subjects, the team has about 60 different specialized web engines to systematically search information on each individual. “They pull up everything. Let’s say there was an obituary, that might tell us about their parents, noting if they were divorced. It might say if they had siblings or mention different key events in their lives,” says Freilich. “Even social media posts come up. These posts could identify if the student was in different school groups.” Then, for each individual selected to be in the database, the team creates a search file with source types on every entry. Afterward, they code the different variables—risk and protective factors—for analyzation.

The Research Students
The project incorporates the work of students at all different levels—undergrad students, master’s degree students, and doctoral students. “We have a young man, Manuel Martinez, who was a John Jay undergraduate and worked on the first NIJ research project. He graduated this past May from John Jay with a master’s degree in Criminal Justice and now he’s working on this project. This guy is terrific. I’ve never met a person with his work ethic. He was volunteering for us at first and now he’s a research assistant,” says Freilich. “We also have Emily Greene-Colozzi, who’s a doctoral student, as our project manager. She’s supervising and training all of our research assistants.”

Martinez was a “searcher” on the first project, gathering all the public information about the subjects, almost like a private investigator. “He found a ton of rich information for us. He’s just very resourceful using the search engines we have.” For the new grant project, Martinez has been elevated to a “coder,” reviewing the information gathered and filling in the values. “The students are invaluable. We couldn’t do the project without the students. It would be humanly impossible to finish the project in the allotted period of time without them. They’re fully committed to both the quantity and quality needed for this project.”

The Hopeful Outcome
The NIJ research grant officially concludes the last day of this year. Currently, the team is busy creating lists and searching information for the first two adolescent categories. “I’m hopeful that we can identify some number of risk or protective factors that consistently differentiate the shooters from the non-offending students and the regular community violence offenders,” says Freilich. “Depending on what those characteristics are, from a policy standpoint, it could be very useful. You might be able to finetune some of the programs we have in place to identify at-risk students. However, if we don’t identify these characteristics that’s also very interesting. Maybe we don’t find differences between the school shooter and the community violence offenders, maybe they both differ from the non-offending students. In that case you would conclude, possibly, that the school shooters are really just a subcategory of youth violence in general. Whatever we’re going to find in this research, it will have theoretical and policy implications.”

One of the driving forces behind the research is the understanding of public fear and anxiety surrounding school shootings, and the wide-reaching harm that they cause communities. “The consequences of a school shooting kind of ripple out. Just being a student at a neighboring school—so not technically direct victims—can cause anxiety. It goes way beyond the direct victims. You have a very wide net of affected people—be it the students themselves, custodians, work staff in the school,” says Freilich. “You have these terrible events that are occurring and they have very direct consequences. But the question going forward has to be, What can be done to alleviate it? In that sense, it becomes about gathering data and isolating some of the characteristics of the offender or the offense itself. If you have those findings, you have to have talks with academics, practitioners, and policymakers about them. Going from the findings to useful policies is really the way forward.”