Speaking of Psychology: Can we prevent school shootings? With Dewey Cornell, PhD

From Columbine to Sandy Hook to Uvalde, the list of schools and communities devastated by gun violence grows every year. Dewey Cornell, PhD, an expert in school violence prevention, discusses how schools can respond when faced with a potential threat, and how students, teachers, administrators, and law enforcement can work together to keep schools safe.

Dewey Cornell, PhD, is the Virgil S. Ward professor of education at the University of Virginia. He is a forensic clinical psychologist, the director of the Virginia Youth Violence Project, and a faculty associate of the Institute of Law, Psychiatry, and Public Policy. During his clinical and academic career, Cornell has worked with juvenile and adult violent offenders, testified in criminal proceedings and legislative hearings, and consulted on violence prevention efforts. He has authored more than 300 publications in psychology and education, including studies of juvenile homicide, school safety, bullying, and threat assessment. He is the principal author of the Comprehensive School Threat Assessment Guidelines, which is an evidence-based model of school threat assessment used in schools across the United States and Canada.

Kim Mills: From Columbine to Sandy Hook to Parkland to Uvalde, the list of schools and communities devastated by gun violence grows every year. Every time a school shooting is in the news, it’s easy to wonder what, if anything, could have been done differently. Were there signs that were missed? At what point, if any, could or should someone have intervened?

Psychologists who study school violence have been researching these questions. Today we’re going to talk about what they’ve found. How common is gun violence in schools? How do most schools deal with the threat of violence right now? How can teachers, administrators, and law enforcement officials tell real threats of violence from false alarms or pranks? When a real threat is identified, what kind of interventions work? And how can schools avoid underreacting to genuine threats and overreacting to false ones?

Welcome to Speaking of Psychology, the flagship podcast of the American Psychological Association that examines the links between psychological science and everyday life. I’m Kim Mills.

My guest today is Dr. Dewey Cornell. Dr. Cornell is a forensic clinical psychologist and the Virgil S. Ward professor of education at the University of Virginia, where he directs the Virginia Youth Violence Project. He is the principal author of the Comprehensive School Threat Assessment Guidelines, which is an evidence-based model of school threat assessment used in schools across the United States and Canada. He’s currently directing a national study of school threat assessment with funding from the U.S. Department of Justice. Dr. Cornell has authored more than 300 publications in psychology and education, including studies of juvenile homicide, school safety, bullying, and threat assessment. 

Dr. Cornell, thank you for joining me today.

Dewey Cornell, PhD: Thank you for inviting me.

Mills: Before we start talking about threat assessment, which I want to do, let’s put this discussion in some context. This episode won’t air until January, but today we’re talking right now in late December, the day after another school shooting, this one in Madison, Wisconsin. So the topic feels especially important right now in this atmosphere. It’s hard for students and their parents not to feel worried about sending their kids to school every day. So I want to start with the question, how common is gun violence in U.S. schools and how worried should students and parents be about the threat of school shootings?

Cornell: That’s a very important question, and certainly when we’ve had another tragic shooting like we’ve just had this week, it’s very hard not to feel that schools are not safe places, that gun violence seems to be pervasive, and parents understandably are very worried about sending their children to school. But I want to offer a little bit of reassurance there. On the one hand, gun violence is a very serious problem in the United States, but it’s actually less frequent in schools than it is outside of schools. And that, statistically speaking, we have 130,000 schools—even though we’ve had a number of school shootings, the average school is very unlikely to ever have a homicide shooting. It calculates out about once every 6,000 years. So although it seems counterintuitive, schools are statistically one of the safest places from gun violence in the United States. We have more gun violence in our stores and restaurants and parks and parking lots and in our homes, by a large margin, than we have in schools. So don’t get me wrong, school shootings are a terrible tragedy. We have to do everything we can to prevent them, but statistically schools are much safer than it feels.

Mills: Around the beginning of this school year, after the shooting at Apalachee High School in Georgia, there was a surge in violent threats against schools nationwide. These were really disruptive, in some cases closing schools for days. Do surges like this usually follow high-profile shootings? And when they do, how can schools and law enforcement tell serious threats from pranks or copycats?

Cornell: Yes, after a school shooting, we do see a surge in student threats, but we also see a surge in sensitivity to threats. So statements and behaviors that might otherwise have been ignored or handled informally become more concerning in the wake of a school shooting to school authorities and to law enforcement. So they also react strongly. So we have both students being stimulated by school shootings to say and do threatening things, and we have authorities who are on edge and worried and reacting strongly to student behavior.

As I said a minute ago, schools statistically will almost never have an actual shooting or an actual homicide, not in thousands of years, but every school has threats. Schools frequently deal with students who say or do things that are threatening or concerning in some way, so it’s a pervasive problem. At the same time, the worst possible outcome is very unlikely, and so that creates a real dilemma for schools. What do we do with students who make threatening statements?

Mills: And that leads us to the Comprehensive School Threat Assessment Guidelines that you were involved in creating. Can you tell us about those guidelines and how does this model of threat assessment work?

Cornell: Sure. So threat assessment is a form of violence prevention that was originally developed by the Secret Service to protect public figures. But within a few years, folks began to realize that the general principles of behavioral threat assessment and management, that’s the formal term for it, could be used to prevent workplace violence, domestic violence. And as we started to see shootings in schools increase in the 90s, the FBI and Secret Service said, Hey, why don’t we try this in schools to prevent school shootings? And that’s when I went from a meeting with the FBI, back to Charlottesville, to the University of Virginia, and I talked with my students and colleagues and I said, Hey, let’s try to develop something.

And so we work with a group of local educators, folks in the trenches, in the classroom, in the hallways to say, how do we deal with students who make threatening statements in a way that’s going to be reasonable and fair, safe, and avoid the two errors of threat assessment? The two errors are overreaction to threats that are not serious and underreaction to the rare threats that are truly serious, where someone poses a serious threat. And so we developed the set of guidelines. We worked with local schools, we tested them, we refined them, we developed a manual, and then we did all the stuff that academics do. That is, we conducted a series of studies and went out and got grants and developed a training program and so forth.

Mills: In the real world, what does this feel like? What does it look like? So if I’m a school administrator and I’m following the threat assessment guidelines, how do I make the judgment that this is a serious threat or this is not a serious threat?

Cornell: So threat assessment is all about collecting information from multiple sources using a multidisciplinary team, so it’s not on the shoulders of one person to make this weighty decision. But with a multidisciplinary team, with school administration, school support staff—that is, counselors, psychologists, social workers—and if they have a school resource officer, also the law enforcement officer, but all of them together, we’ll gather information.

So we have an interview protocol for this student who’s been identified as the student of concern. We have an interview that we use for witnesses and reporters of concern. We look at the student’s background and history, and depending on how complex and serious the threat is, we may be able to resolve it quickly and easily. A student was just joking or they were expressing themselves in an exaggerated way. But if there’s reason to be more concerned, things don’t quite add up where we’re not really confident the young person is being straight with us, then we do a more intensive assessment with more extensive interviewing and gathering of data. We might involve law enforcement. In some cases, we might conduct searches for weapons. We might do a home visit.

So our guidelines really have a kind of calibrated step-by-step approach. We have five steps and some threats are resolved at step 1, some at step 2, some at step 3, and so forth. So it’s a calibrated process based on the complexity and seriousness of the case.

Mills: What does the research say about how effective threat assessment is? Do you have evidence that it’s prevented school shootings? And are there other outcomes that you’re looking for in your research?

Cornell: Yeah, so there’s a lot of outcomes that we’re looking for. We’ve done a half a dozen controlled studies involving hundreds of schools. We’ve done some implementation studies with thousands of schools. And so we know that most cases can be resolved without removing the student from school, without expelling them, without arresting them. And we can provide them services because we identify problems that they’re having, peer conflicts and bullying, and we’ve got lots of data to show that that can be done. It can be done effectively and safely. Very few threats result in some kind of violent action. We’ve had some fights, but no shootings, thankfully, no homicides.

But the question of course people want to ask is, Well, have you prevented a school shooting? Well, certainly we have cases where students seem to have been very serious about committing a homicide and they’ve been resolved, and we’ve had no homicides. But let me just say, scientifically, you need a randomized controlled trial. And school shootings are so rare. As I mentioned, fatal shootings are about once every 6,000 years for the average school. If you think about that from a statistical point of view, you need thousands of schools that you’re going to follow over multiple years to be able to demonstrate that you have reduced the rate of school shootings, because school shootings are quite statistically rare.

And so I can’t claim that we’ve done a randomized controlled trial of school shootings, but we do have controlled studies to show that we can reduce a lot of conflicts and problems in schools. We can show that students feel safer in those schools. Teachers feel safer in those schools, and about three-fourths of the students get referred for support services. So we identify a lot of problems like bullying or peer conflicts or maybe mental health needs. We have a lot of kids with special education with disabilities that we need to make referrals to their special ed team to maybe enhance their program or their plan.

Mills: What are the roles of the various people within the school regarding threat assessment? For example, teachers, administrators, other students, what do they have to do to be effective in stopping school shootings before they happen?

Cornell: Yeah. Well, safety is everyone’s responsibility. Parents, teachers, students, support staff. Everyone needs to be observant. And if they have concern, they need to report their concern. That’s all they need to do. They don’t need to do a threat assessment themselves. They don’t need to make a determination that someone is or is not dangerous. If they are concerned about something, the school should have multiple ways for them to report it. Of course, we’d like them to just come forward and talk to a trusted adult. That’s the most effective way. But some schools will use an anonymous reporting system, that’s fine. It doesn’t give as much information often as we would like, but in any way that a concern can be reported, then the team, it’s their responsibility to gather additional information and make that kind of determination about the seriousness of the threat.

But we need everyone in the school to understand that there’s a process, that it’s a fair systematic process, that it gathers information, and it’s not going to mean that every student is removed from school. That’s kind of a problem we run to in threat assessment because parents and teachers and students are worried, and if a rumor gets out of hand and someone thinks that a student is dangerous, the first thing that people think is, well, let’s get them out of the school. Let’s remove them from the school. That turns out not to be an effective safety response, and it’s certainly not helpful to the student. So we want to, in most cases, provide services and support, resolve a conflict, resolve an underlying problem, and then move on from there.

Mills: So if you’re not going to remove a student, what are the alternatives? What’s something that does work when you’re seeing a student who is problematic in that way?

Cornell: Okay, so every threat is different. Sometimes a threat is just a joke, and it can be resolved with an apology and clarification. Sometimes the threat is because somebody has had it with somebody and they’re itching for a fight. And so if you work in middle school, middle school kids will get into fist fights. And our school counselors and psychologists, they mediate conflicts between youngsters, and we certainly want to support that, and it usually works. And what we found is that very few threat cases result in some kind of physical assault, about 5%. And those are almost all fights. As I said, very few injuries in the state of Florida. We looked at 23,000 cases, and we found almost all of them were resolved without any kind of assault or fight. We had 10,000 cases where we looked at what kind of injury occurred, and less than half of a percent had a serious injury where someone maybe had to go to the ER for some kind of injury. So those are very, very low rates.

And most of the time we’re trying to draw upon the school’s support services. So a school might have multi-tiered systems of support. They certainly have special education services. They may have PBIS, they may have restorative practices, they may have anti-bullying programs. The whole range of support programs that we already have in schools can be selectively applied in most threat assessment cases.

Mills: Now, I read a story in the newspaper just the other day about how students are using cell phones to spur violence and record each other having violent altercations. And yet you’re making it sound like it’s not really a problem. I mean, is it a problem? Are cell phones in school essentially spurring kids to act out more?

Cornell: Well, I’m not trying to imply that we don’t have a problem with student behavior in schools, and we’ve had some increase in aggressive behavior, and we have a lot of mental health concerns that we’re not fully staffed in our schools to be able to manage. So I don’t want to sound like I am dismissing the problem, but I’m saying the problem is not so much school shootings, which are terrible, but very rare. But under the radar screen, we’ve got a lot of kids having conflicts and problems and bullying and those things we need to deal with. With a lot of the evidence-based programs that the psychologists have developed for use in schools

Mills: In the past several years, there’s been a lot of debate about whether police officers, who are often called school resource officers, should be working in schools. What’s your view? What role should law enforcement play in school threat assessment?

Cornell: Yeah. Well, we’ve seen some cases where it seems that law enforcement has overreacted. And after a high-profile shooting, we’ll see cases of a 10-year-old being arrested or 12-year-old being arrested. And in many of those cases, when we dig into the facts of what happened, it wasn’t that severe an act. It didn’t justify arrest or criminalization. Now, there are some cases where arrest is necessary, of course, but this has created a perception that maybe we need to get law enforcement out of schools. But I think the problem really arises when law enforcement are not in the schools working as part of the team. So what we do in threat assessment is we want them to be part of the team. We want them to go through the training, we want them to understand the approach that we use. And we found that law enforcement is very receptive to this.

And when we have law enforcement as trained members of the team, we get very good results. Fewer than 1% of the students are arrested, and law enforcement officers in schools can have a positive effect when they’re not enforcing discipline, but being role models and counselors and sources of support in the school. Now, I am concerned that law enforcement officers might not be adequately trained or prepared or selected to work in schools. So we’ve got a lot of work to do to improve that, but I certainly think pulling them out of schools is a recipe for trouble. Many schools have found that when the officers were removed, who’s going to break up the fight? If you have a melee in the school, in the lobby, are you asking the history teacher to step into harm’s way and do something about it? And of course, students also in our surveys, the vast majority of students feel that the presence of the schools resource officer makes them feel safer in the school.

Now, this is across racial groups. This is true for Black, Hispanic, White, other racial groups. We’ve done statewide surveys of that. I know there were situations where things have gone poorly, where we need to have interventions to make things go better. But on the whole, I think law enforcement can be a positive force in schools if they’re well trained, supervised, have a well-defined role.

And yes, I am concerned about the students with cell phones who are taking images of fights and putting them on the internet, and that has a great contagious effect. Everybody wants to get in on the act and be part of doing something as sensational as that. So we have to do a lot of education with our students around the hazards of social media, around appropriate behavior, and we need to have some restrictions on the use of cell phones in our schools.

Mills: You mentioned a moment ago about sometimes overreacting by law enforcement and arresting 10-year-olds or younger children who really are not seriously threatening. And I’m just wondering whether threat assessment works differently for different ages or grade levels. I mean, how should a school respond to a threat from a very young child, like in elementary school?

Cornell: Yeah. Well, you always have to consider the child’s age, their developmental level. If they have a disability or any other kind of mental health or educational condition, you have to consider that to understand where they’re coming from, what they meant and what they’re capable of doing. We actually have more threat assessments conducted in elementary school than we do in secondary school. There’s more of those kids, and they’re immature, and they will say and do things impulsively because of their age. And in most of the cases, the vast majority, 80% of those cases, it’s resolved as what we call a transient threat. That’s not serious.

As kids get older into middle and secondary school, they’re a little less impulsive. They make fewer threats, but those threats have a little bit more concern behind them because they’re more capable of injuring someone. And so we have to adjust the threat assessment process to the age of the child. And of course, by using a school-based team, we’re able to do that because we want the folks who are in the school, who are familiar with their school, their students, the culture of their school. We want them making these decisions rather than maybe an authority from outside who’s using a uniform set of standards. And I think it’s a big difference between an 8-year-old saying something and a 16-year-old saying something, and our expectations for their behavior are different. And unfortunately, if law enforcement or outside authorities of any kind are applying a strict uniform standard that doesn’t consider the age of the student, there’s a potential for overreaction there.

Mills: Well, how can schools avoid underreacting to serious threats and overreacting to unserious threats?

Cornell: Well, that’s why they need an evidence-based threat assessment program. It’s a systematic way to gather information, share it with colleagues, and make a reasonable decision. Threat assessment isn’t perfect. It’s not guaranteed to always be correct. But what we found is that we can dramatically reduce removing kids from school, kids being punished rather than helped. And that’s very important because as I said, almost no school will ever have a shooting, but every school is having students make threats. And so there’s a big problem there. If we were to apply zero tolerance, and many schools use zero tolerance, they say, you’ve broken the rule and regardless of the circumstances, regardless of your age, whatever, we’re going to remove you from school. And what research has found is that removing kids from school doesn’t improve their behavior. It doesn’t make the school safer. It increases the likelihood that the student is going to misbehave and drop out of school down the road. So we want to be very careful about using school removal. I would remove a student from school when it’s just really clearly not safe, and it’s in the child’s best interest to have a different, more structured and supportive environment than the regular classroom or the regular school. But what we have found is that in the vast majority of cases, that’s not necessary.

Mills: Is there anybody in this whole mix who is more critical or pivotal than anybody else? Is it other students who report the threats or teachers or guidance counselors or even parents?

Cornell: Yeah. Well, when it comes to reporting concerns, we want everybody to know how and be willing and to see the difference between snitching and seeking help. But what we have found when we study school shootings, school shootings that have occurred, and school shootings that have been averted, that usually it’s a student who has key information, who knows something. And so we very much want to encourage student reporting, and students are going to be more willing to report when there are trusted adults in the school, when they have confidence that the adults in their school have their best interest at heart, that aren’t going to simply take a zero tolerance approach, that the school climate is that discipline is fair and that they’re treated with respect.

We’ve actually done a number of studies of what makes students willing to report a threat, and a positive school climate, what we call an authoritative school climate, not authoritarian, but authoritative—that is, they feel that the rules are strict but fair, and that the adults at school support them and respect them and have their best interest at heart. That kind of climate is what we want to see in our schools. And then students are willing to report threats. Parents are willing to report threats as well. And I’ve worked on a number of cases where that was really critical information.

Mills: What is your view of active shooter drills? I mean, many schools are requiring them now, and they’re trying to teach students how to respond to an active shooter in the building. Now, you’ve talked about how unusual shootings are in schools. Should we be doing this? I mean, this traumatizes a lot of kids.

Cornell: Yeah. Well, I am very concerned about active shooter drills, particularly the ones that involve some kind of reenactment, some kind of simulation, ones that involve weapons being displayed or people being simulated being shot or attacked. I think those are harmful and not helpful. I think some training and orientation of staff about how to deal with an intruder, and most of the time the intruder’s not going to be an active shooter. There are other kinds of incidents that schools have to deal with—someone coming into the building that shouldn’t be there. So there are reasons to practice security in a school, but I think the reenactments and the simulations have gone too far. I think if law enforcement wants to practice their SWAT procedures, that should be done on a weekend when there aren’t students present, I think trying to practice something live is going too far.

And the same is true for a lot of the security measures. Schools are spending literally billions of dollars nationwide on high tech security measures, facial recognition, gunshot detectors, metal detectors, bullet-resistant glass, and so forth. And these are extraordinarily expensive. And at the same time, these same schools don’t have enough school counselors. They have one school psychologist covering five schools, schools, they’re understaffed. And so their ability to implement prevention programs and to build relationships with students where we can get information that can prevent, head off a shooting, that’s a huge gap. So we need a better balance between security and prevention. And unfortunately, the emphasis on these high-profile shootings that are very dramatic and troubling and distressing, that leads people to make an emotional decision and school boards to make an emotional decision, it says, all right, let’s batten down the hatches. Let’s make our school building into a fortress.

But I think the return on investment for that is not very good. If you want to take an economic perspective, and from a psychological perspective, it’s not a very good return in terms of building trust in our students and having enough staff to work with our students. Let me mention one other fact that people sometimes overlook. They’re focused on the idea of a shooting in the building. So they try to lock down the building, they try to keep everybody out of the building, and they try to lock all the doors in the building. But if we look at school shootings that have occurred, only 30% of school shootings occurred in the building. So it’s the parking lot, it’s the football game. It’s places outside of the school, which is consistent with gun violence in general. Gun violence is more common in these other environments than in the building.

So if we’re concerned about security, just making the building secure is not a good expenditure of funds, because the shooting is more likely to occur outside of the building. And actually, if you have a young person that you’re concerned might be potentially violent, they’re much more likely to shoot somebody outside of school. So if we’ve got a counselor, if we’ve got a psychologist, if we have a social worker working with a young person that we’re concerned about, I would rather be focusing on preventing them from committing a shooting anywhere in the community rather than just the idea that we can prevent them from doing it in the hallway of one particular building. So I think we’ve got to think more broadly about prevention here and less narrowly about turning the school building into a fortress.

Mills: I think a lot of what you’re saying will give hope to our listeners, many of whom are our parents, that they don’t have to worry as much as perhaps they are about sending their kids to school.

Cornell: So let me tell you one thing that I like to tell parents, and it’s kind of a hard message to get to them when they’ve thought about school shootings and they have imagined how horrible it is, and I had three children in the public schools, I’ve got grandchildren. So I’m certainly able to look at this from a parent’s perspective, but I know scientifically that if a parent is afraid to send their child to school because they think of the risk the child could be shot, then that parent logically should never let their child go to a restaurant, never let their child go into a convenience store, never let their child go into a parking lot or a park, because those areas have double, triple, quadruple the shootings that school buildings have. Schools ironically are safe havens. In many, many communities with high crime schools are the safest places for our kids to be.

And I know it’s hard because as a parent, we can imagine those school shootings and they’re so horrendous, and we’ve got to do everything we can to prevent them, but we’ve got to prevent the shootings that occur all over our society. We have about 300 shootings every day in the United States. We have over a hundred people killed by gunfire every day in the United States. We just don’t hear about those cases. We hear about the ones in the school building that are so heartrending, so heartbreaking, but less than 2% of homicides of school-age people occur in schools. Less than 2% of the homicides occur in school. 98% occur outside of school. So I want measures that focus on protecting all of our kids and not just in the confines of the school building.

Mills: So just to wrap up, what’s next for you and your research? What are the questions you’re still trying to answer?

Cornell: Well, oh gosh. There’s lots of questions. Because we want to make this training program effective. We want to make the implementation effective. We want to make sure it’s fair and equitable. So we’ve done a number of studies on that. We currently have a national study with districts from six states implementing the model. We’re measuring the fidelity of implementation. That is, how high quality are they following the procedure? Because schools have limited resources and staff are often burdened with lots of responsibility. So what can we do to make sure that they follow the threat assessment protocol fully, and if they do, do they get better results? We want to show that.

We also want to show the equity and fairness of threat assessment. I know in school discipline, Black students are removed from school at disproportionate rates, about double the rates of white students for school discipline. In general, what we found with threat assessment is that disparity shrinks almost to zero. Our differences in our latest study between White and Black students, were about 2%, not 200%, 2%. So we’re getting very equitable outcomes across racial groups. Also across groups defined by disability. We know that in the United States, students with disabilities get removed from school at one and a half to double the rate of other students without disabilities, but not with threat assessment. If we can route those students into a threat assessment, their outcomes are no different, no different than the students without disabilities. Same for families with low income. We know that low income families often get fewer services and more punishment, but in large scale studies of threat assessment, we find they’re treated comparably.

So we’re very optimistic that we can get some very positive benefits from students quite apart from the horror of the school shooting. We can discipline students in a more equitable manner, and in most cases, 75% of the cases they’re receiving services rather than being put out of school.

Mills: Well, Dr. Cornell, I want to thank you for joining me today. This has been very interesting.

Cornell: Thank you very much for the opportunity. I’m happy to follow up if anyone has questions, wants to contact me. We’ve got lots of research and materials to share.

Mills: Oh, great. Thank you. 

You can find previous episodes of Speaking of Psychology on our website at speakingofpsychology.org or on Apple, Spotify, YouTube, or wherever you get your podcasts. And if you like what you’ve heard, please subscribe and leave us a review. For those of you who want to know more about psychology on the work of APA, you can join us as a supporter. You don’t have to be a psychologist. The benefits include a subscription to our award-winning magazine, Monitor on Psychology, as well as discounts on our books and other publications. So if psychology is your passion, or if you just want to know more about the science and discipline, you can become a supporter by visiting at.apa.org/speakingopens in new window. That’s A, T as in Tom, apa.org/speaking. If you have comments, questions, or ideas for future podcasts, email us at speakingofpsychology@apa.org. Speaking of Psychology is produced by Lea Winerman.

Thank you for listening. For the American Psychological Association, I’m Kim Mills.