Causes and Consequences of School Violence - Plenary Discussion, NIJ Virtual Conference on School Safety

 >> Good day, everyone.

Thank you so much for joining us today.

My name is Basia Lopez, and I'm a social science research analyst at NIJ, and on behalf of NIJ, I want to welcome all of you to this special discussion session panel on causes and consequences of school violence.

This panel includes subject matter experts who have explored the root causes and consequences of school violence from different perspectives.

We will also consider policy implications associated with their work.

Those experts, the discussions include Joshua Freilich, Jillian Turanovic, Patricia Campie, and Dorothy Espelage.

Our moderator, Emily Dolittle, she's from the Department of Education.

She will moderate the entire discussion.

Before I give it to Emily, I want to encourage the audience to use Q and A option on your screen to enter your questions for the panelists.

With this, with no further ado, I will give it to Emily.

Welcome.

>> Thank you so much, Basia, and hello to everyone.

It's such a pleasure to be here at day's two plenary session for the Comprehensive School Safety Initiatives Conference.

I was really thrilled to get the invitation to come and moderate.

So in thinking about setting up the session today and the best way to structure it, I was really focused on the theme for the conference this week for all of you and thinking about bridging research to practice, and I was thinking about a challenge that we often face in my work that I do at the Institution of Education Sciences, which is the research arm of the US Department of Education, where we have a very similar charge: how do you take this really important foundational research and actually change what's happening on the ground? How do you make changes in practice? And so I've set up some questions that I'm going to be posing to each of our panelists today to help to try to really dig in deeper to think about what are the lessons learned from this very foundational exploratory research that our panelists will be describing for you that are really looking at root causes and consequences of school violence, and how can we take that to really advance our understanding and really then make changes in terms of what's happening in schools around violence and safety? So what we're going to do is I just want to take a really quick moment to briefly introduce our four panelists to get us oriented, and then we're going to jump right in with my questions.

And as Basia mentioned at the beginning, as you're listening to my question and the responses from the four panelists, please think about your questions that you might have for the panelists that are related, and we'll try to take them as we go.

Otherwise, we will definitely address questions at the end with time remaining to make sure to include your thoughts and ideas.

So in terms of introducing our panelists, Josh Freilich is a professor in the Criminal Justice Department at John Jay College, and he has had two projects funded through the CSSI initiative.

Both use open-source data to understand the root causes of school violence, and so we're going to hear more from Josh about some of that work that he's engaged in.

Our second panelist today is Jillian Turanovic.

She's an associate professor in the College of Criminology and Criminal Justice and director of the Crime Victim Research and Policy Institute at Florida State University.

Her CSSI project is a meta-analysis of over 700 studies that look at the various sources of school violence at various levels that we care about: individual, institution, community, and broadly, her research is looking at youth violence and victimization, and then Trish Campie, who is a principal researcher at the American Institutes for Research, AIR, has a CSSI project that looks at school and community factors and understanding school violence, and her broader research interests are looking at gang-involved youth and young adults in the United States and Latin America and how to think about preparing communities and schools to sort of respond to and prevent violence, and then finally, last but not least, Dorothy Espelage, who is the William C.

Friday Distinguished Professor of Education at the University of North Carolina at Chapel Hill.

She will be talking a bit about her CSSI project, which is a systematic review and meta-analysis of the consequences of school violence, so kind of focusing at the other end of it, and her broader research looks at things like bullying and peer victimization in schools.

So without further ado, we are going to jump right into our panel discussion, and so, again, just to remind you of the structure here, I'm going to pose a question, and we'll just go in order.

So I'll start with Josh, and we'll go in order and give each of you a chance to respond to my question kind of in the context of your CSSI work, and then what we'll do at the end, as a reminder to our audience, if you have a question that's pertinent to this issue, please submit it, and we will try to get to that.

Okay.

So we are going to jump right in.

So, John, so one thing that really struck me when thinking about the CSSI initiative and trying to understand school violence and school safety, I was really struck by this charge to really bring in and consider multiple perspectives, and we can think about that in terms of sort of, like, who is involved in school violence, like, students, teachers, administrators, families, the broader community, but it's also this idea of different theoretical perspectives, and one particular kind of theoretical perspective or framework that's very common in education is the multi-tiered systems of support, or MTSS, and I was wondering if you could talk to us a little bit about the work you do looking at school violence and sort of what your theoretical frame is when you do that work and whether you've really thought about how this perspective from an education may complement, inform or kind of change up the way you think about this issue of school violence, kind of what is bringing that particular theoretical perspective due, or how does it inform the kinds of ways you think about the causes and consequences of school violence, and so I will turn it over to you.

>> Oh, well, thank you.

Good afternoon.

I first want to just thank NIJ for including me in today's panel, and my co-PIs and I were very grateful for all of NIJ's support.

Our first NIJ study was by myself, Steve Chermak from Michigan State University and Nadine Connell now from Griffith University.

We used open sources, public information to create a database of all school shootings that occurred on K-12 school grounds in the 50 United States.

We focused on the time frame from 1990 to 2016, and we only included cases that resulted in at least one injury.

We focused on both events as well as offenders, so individual level attributes, including many social control constructs.

We identified over 650 school shootings, and they ran the gamut from intentional shootings to self-harm to accidental discharges.

We also identified 252 known intentional shooters, so these individuals that we knew who their names were, and these 252 were adolescents.

They were under age 19, and we found that a large number of them dealt with school, personal and family problems.

They may have benefited possibly from MTSS' interventions for individual support.

What I mean is that the average age of these adolescent school shooters were 16, around 57 or 58 percent were current students, but close to 10 percent had dropped out.

Over a quarter had psychological problems such as depression, suicidal thoughts or behavioral disorders, and this is what we documented in the open sources.

Another third had family problems, and this includes family violence or abuse, conflict in the home, as well as parents with substance abuse and criminal histories.

Over 20 percent of the school shooters who were adolescents had been suspended or expelled at some point, and another 13 percent had failed a class while in school.

Close to 30 percent were gang members, and a little over 30 percent had a criminal record.

Twenty-nine percent had also suffered peer aggression before the shooting had occurred, so in sum, those particular individuals with those characteristics that I just listed, it's possible that an MTSS intervention geared towards them may have been useful.

>> Thank you so much, Josh.

So, Jillian, I would like to pose the same question to you.

If you could tell us a little bit about what theoretical lens you bring to your work and then kind of talk us through whether you've thought about sort of how this MTSS framework may or may not apply to the kinds of things you're thinking about.

>> Sure.

So I'm a criminologist by training, and so by nature, I approach the study of school violence originally when we're carrying out this meta-analysis of this massive body of research, interdisciplinary research that has been produced on this topic, and when thinking about what key individual, school level and community level factors that we wanted to code for and assess in regards to school violence, you know, originally we were thinking about traditional criminological predictors, things like low self-control, social bonds to parents and to school, substance use, antisocial attitudes, deviant peer influences, neighborhood disadvantage, school disorder, factors such as that.

When we got into the literature and actually started coding for studies, we realized that taking a purely criminological theoretical approach was going to be too narrow because especially in this literature where so much literature is produced in education, in developmental psychology, in social work, there was a much broader focus on social skills of children, on their peer and group dynamics, social competence and different elements of cognition.

And we realized rather quickly that we needed to take more of an interdisciplinary perspective and to be inclusive of all of these factors, and I'm glad that we did because some of those non-criminological measures were risk factors that we were coding for ended up to be the strongest predictors of school violence, things like social competence, which refers to how well kids can interact in social situations.

It refers to their social skills, their ability to collaborate with other youth.

These factors really emerged as robust risk factors for school violence, and so I think taking a broader theoretical lens is needed when approaching the study of school violence and to recognize as well that there are factors that are influencing youth's behaviors and vulnerabilities at school that extend beyond the individual.

So there's factors that need to be accounted for in classrooms, at the school level, and even if it appears that, you know, youth characteristics are what really need to be targeted by interventions.

We have to also be mindful that a lot of these sources of antisocial behavior and antisocial attitudes originate in the home and in the community.

And so I think, you know, the MTSS framework is one that I think, especially criminologists in my field should be more inclusive of and to start thinking about interventions in terms of, you know, universal factors, targeted factors and indicated factors.

So I hope that provides at least some insight into my approach and that that's useful.

>> Thank you so much for that, Jillian.

I think, if I'm understanding what you're saying, you're kind of -- I mean, I think you're seeing like MTSS is potentially a useful organizing framework...

>> Yes.

>> ...for these multitude of factors that you're trying to kind of sort through and understand, so that's...

>> Exactly.

>> ...helpful.

Yeah, no, thank you.

Okay.

Trish, I'm going to pose the same question to you.

If you could tell us a little bit about sort of the theoretical motivations or kind of how you approach your work and whether you've thought about sort of this idea of MTSS and how it may or may not intersect with what you're doing.

>> Well, thank you for this question, Emily.

I think it's a really interesting question, and I also wanted to stop and just thank Basia and NIJ for the opportunity to talk about our work.

You know, our work is led by a group of people, not just me, so I want to give a shout-out to Anthony Peguero, my co-PI, and our partners at John Couris and the Wandersman Center and all of the amazing staff and the schools that we're working with.

So in our study, we're really looking at sort of the dotted line between communities and schools when we think about root causes of violence, and so it turns our theoretical attention towards the structural causes of violence that exist in the community and that staff and teachers, parents and students walk into the school with and then looking at what's happening in the school to either attenuate those things, so reduce whatever issues are coming in to the school or amplify the issues and actually make it worse for folks when they go back into the community, and so when I think about the tiered system of supports, I think maybe that is in some way sort of the evidence of the inequity we see in the broader community.

So when you think about structural causes of violence, you think about inequities in housing, in healthcare, systematic bias in systems, things like that, and so people are differentiated in a community by the amount of inequity they face, and so we're not -- That's our approach, right, is to understand those different inequities and level of risk in the community, and it strikes me that maybe the tiered systems of support, again, are a bit of a reflection of those inequities but kind of turned into the consequences of them, right, where you have select groups of students that are really experiencing those inequities in the most gravest way, right, through their behavior and then the selected -- and the other students -- And so in our study, we're really trying to understand how those differences between these structural inequities in the community, how that affects the readiness of students and teachers and parents and business leaders and neighbors and faith people to actually work together and pull in the same direction about even what is school safety.

The definition alone, these groups all have a different view of that, and obviously if you have a different view of the problem overall, it's going to be difficult to work in a comprehensive way on a solution.

So yeah, those are my thoughts.

>> Thank you so much for that, Trish.

Can I just -- I want to probe really quickly on one thing that you said which I was really intrigued by.

You talked about how sort of those tiered systems within the school sort of are this kind of interesting reflection of those inequities that children are bringing to school with them.

I'm sorry.

Can you say just a teeny bit more about this? >> Yeah, so maybe, I mean, honestly I just thought about this in reflection of your question, but maybe there's something to that where the young people who are acting out the most and who have the behavioral needs that are resulting in a more intensive level of attention by the school, maybe those are the students that if you looked at it through a structural inequality lens, you might see that they actually are the ones who have been experiencing the highest rates of victimization in their community, their families have the least amount of access to employment, healthcare, housing.

It's an empirical question.

You know, we're not studying tiered systems of support, so we're not going to answer the question through our study, but it did strike me when you asked the question that, you know, maybe there is a relationship between sort of causes and then the consequences for which kids are in which groups in the tiered system of support.

>> So interesting.

Thank you.

No, I appreciate that.

That's really great.

Okay.

So, Dorothy, I'm going to now turn it to you, same question.

MTSS, how does this fit into your work and your approach to things? >> Yeah, so I am a psychologist and educational psychologist by training, so the MTSS framework is one that we utilize across all of our research for the last 25 years.

It's changed names in those 25 years, but it's still [Indistinct] tiered intervention.

So in relation to the particular paper that was published in "Psychological Bulletin" that Emily just described, which was a meta-analysis that I conducted with my colleague Dr. Polanin at AIR, as well as my colleague, Jennifer Grotpeter, Dr. Grotpeter from DSG.

In that -- So it's helpful to understand what we found there, and I'll do that briefly, and then talk about the results really did lend itself to some of the work that we're doing universally or targeted or indicated, and in that particular paper, so we screened 21,000 abstracts and landed with 114 longitudinal school-based studies with 765 effect size is about 95,000 kids, and what we were looking at were the consequences of school violence, and those consequences of school violence range from mental health symptomatology like depression, suicidality, self-esteem as well as school performance, grades, skipping school, as well as conduct disorders and delinquency.

And interestingly, we looked at different types of school violence, so bullying, which is an area that I've studied for a number of decades, physical fighting as well as other forms of aggression and violence, again, looking at the consequences, being involved as a perpetrator or as a victim, and what we found, although there's some nuances, and I encourage you to go read that article, generally perpetration was associated with conduct disorder and delinquency whereas victimization was more strongly associated with mental health outcomes.

So when we start to think about perpetration as perhaps a signal for subsequent conduct disorder delinquency, you might argue that that is a signal that is much more where that child may need a targeted intervention, right? So you might think of a threat assessment using kind of Dewey Cornell's approach depending on the type of violence.

When you look on the victimization side and you think, "Okay.

We know that there's an association between victimization and mental health, that would suggest that screening universally could happen to identify these kids that," because the victimized students don't often come to attention, so a universal approach might be helpful in identifying those kids, but when we think about perpetration, perhaps that more targeted indicated.

And so MTSS, PBIS, this has all driven my work in bullying prevention, so when we think about the theories underlying all of that, it is behavioral expectations but a positive school climate, which might be universal but also understanding and recognizing from our findings that some kids may need more skill development such that aggression, bullying is not the way in which they manage conflicts, whereas victims may need to be encouraged to develop the skills to manage the conflicts that where they're placed in victimization.

So for me, they go hand-in-hand, Emily, but I appreciated the question because it made me think more deliberately about the work that we're doing, and to Trish's comment, which is a wonderful comment, absolutely we look at the tiered system, and we look at the inequities, the number of African American kids, say, that are in targeted interventions or have IEPs or 504s, it is in some ways the bigger idea that mirrors the inequities and structures is spot on.

>> No, that's great.

Thank you, Dorothy.

No, and actually -- so this -- I love your response because it's such a great segue to my next question for the panelists, but I did -- So Basia just indicated there didn't seem to be any audience questions, so I am just going to keep going forward, but I did want to remind the audience that if you have questions, please submit them, and we will take them as we go.

So I'm thrilled by your response, Dorothy, because it's such a perfect setup.

I didn't plan it this way, so thank you.

That was great.

So, Josh, so this is a question that I asked somewhat selfishly when I was thinking about being the moderator because I realized that I had this very naive view of these issues around school violence, and I notice that sometimes we talk about school violence and preventing school violence, and sometimes we talk about victimization, right, as part of this situation, this story, what's happening, and I pretty naively always kind of thought of, like, victimization, violence kind of like flip-sides of the same thing, right? Like, "I'm the victim of violence," or "I'm the violent person, and I victimize others," but, you know, it's way more nuanced, right? It's not just sort of like -- There's a lot to unpack there, and so the question I wanted to pose the group, and, Josh, we'll start with you, is, what does your research say if anything about whether this is an important kind of conceptual distinction between victimization and violence and sort of how you think about those constructs in this bigger story that we're trying to unpack and understand about sort of the causes and consequences of school violence.

So if you could speak to that, what I sort of as sort of a very simple, like, victim-violence relationship, what does your research tell us about this connection? >> Yeah, I mean, it's a very interesting issue.

The wider criminology literature, as you kind of noted, it's long noted the overlap between offending in violence on the one hand and then victimization on the other in that many offenders are often victims.

So the first NIJ study that we conducted, we mainly focused on offenders.

We didn't engage the victims, per se, but we did look at school and situation factors, as well as offender level characteristics, so what's possible is some of these same attributes might also access for school victimization.

What's perhaps a little bit more interesting is -- And I kind of alluded this to before in sort of listing some of those risk factors that the school shooters had.

We found that 29 percent of the adolescent school shooters, they suffered peer aggression prior to the shooting.

So what that means is they suffered some form of victimization from their peers before they actually committed to shooting, and the peer aggression variable, the way we coded it, we were including physical aggression towards the shooter, so it's they suffered an assault, somebody jumped them or somebody shot at them.

We didn't include if they were only being teased or being made fun of.

We also found that 33 percent of the adolescent school shooters, they faced a lot of family problems in the home, and sometimes this included physical or sexual abuse.

My colleagues and I, we recently completed a paper where we crafted 20 case studies in adolescent school shooters.

Now, it's not perfect because we selected cases that had more open source information on them, so they're not fully representative of all of our adolescent school shooters.

But I think the case studies that we conducted, they kind of highlight in vivid detail some of the chaotic lives that our adolescent school shooters experience, which included victimization.

Seven out of the 20 suffered peer aggression.

Six out of the 20 had physical or sexual abuse, and at least two we found in the open sources had witnessed domestic violence in the home.

But I think the takeaway, at least from our study is that some number of our adolescent school shooters faced adversity and that this often included in some cases past victimization that occurred before they committed the school shooting.

>> Thank you, Josh.

I appreciate that.

So I do see we have some audience questions that came in, but I think what I'm going to go ahead and do is finish this round of questions with the panelists on this topic, and I'll pick those up, so thank you for submitting those.

So I want to turn to Jillian now.

If you could talk a little bit about this distinction between violence, victimization and how it relates to some of your research.

>> Absolutely.

So with respect to our NIJ-funded meta-analysis which analyzed nearly 800 studies of school violence and victimization and close to 8,800 effect size estimates, we focused on 30 different predictors at the individual level -- in total, at the individual level, school level and community level, and we assessed the correlates of perpetration and victimization separately, and so what we found, first of all, was that, not surprisingly, there was evidence of an overlap between school violence and victimization, so antisocial behavior was a strong correlate of school victimization, and prior victimization was a strong correlate of school violence perpetration.

So we certainly found evidence of an overlap there where if youth are at risk for one, they may also be at risk for the other.

We also found some overlap that existed among their strongest correlates, but really stood out was that they tend to be influenced by unique sets of factors, at least this is what our meta-analysis found.

So this means that victims of school violence and those who perpetrate it are not always the same individuals.

So for instance, some of our findings suggested that LGBT youth, for instance, were at moderate risk for victimization at school but a very low risk for perpetration of school violence, and so there appears to be this vulnerable -- youth who have markers of vulnerability for victimization but not necessarily the same risk factors for perpetration, and I think that the bullying research for years has recognized these sorts of differences, acknowledging that bullies and victims of bullying can be distinct in several ways.

So for example, some of this work suggests that bullies have low empathy, that they tend to be impulsive, domineering, uncooperative, and depending on the age range that you're looking at with kids, some may even have larger social circles, right? But in contrast, victims of bullying tend to be socially anxious, submissive and withdrawn, have much fewer close friendships and maybe seem as if they just don't fit in at school.

So while there's certainly overlap between victimization and perpetration to consider, I think future work can really benefit from a greater appreciation for the various ways in which victims and perpetrators of school violence can be unique from one another, and I really do feel as though, you know, based on our research findings that violence and perpetration victimization are very qualitatively different phenomena, and I think we can reach a better understanding of both and better be able to treat and prevent both if we respect those differences.

>> No, that's great.

Thank you.

No, it's really interesting the way as you were describing that.

Like, I was really struck by, right, the overlap, but you're saying, "Yes, there's overlap, but there's some very important differences, too, and to really unpack those is -- We can't lose sight of that, as well." I appreciate that.

Okay, so, Trish, so I'd like to turn it to you, if you could speak a little bit about this victim violence distinction in your work.

>> Yeah, so in addition to what Josh and Jillian were saying, another aspect to this that we've seen and continue to see is around symbolic violence, so it can also be the case that a student, a young person, can feel threatened and feel at risk of harm even though they've never been victimized physically by someone.

So think about a student that identifies as LGBTQI.

Maybe in their community at school, around school, in the neighborhood, you know, other LGBTQI youth are being assaulted, are being harassed.

It's never happened to them, but because they belong to a group that is being assaulted, they symbolically feel afraid, right, so maybe they stop walking to school on a certain route because they want to avoid trouble.

They don't take the bus anymore.

They get to school, but they don't really speak up.

They change the way they're dressing, and eventually, maybe they disengage completely, which in their minds makes them feel safer, but now they're vulnerable to being on the street.

They're not in school anymore, and if home is not a safe place, then what do they do? In our NIJ study, our current study, one of our communities is a small rural community in the Central Valley of California, 11,000 people, and from the very first meeting with the school administration we heard about students who were worried that immigration and customs enforcement would be taking their families, so students were afraid to leave the house to come to school because there were stories about students coming home, and their parents or grandparents were gone.

And so the symbolic violence that can make people afraid, young people afraid, can often be just as detrimental in the short-term and the long-term for their engagement in school, them having the optimism, the attention to do what we want kids to do in school, right, which is to engage and have healthy peer relationships and great adult relationships.

So symbolic violence, I think sometimes we overlook that, and it's actually probably a lot more common than we think, so we're hoping in our study that's one thing we'll start to see in our data.

We still have a ways to go, but I think that's part of this complexity that you're talking about, Emily.

>> So I find that really fascinating.

This is, like -- That's a new term for me, symbolic violence, so I'm really trying to wrap my head around it, and at first, I thought you were getting at this distinction, like, between physical and verbal abuse, right, or violence and sort of that, but I feel like that's -- You're talking about something slightly different, which is just because of who you are, like, your identity, you have this perception that you're more at risk for experiencing violence even if you haven't necessarily yourself.

>> Exactly.

>> So that's really interesting.

>> Well, and you can -- I mean, you can think about it in terms of women.

I mean, this is really common all around the world with gender-based violence.

You know, if a woman is walking down the street by herself at night and there's a man, you know, behind her, she's going to have a much higher level of fear than if a man is following a man, right? So it's just -- It's a really simple kind of comparison, but that's kind of what we're talking about.

>> No, that's really interesting.

Thank you.

No, that's -- Yeah, I think that's a really -- and it really -- It says some important stuff about that victim violence distinction, absolutely.

So, okay, so, Dorothy, so turning to you now, so tell us about your thoughts on this.

>> Well, I think Jillian did a nice description of the bullying literature.

We know that the overlap between perpetration and victimization typically in K through 12 largely settings, you know, about eight percent of kids that would be identified as a perpetrator and a victim around bullying, so there's much more overlap in bullying.

We also know with many of these roles that there's more stable -- If you're a perpetrator of more physical violence, the stability of that role is higher than, say, different types of bullying, whether it's verbal, and that can go up and down is dependent on school-level characteristics.

Also, what our meta-analysis didn't do is tell you anything about what those kids were bringing to the school that Josh was just talking about, right, so we're almost doing these meta-analysis as if these are, like, blank slates and kids coming in, and then we follow them longitudinally, look at the consequences in this that we say, "Hey, perpetration of violence at school are victimization of violence at school," when we know in fact that many of these kids are what we call polyvictims and that they've experienced victimization in the home, in the community and that places them more at risk, so in some ways, it depends.

The overlap depends on the type of violence you're talking about, and the stability of those roles depends on if there's no intervention that's happening in those roles.

There's some kids that are very popular as bullies.

There's heterogeneity in those behaviors.

They're not all, you know, antisocial, and so I think you'd have to think about development, the context and what the kids are bringing to the table, whether or not there's that overlap, but certainly, there's what we call pure perpetrators, and then there's those that are perpetrating because of victimization.

Then it gets even more complicated when you think about the special education context.

Many kids are, you know, pushed into special education because of behavioral issues, when in fact we've shown this -- My former student Fred Rose has shown that kids that are perpetrating violence within special education contexts actually are doing so for protection or reacting to victimization, so that's a lot of literature, Emily.

>> Yeah, I know, right? There's a lot to unpack there, absolutely.

Absolutely.

Okay, so I'm looking at the time, and I know we have had a couple questions come in from the audience, so I'm going to pause here with the panelists and take a look at what has come in.

Looks like we have three questions.

Okay, so what I'm going to do is, I'm just going to pose these, and if any one of the panelists feels like they would like to respond or have the response that they'd like to share, just please -- I don't know what's best for our technicians, but raise your hand or jump in.

So the first one is, in terms of data collection looking at students who are placed in special education or students diagnosed with autism or students who are just in kind of more general ed students, is data on violence like that broken down by things like education status, special education status? I'm going to jump in and say I'm pretty sure yes, that absolutely we get that broken down that way, but you all can confirm that, as well, correct? >> Yeah, so I -- You know, we're in this space, too, of trying to understand violence and the unique aspects of violence for kids in special education versus those in general education, but you also have to understand that, you know, one out of four kids are undiagnosed in general ed classrooms with a disability, so that gets a little complicated.

The data are messy.

They're there.

It just really depends.

It varies by districts depending on their coding scheme and what they track and how unfortunately they implement IEPs versus 504s, so it's a messy data set, and I think we need to get on the same page of tracking data better consistently across districts and states.

[Indistinct] on that.

>> Yeah, no, Dorothy, I'm glad you added that because actually that's a very important kind of caveat to the fact that, yes, that data exists.

There's, like, lots of very complicated issues around how kids get placed into special education services and who's identified or not, and yeah, absolutely.

>> Yeah.

>> Anybody else? Anything they want to add? Somebody had -- This to me is more of a comment, but so I might address it.

Somebody asked, "So what do we do about this inverted triangle in schools with multitiered systems of support where you have a really high percentage of kids who fall into tier three and tier two and where, you know, like, a lot of kids need more than tier one?" and that is a very interesting -- That is very problematic on sorts of all different levels, and this person says, you know, "Especially at low resource schools." I mean, I think that is actually something very interesting with respect to school violence but just sort of in general if you find more kids falling at those tiers, you know.

I think at a very simplistic level it says something pretty fundamental about your tier one.

Like, you know, what are you sort of providing to everybody, and maybe you need to rethink, like, do you have enough just sort of basic, universal preventative resources in place, but I don't know if any of the panelists would like to comment on that with respect to school violence, if you sort of are dealing with schools that are low resource and a lot of the students are in a more tier two, tier three kind of space with respect to school violence issues.

>> So, I mean, for our study I think that's one of the things that we're really trying to tease out, sort of these different dimensions of risk and need in a community and how different combinations, high risk, low risk, high need, low need, sort of what that then translates to into the school environment and what the school is able to, but it's -- You know, it's quite likely that a school that is serving or situated in a high-risk, high-need community is going to both have a higher proportion of students that fall into tiers two and three and also not have the resources to be able to support those students because the broader community, right is under-resourced, which is probably why there's more kids in tier two and three to begin with.

So, you know, it seems like an obvious connection, but it is an empirical question, and it's true that schools that are in communities with low need and low risk, they still have school violence.

They still have -- So it's not, you know, this completely linear relationship, but certainly, at the high end of risk and need, there's more solutions probably in the community that need to be addressed in order for that school to have a chance with universal prevention or anything else they're doing.

>> Such a good point.

Yeah, absolutely.

>> Yeah, I'll add to that.

I think that this is a real issue for the fields right now because we have done tier one universal intervention programs to large scales of RCTs and then found that the only kids that moved on the needle of violence were the ones that were the most violent where they had room to move, so we're questioning whether or not every kid needs, say, Social Emotional Learning training, or should there be more -- Should you put your resources more into the tier two and tier three? And this is a discussion that went on a lot.

I worked a lot in Chicago public schools, and the inverted triangle when I saw this comment, I was like, "Yes, that was what we were working with," so we had to re-evaluate, what is universal, and can we re-allocate resources such that we're getting to the kids that have the greatest need versus during these universal programs where the kids are like, "Why are you teaching me how to regulate my emotions when I know how to do that, but yet a kid next to me may not but may need more than a universal program?" So I think it's really a problem for the field.

>> Yeah.

That, yeah, that is -- Sorry, I'm getting some feedback there, Dorothy.

So great.

Thank you, guys.

So I'm looking at the time, so I want to make sure we still have time to go through some more questions.

There is one question here from the audience, but I'm going to keep it to the end because it seems like it's speaking to some broader issues that I want to make sure we don't kind of short shrift at the end.

So the question I wanted to make sure we had time to get to was one that I think is really important, and it kind of gives each of you a chance to brag a little bit about your work and what you do, but it comes back to this theme of a conference, which is, you know, bridging research to practice, and how do you actually do research that can make a difference for the users of that research, for the people who are most impacted by what you're learning? And a big piece of that is this idea of dissemination, so, you know, how do you tell the world what you've learned, right? So given this money that you've gotten to do this research on the causes and consequences of school safety and violence, one thing I think it's really important to do, and one thing that I want to make sure you each have a chance to do is, like, what do you think is the most important finding coming out of your CSSI-funded work? Like, if somebody stopped you on the street and said, "Wow, Dorothy," you know, like, "I just got to know, like, if I know nothing else -- If I learn nothing else today, like, what is the, like, bottom line most important thing you want to tell me about your CSSI project?" But now going in order -- Sorry.

I know we keep this order thing going, but it helps me to stay organized.

So, Josh, if you could tell us a little bit about, like, what is, like, this really critical finding that you feel like we've got to know and understand if we're going to really advance this larger kind of, you know, concern about really taking research like this and turning it into something that we can act upon and make changes with? >> Yeah, I mean, I think that our most important finding is that the school shootings we identified, they really include desperate acts that are occurring on the school grounds.

You know, so as I mentioned before, we have intentional shootings.

We have accidental discharges.

We have self-harm and suicide.

We even have a few justifiable shootings in our data, but what's probably more interesting is, when you look at the offenders, we have youths and adults.

We have current students and nonstudents.

Right, so a lot of the shootings that get a lot of attention, like Columbine where you have current students, you know, committing mass murder, they're pretty atypical to a lot of the shootings that we actually see.

When we look at the characteristics, some of our shootings are occurring during the day when school is in session, but more than half the shootings are actually occurring when school is not in session, so it's going to be at night.

It's going to be before school starts in the morning or during summer and winter break.

Some shootings are very clear school-related.

You have a school motive.

You have a current student that is an offender and is a victim.

But many involve nonstudents as offenders and victims, and you have nonschool-related motives, like domestic violence or drugs or gang-related, and they're occurring outside the school building in a yard or a parking lot.

So I think the takeaway is, is that a one-size-fits-all approach may not work.

You know, to respond effectively to self-harm might require a different policy possibly than an intentional shooting.

But I think it's also important to realize that we have the shootings that are clearly school-related, and then we have other shootings that really just occurring on the school grounds, and in many ways, those shootings occurring at night involving gang members, you know, in a parking lot or a yard, they're really manifestations of community-level problems that just happen to occur on the school grounds.

So when you think about responses, a response to those shootings at night by, let's say, you know, nonstudents, that might be better responded to by a community-school partnership dealing with larger social issues, right, to respond more globally whereas shootings that involve current students in a classroom during school hours, you might want to have more focused on the school structure itself.

So, in the future, we really hope to develop a formal typology where we could try and develop more refined and targeted policy interventions.

>> Thank you, Josh.

That's really interesting.

I know when we talked about this earlier, I was really struck by that finding, right? Like, so much of it is just not really connected to the school at all, so that is really interesting.

All right.

So, Jillian, I want to pose the same question to you.

What do you think is the biggest, most important takeaway from your work? >> Well, aside from, I think, in doing a meta-analysis like this that's so large, I mean, the key findings are, okay, what's the strongest predictors of school violence, and what's the weakest? And, for us, what we found -- I mean, the strongest predictors of school violence were antisocial behavior, deviant peer influences, antisocial attitudes, victimization and peer rejection.

And for school victimization it was prior victimization, social competence, antisocial behavior and peer rejection.

But what our meta-analysis also showed was the really weak predictors, right, so things like extracurricular activities and students' routine activities on school grounds and school security devices were among the weakest predictors in the meta-analysis, but I mean, those are key findings, but I think what really stood out to me as a major implication of this work was that the etiology or the nature of school violence may be different from street violence given that several traditional risk factors for violence in the streets and in the community did not perform strongly in the school context.

So demographic correlates to school violence, at least in this meta-analysis, were different, were things like race, sex and socio and economic status were not strongly related to perpetration or victimization, and major theoretical predictors of offending and victimization in the committees like opportunity and self-control perspectives just did not generalize well to the school setting.

You know, across the board, the effects for self-control and social bonds were modest.

Routine activity factors were virtually irrelevant, and even risky lifestyle indicators that are such strong predictors of youth for victimization in the community were weak predictors of victimization at school.

And so, alternatively, you know, predictors such as peer rejection and social confidence seem to hold much more promise, and these are components of, you know, broader developmental perspectives that emphasize peer hierarchies, social status and social vulnerability, and so I think the key implication for my work is, you know to really urge scholars to better incorporate peer and social dynamics into the study of bonds and victimization at school, into their intervention efforts, as well.

You know, and to also be mindful of the ways in which school violence may be unique from violence that occurs in the community.

Obviously, you know, to build off what Josh said, you know, there is an extension of violence that occurs in schools that is an offshoot of violence in the community, and so we don't want to neglect that that, you know, overlap exists, but I think what was a key finding from our broader work is that there are elements of school violence that are unique from youth violence generally, and we need to be mindful of those factors.

>> Thank you so much, Jillian.

That's really great.

All right.

So, Trish, going to pose the same question to you.

>> Yeah, so I think this is such an important question.

You ask such great questions, Emily.

And I've really been thinking about sort of moving past dissemination and research to practice to engineering, so how do our results -- How can they be specifically engineered into a specific thing that this specific district and that specific district can use, so I've really been thinking about this a lot, and I think for our study, which we still have time to go, so we don't have our final results yet, but for our study, I think understanding this readiness factor, this -- you know, the timing for a student to be on the same page as their peer, as their parent, as a teacher, as a community member, to actually work in a comprehensive way to think about safety from these multiple perspectives of social, emotional and physical, and our study really was prompted by experiences we had had on our team working with really different types of districts that -- You know, one district was really innovative from a research perspective and had removed suspension as a consequence completely, so if a student acted out, and teachers had this matrix of consequences, things they could do, and the district took suspension off.

And they spent a couple of years, like, really figuring it out, but when they unveiled it to teachers and parents, they realized they hadn't really engaged teachers and parents in that conversation to the extent they should have, so teachers felt like options were being taken off the table.

Parents were calling to complain because, "Why didn't you," you know, "suspend that student who, like, punched my kid?" So it really just spoke to this idea of people being ready to do different things at different times, so talking about the solutions, right? Another example is, you know, after the Parkland shooting.

I mean, look at those kids! Those students mobilized a national network of students to walk out of schools.

But in terms of engineering an actual change, they ran into problems because policy makers weren't ready to pass laws, right? Maybe the funding streams weren't there, so I think our study -- The promise of our study beyond looking at the root causes and the interaction between community and school that Jillian was just talking about, but the solutions, right, like engineering solutions and being able to maybe not have a typology but at least be able to say, "Hey, if you want to talk with business leaders in your community about school safety, you need to think about it from their perspective of how they are impacted." If they see that shooting that Josh was talking about.

They see that on the news.

It happened in the parking lot of the school.

Guess what? They associate it with the school whether it was associated or not, right? So you have to think about it from the business leader's perspective if you're going to engage them in the solution, and so I'm hopeful that our study can help districts engineer their solutions to engage these different stakeholders from students all the way through the different levels of the community that need to be involved if you're going to take a comprehensive approach to this issue.

>> So interesting.

You keep giving me these new ideas, Trish, this idea of engineering solutions.

That's good terminology.

I think that's a good one.

So, here, so, Dorothy, I want to give you a chance to respond to this, your most important kind of takeaway, finding, message from all this.

>> Yeah, so when you do meta-analysis, you feel exhausted, but then you feel so wonderful when it's done, and I felt that way with this one.

And the reason is, we did this longitudinal meta-analysis because the research literature is all over the place, largely cross-sectional, school violence, what are the consequences of school violence, and so being able to do this meta-analysis and say, "Okay, those kids that perpetrate school violence, that is a strong association with later delinquency, support some programming to address that or identifying those kids to try to get them off of that trajectory." On the victimization side, you know, we've known for a long time that being victimized through bullying, chronic peer victimization, that that's a potent risk factor for mental health outcomes.

Here we found the same thing, that victimization was associated with depression, and why is that important? It's important because the youth suicide rate is the highest it's ever been in ages 10 through 14, especially among certain demographics, and so for me, I think it just summarizes that we need to continue to chase programming that would be effective at the school level and really giving victims of school violence the strategies to destigmatize help-seeking, to identify trusted adults, so I think it just kind of confirms that we need to be doing the things we're doing around Social Emotional Learning, around building a positive school climate, and then also, I think we need to think about schools and recognize, this is what Jillian is saying, that there are some schools where violence is adaptive, where violence is perpetrated by some really popular kids, and so we have to understand the peer dynamic and how the school climate, the behaviors of the teachers are impacting this, as well.

So this really -- This meta-analysis needs to be followed up with a bigger one, and maybe Jillian is doing that, to look at the extent to which the consequences can be tied to these antecedents that were demonstrated in the larger meta-analysis that Jillian talked about.

>> Excellent.

No, thank you.

That's really helpful because I was thinking about that with your meta-analysis, sort of the -- I think of it as, like, this ability to be very precise, right, about the -- take this cross-sectional work and really think through the time line, which is so, so critical.

So this brings us to the final question that I want to pose to the group, and it's one I really debated about whether I would even ask it because it's sort of cliched at this point.

But it's sort of like the elephant in the room, and it just seems so relevant to the topic at hand, which is school violence, so school-based violence, school-based victimization, what's happening in schools.

And the question is this: We have all been impacted by the coronavirus pandemic in so many different ways but schools especially, right, with remote learning, people at home.

You know, it has just been disruptive on so many levels.

And so, obviously, it's posing challenges for your research even if you're doing a meta-analysis, right, even if you're looking at work that's been done.

You're being affected by the coronavirus pandemic.

But my question for all of you is, because so much of your work is, like, really trying to understand what is unique about violence that happens at school, how has the coronavirus pandemic challenged the ways you even think about school as a setting for violence? And this is kind of a big abstract question, but, like, I would love to hear some of your thoughts about this and sort of what kinds of -- kind of thorny intellectual challenges this is sort of creating as you're thinking about your work and this issue of, like, what is unique about violence that happens at schools? So, Josh, I'm going to pose this question to you and get your thoughts on that.

>> Yeah, I mean, it really is a great question.

I really -- I just wanted to first touch upon something that Trish had said about even if the motivation is not school-related, it could still be associated with the school, and it's such a clever comment because part of our inclusion criteria was the shooting had to occur on K-12 school grounds.

So if a shooting just incurred right in front of the school, but it wasn't technically on school grounds, we excluded it, and we went through, you know, well over a thousand cases.

We probably ended up excluding, like, five to 700 cases that we considered for inclusion, and a lot of them just did not occur on the school grounds.

But their portrayal in the open sources invariably went into the school, you know, right in front of the school, at the corner of the school.

It was repeated, you know, countless times in the article.

So I think it's a point very well taken that, to the average person or to the average businessperson, just reading about that shooting, they would have associated it with the school, regardless of where it occurred.

You know, in terms of the pandemic, it's such a, you know, a great question because obviously, you know, the move to online teaching, they're not going to be on school grounds, right? So, you know, it would impact our inclusion criteria.

But, you know, it just got me thinking.

You know, how could -- When you think about the pandemic more globally and the responses to it, how would it impact some of the findings that we have? You know, and I think that when you look at some of our school shooters, the adolescents who had such chaotic lives and which suffered, you know, personal adversity, and some of those dealt with problems in the family.

You know, it makes me wonder that if they're not going to school, if the school is closed, and it's moved to an online platform, and if they're spending more time at home with their family, whether that magnifies any of their adversity and how that would play into students like them in terms of future school shootings.

And then when we think about those shootings, more than half of the intentional shootings that are less school-related, you know, and more community problems that occur on school grounds, again, if you think about the lockdowns and some other policies, you know, if they're leading to more free-time and maybe increasing peer interactions, how does that play into possibly magnifying some of those community problems that do tend to spill onto the school grounds? You know, so I guess in terms of the research that Steve Chermak and Nadine Connell and I have conducted, you know, it's kind of an open question if and how the pandemic and the responses to it has magnified any of the risk factors either individually or in the community and the adversity that many of these adolescent school shooters face both at home and inside their communities.

>> Thank you, Josh.

So that's really interesting to think about it as sort of like this magnifier, right, as sort of what you've -- these relationships that you're learning about.

So, Jillian, what are your thoughts about this? >> When I think about school violence, I don't necessarily only think about school shootings and, you know, the most serious ones of weapon-based offenses.

I think, too, about aggression, and when you think a lot about relational aggression and bullying and peer dynamics, you know, a lot of that doesn't just occur on school grounds.

It also occurs online, on social media, and is cyberbullying is a concern.

And so I'm curious about the ways in which the coronavirus has, you know, moved those dynamics and those bullying and relational aggression to an offline context and how much that has affected youth who were already being bullied in school, if they experienced a reprieve from that during coronavirus despite, you know, many other challenges, or if that bullying moved to an online setting.

So I think it would be really interesting to examine that in future work.

>> Yeah, no, definitely.

To think about sort of how all those dynamics maybe have just been transplanted into this really odd online world that we all live in now, right, just like we're right here with this virtual conference.

So, Trish, what are your thoughts about this, this school setting idea? >> Well, you know, we're trying to see this as an opportunity for our research, the pandemic, and so we've revised our survey instruments to be able to collect information from parents and students and staff around their experience, whether they're distance-learning or in hybrid situations or on campus, and a couple of reflections.

One, you know, for us, it's not always easy to have a teacher or a parent think about school safety outside of physical terms.

A teacher might be more likely to understand the social and emotional stuff if they've been trained in that, but not all teachers have been.

But people generally think about physical safety.

But the pandemic actually has opened up that definition of safety into this broader understanding about public health that I think is -- could be really both fascinating to understand and see how people might change their definition of school safety now, to think about health-related things, but also how they see each other.

So the dynamics, the relationship dynamics between groups that maybe were formed through the prior experience, right? So, you know, schools talk all the time, teachers talk all the time about how hard it is to engage parents, and especially parents who are running around with multiple jobs, are not native English speakers, right, all of these issues.

Through the pandemic, right, might we find that parents and teachers start to have increased engagement or empathy for each other in ways that they didn't have before, and maybe the opposite with students.

Maybe students are less engaged with teachers, would be a bad thing, right, in most cases.

And then with their peers, you know, to Jillian's point, I mean, I've seen research go both ways on this so far, where some adolescents feel more isolated, more prone to harassment, online bullying, and I've seen it's the exact opposite, research being reported where students are using their online networks in this coping mechanism kind of way, and their mental health has been improving, and that's a study that's been going on for the past 15 years, every 2 years.

So, you know, I think the jury is still out.

But for our study, you know, we could really learn some things about does a disruptive event like a pandemic or, you know, the Parkland shooting or the -- George Floyd's death, does a disruptive event at a social level kind of, like, disrupt relationships in such a profound way that you can remake them in a more advantageous manner? So we'll see, but, I mean, we'll at least have some data to look pre and post-COVID with engagement and climate and all that kind of stuff and the definitions of safety when you're learning at home versus learning at school versus a pandemic and not a pandemic, I mean, it's going to be fascinating to peel that onion.

>. Yeah, thank you, Trish.

So I'm looking at our time.

We have 5 minutes left.

So, Dorothy, I want to give you a chance to respond, and then if time permits, I think I see one good question in here that would be a really good one to end on.

So, Dorothy, if you could -- a little bit about this concept of school setting? >> Yeah, so our team have struggled for a long time of trying to implement school violence prevention programs.

Pushback from schools, there's not enough time, too much on the plate, testing.

So we, pre-pandemic, were trying to figure out how can we alleviate some of the burden for the programming, whether it's social emotional learning or PBIS or tiered systems or gang violence programs or these types of programs.

And so we had been working to leverage technology for a really long time.

So we had developed a text messaging program.

Most of our professional development had gone online, right? So we were trying to think, at the same time of not burdening the schools, but also recognizing that conflicts, just when the bell rings at the end of the day, it doesn't end.

They leak over into, you know, family life and to social media for these kids, and so this was an opportunity for us to then say, "Okay, how can we, given these kids are not getting social emotional learning in some cases through their Zoom meetings or Google Meetup, how can we leverage technology for that?" So now I think, moving forward, we will probably always, in our applications, think about what would happen if we have to shift to virtual learning again? So I think that just generally, it was like, "Ah, what are we going to do?" But at the same time, it's really made us think about school violence outside of the brick-and-mortar and contemplating, like, "What is Zoom climate? Like, are we having a good positive climate?" Like, there's so many questions to ask.

How do you deliver some of this programming that's been known to reduce school violence? So lots of questions, as Trish said, that remain to be answered.

But I think we always have to think, can your program pivot to online if this happens in the future? >> Dorothy, truer words could not have been spoken.

I think it has forever changed the way we think about the potential of distance learning and what that means and what you can do in the space, absolutely.

So this -- So we're getting close to the end.

There was this one question I saw that I would like to try to throw out there to the group, and maybe -- So, yeah, so we don't have a lot of time.

So let's be sensitive to that, and it's kind of a big question.

But for each of you, is there sort of one kind of, like, big outstanding question that keeps you up at night about the causes and consequences of school violence, where you're like, you know, "Wow, there's just still this one thing that, like, if only we could figure that out, that would be really the key to advancing this field?" So, Josh, do you have any thoughts about that? Is there something that really perplexes you still in this work? >> Sure.

You know, the current study that we completed, we relied upon criminology theory.

But criminology theory is really designed to explain what differentiates offenders from non-offenders.

Everybody in our study, in the first study, was a school shooter.

So Steve Chermak and Brent Klein and I, we're actually starting a new study where we're trying to gather a sample of non-offending students from the same school as school shooters and then look at some of these individual-level factors to see if, in fact, they can account for variation across the non-offending students and the school shooters.

>> Trish -- I'm sorry, Jillian.

I'm forgetting my ordering.

>> Okay, to be quick, I think for me the biggest questions surround how is school violence unique from violence in the community, and how do you who have many risk factors that you would think of as markers for school violence? They have a history of antisocial behavior.

They've been exposed to adversity and trauma.

What predicts whether they go on to commit violence at school or not? You know, I think that's a key question that we all would like to know, how to better predict and identify who will actually go on to commit violence at school, despite having many risk factors.

Because many, many students have risks, yet they never engage in violence at school.

So that's a question that I think we all are going to struggle to answer.

And in addition to that, what are the situations and contexts in which violence in schools unfolds? So those three things, really, I think are going to challenge me for many, many years.

>> Excellent.

No, thank you, that's great.

So just want to give time to Trish and to Dorothy.

So, Trish? >> Well, for me, it's just right-sizing the solution to fit what the need is within the school and within the community and being able to engage everybody in that process rather than just leaving it to the school to figure it out.

>> Excellent, thank you.

And, Dorothy, you get the last word here.

>> Oh, always.

Yeah, so I think for me, I'm really concerned about when kids do come back full-force, 100 percent, how we're going to manage the mental health issues and will some of these mental health issues that they're bringing back because of COVID, inequities, racism, then will we see an uptick in school violence, and how can we kind of get ahead of that, is a concern.

>> Wow.

Excellent.

Well, wow, okay, we are right at 3:30 on the dot.

So I just wanted to thank the panelists so much for engaging in this conversation with me.

I thought some really interesting ideas were raised.

Obviously, so much more research still to be done, and I -- So I think I'm turning it back over to Rasha? Or I'm not sure, sorry.

Here she comes.

>> Yes, I want to just give a big hand of applause to all the panel.

Thank you very much to our audience for your questions, and, Emily, thank you very much for your contribution to moderate this discussion.

We will see you at another session.

Thank you.

>> Thank you.

>> Bye-bye.

>> Thank you.