School Safety Considerations for Distinct Student Populations - Breakout Session, NIJ Virtual Conference on School Safety

>> Good afternoon.

Welcome to a breakout session of the Comprehensive School Safety Initiative Conference.

I am so glad that you could join us today for our session for School Safety Considerations for Distinct Student Populations.

My name is Barbara Tatum Kelley.

I'm a social science research analyst with the National Institute of Justice, and my primary interest areas are in that of adolescent development and its relationship to victimization, delinquency prevention, Juvenile Justice System improvements, mental health, youth gangs, and education.

So there are many factors that come to play in the development of the youth, and this session offers a fascinating set of research projects, all of which examines equity and disparity and discipline and school safety across distinct student populations.

In this session, we will have four presentations, and then with the remaining time at the end of the session, we will address questions and answers from our attendees.

Please enter your questions into the Q&A section of the program -- I think you're go getting to be pros at this -- so that we can see them when we get to the Q&A time.

So, without further ado, I would like to introduce our first presentation.

It will be made by Nicholas, or better known as Nick Read, and Melissa Ripepi, who will share their insights on distinctions between urban, rural, and suburban school districts.

Specifically, are their school climate safety disparities? Are their distinctions in terms of how schools are embedded in the community health content? So, I'd like to welcome Nick and Melissa, and I can't wait to hear your presentations.

>> Great.

Hopefully everyone can hear me okay.

All right.

I'm going to dive right in here.

I'm going to be setting the stage, talking a little bit about the ReSOLV study, Research on Lowering Violence in Communities and Schools, a very little bit about the study as a whole and turning things over to any colleague, Melissa, who is going to dive more deeply into one lens, through which we're conducting the study.

As Barbara said, we're looking particularly at school climate safety and inequality, kind of setting the stage for a range of student populations by drilling down into the context of place and community.

As I said, I won't be covering a lot about ReSOLV specifically, but we do have a great overview video that I encourage you to check out on YouTube.

The link is here.

We can make sure that's disseminated to those that want it.

But, yeah, that will give you much more information about the study as a whole than I will have time to do today.

But the quick version of it, you know, as my colleague and principal investigator, Trish Campie, talked about yesterday, this study is looking at the root causes of violence, school safety concerns among students, the school, the community, really, the totality that makes up school.

It's not just the building itself, as we all know.

Specifically, you know, we're looking at the element of readiness, school and staff readiness, the individuals within that, the whole and their readiness to recognize and counteract those root causes of violence.

As we all know, school settings, both amplify and attenuate those risks for school violence, and at the end of the day, we're interested in how all of this impacts school safety and youth outcomes.

So, more specifically looking at our range of research questions, four main questions kind of guide the study.

We're looking at the ecology of risk and protective factors within schools that influence these outcomes.

We're looking specifically at readiness and the readiness to mitigate those risk factors and how that's associated with improved outcomes.

Specifically, we're looking at the core components of school, individual, and community readiness -- we'll talk a bit about this a little further -- to mitigate those risk factors.

And then how many associations between readiness and school safety and student outcomes, how do they vary over time? Of course, we're not immune to the world that we operate in and we're certainly in a different place than we were when we started the study three years ago, and where were February 2020, so we've kind of adapted some of our research focus and sub-questions around the post-COVID era.

Of course, this is inclusive of not just the pandemic but the period of political unrest, social unrest that our communities find themselves experiencing alongside of us.

So, looking and shifting a little bit, saying, you know, how do our students, families, and stakeholders, how do they think about school safety in the context of the COVID pandemic.

What does student safety even look like when students are learning at home and participating in hybrid learning models, and then what is the specific role of schools right now in relation to community wellbeing, and vice versa.

And, again, how do different risk and need factors among the populations that we're studying how do those impact the thoughts around this time we're living in? So, I mentioned readiness.

It really is a key component of our study.

I cannot do this anywhere the justice that our colleagues can.

I'm extremely thankful to our partners at the Wandersman Center and the GAMM Course Group.

I really have championed this notion of readiness and their readiness equation, which is readiness equals motivation, innovation-specific capacity, and general capacity, or R equals MC squared.

I urge you to check out those groups for far more information on readiness.

But for our study, we're looking at the difference factors associated with the high readiness-low readiness among the groups that we're looking at, the individual students, staff, parents, community members, and specifically interested in the factors of readiness, are they different within each of these groups? What do they mean for a student to be ready versus a whole district to be ready? And a big question for us is, do these groups have to be ready at the same time? If we're talking about implementing a comprehensive school safety approach, what does readiness mean? And if one group is ready and another is not, what are the potential consequences of that.

So, the districts we're working in, we're specifically in California, as proposed for the study.

We're working in the Mendota School District.

Mendota is a small rural city in Fresno County.

We are working in one small part of the Los Angeles Unified School district, their local district south, and then as well as the Hanford Unified High School District.

Hanford is the county seat of King's County.

The districts themselves really, by design, were to be one rural, which is Mendota, small rural county, again, in Fresno, a small rural city in Fresno, about 11,000 individuals.

Ironically enough, our suburban district is actually the south region in L.A., with suburban characteristics, and our urban district is Hanford, which is in, as I said, the county seat of King's County.

We specifically wanted to look three distinct settings for the study.

The study is largely relying on a pocket of extant data.

Right now, we're focusing on seven years of data from three years before the study started to three years after.

We're hoping to extend this work, given the delays and what school looks like in the 2023, but we're collected extant data around school and student administrative data, student achievements, student outcomes, school discipline, things like that.

We have a robust crime dataset through the California Department of Justice, which includes both individual crime and crime incident-level crime.

We're extremely fortunate to work with the department directly and get their Criminal Offender Record Information System data around not only the incidents themselves but also criminal histories for anyone that's arrested during the study period.

And then we're using census data through the American Community Survey, as well as a range of sources of social determinants of health data and other pockets of extant data, just kind of fill in our direct research, which involves three years, and, again, hoping to extend this direct, direct student, parent, and staff surveys, building off the California Healthy Kids and Healthy School surveys, as well as the district taking on some of the surveys we're developed directly.

We conducted and will continue to conduct student focus groups, as well as parent, staff, and community stakeholder interviews.

And then we're doing some community mapping of places and spaces, and I'll touch on that very, very quickly, and you'll see why we're looking at that when Melissa takes over.

So, as I mentioned, we're looking at the relationship between school safety, readiness with school level of student outcomes across various time periods.

And if that's the playground of the study, really, you know, the sandbox that Melissa are in are really in the realm of moderators.

The moderating factors of community risk in need, and, by that, we're looking specifically at community disadvantage, community crime, and as we use social determinants of health.

So, again, just kind of reflected back on the specific research questions of focus group and this presentation in our group, it's looking at that range of risk and protective factors within schools, and looking at the impact of the readiness of individuals, of the school and the community, to mitigate those risk factors and what are the associated outcomes.

As I mentioned with the districts that we selected, we also wanted to find, within those districts, the continuum of risk and need.

And you see there, very basic, we want to find those pockets of neighborhoods that represent the A's, B's, C's and D's here, those that are high risk-high need, low risk-lower need, and everything in between.

And just to illustrate that quickly, this is some of the mapping that I've talked about.

These are not our maps, but we've kind of modeled some of what we trying to do.

We want to start to overlay some of our datapoints, whether it be crime or social determinants of health, community or concentrated disadvantage, and literally lay those on the map so that we can start to see graphically those differences and those similarities that exist.

We're not comparing the three districts.

They will each be their own case study.

But within those districts, we want to literally map out kind of not only the contextual factors of crime and social determinants of health, but, also, those specific places and spaces, and Melissa will talk a little bit about this, you know, where are hospitals? Where are playgrounds? Where are hospitals? Where are abandoned houses? And just get that contextual feel through the geography.

So, just a couple of points to kind of illustrate where we're going with this.

This is 2016 crime data, but we've continued to update this.

Here, you'll see our three school districts at the city level.

And what might be surprising to you here is when you're comparing a small city like Mendota, of a little over 11,000, with all of Los Angeles, over four million people, Mendota, per capita, has a higher crime rate, and you can see it's almost approaching double of what Hanford looks like.

And while crime is kind of blunt instrument to look at context, it was part of our proposed study to NIJ to look at the impact of high-crime areas, low-crime areas, risk factors and protective factors.

So this just gives you an initial glimpse of the range of these three districts.

And, of course, we want to dive deeper into each one of those districts and look more specifically at communities and neighborhoods and not so just at the city level.

One example of that is our community disadvantage index that we have.

We see here, these are three of the three census tracks, U.S. Census tracks that make up Mendota.

And you can see here across the five metrics for concentrated disadvantage, which is poverty, public assistance, female-headed households, unemployment, and individuals younger than 18, looking at the Z score shaded here, there's a range within each of those or across the three census tracks here.

And if you look at the final column, looking at the overall index score for concentrated disadvantage, there's quite a range even in this very small geographic area.

So, all that's to say is, there is not one Mendota that we're looking at.

There's at least two, three, four, seven different Mendotas that we're really taking into context and thinking about; okay? What does school safety look like in the context of the individual communities that make up these districts? So hopefully I've left Melissa some good time her.

She's going to advise of her work and our work around social determinants of health and how this particular lens is really kind of shifting some of the ways we're thinking on the study and really providing a detailed context to all of the kind of blunt instruments, as I've said, whether it's crime data or disadvantage.

I'll turn things over to her.

>> Thanks, Nick.

You can go back to the other slide.

So, the social determinants of health, the standard definition are conditions in the environment in where people live, learn, work, play, worship, and age that affect a wide range of health and quality of life, outcomes and the rest.

So, according to the World Health Organization, this unequal distribution of health damaging experiences is not, in any sense, a natural phenomenon.

Rather, it's a result of the combination of structural inequalities that are imposed on populations of people through social, political, and economic institutions of a place, so it's very place specific.

You can go to the next slide.

The invisible force sort of connects structural forces to biology and stress.

But it's a specific experience of stress that's elevated and sustained.

So, these terms, tolerable, toxic, and one that's missing, positive stress response, were developed by the National Scientific Council on the Developing Child in a study they undertook to understand how the experience of adversity in early childhood affects the development of children's brains and bodies.

But the classifications can be used to understand the kinds of stress that makes the population vulnerable to health outcomes at all ages.

So, the first two, I describe sort of normal healthy responses to stress.

But the last one, toxic stress response is what we're concerned way.

It comes from exposure to strong frequent prolonged adversity, you know, a constant state of uncertainty, in which safe spaces are difficult to access.

The stress response or negative impacts are really triggered through the social determinants of health.

So next slide.

So, this is where I want to draw a distinction between determinants and health indicators.

So, the social determinants fall under these broad categories of social inequities, institutional inequities, and living conditions, and these are what sort of work as delivery systems for toxic stress.

So, changes that impact these determinants can then, in turn, impact help in positive or negative ways downstream.

Next slide.

Toxic stress that manifests as a result of determinants to physiological responses that can interfere with memory function, the ability one has to control impulses.

It can increase anxiety, arousal, aggression.

It can disrupt the ability to express pleasure.

It can wreak havoc on hormonal systems, cause the immune system to malfunction, suppressing in some ways, making it overactive in others.

So, through these physiological responses, stress has been embodied as obesity, diabetes, and susceptibility to communicable diseases like the flu our COVID-19, autoimmune disease, high-risk behaviors, acts of violence, heart disease, and so on.

So, the important thing here is that these are also described as upstream and downstream, but, really, it's more of a circular connection.

Poor health outcomes and high-risk behaviors further intensifies the experience of upstream determinants, which, in turn, exacerbate downstream outcomes.

Next slide.

So, filling in the gaps, this is really a framework that's an ecological approach to understanding health.

But the important thing is that the undercurrent of it is really the experience of being safe, of being able to rely on a safety net so trust can be cultivated within a community or a school treatment or prevention of a health outcome or indicator, really, within this framework prior to treatment of inequalities and experiences of trauma that often accompany them.

So, I saw recently saw that all data is health data when looking at social determinants of health, which is kind of true.

But it really encompasses a lot of the focus and metrics that are being used within our project results, including community crime and safety, concentrated disadvantage is folded into this, as well.

It's sometimes considered thought to be a social determinant of health, and it's often used with these studies to help identify socially vulnerable or isolated populations.

School readiness has been identified as a central indicator in child and adolescent wellbeing by the Lufkin Healthy California Task Force.

So the determinants can really be function-driven connected between all the disparate areas of our work, which leads to questions our group hopes to answer through this data.

Next slide.

So, these first three questions tie back to the previous slide.

The interactions and the bridge will focus on health indicators and barometers, the structural inequalities, and the embodiment of trauma, and as indicators for the presence of buffers, which are equally important to name and highlight.

So, focusing on one health outcome will unpack this sort of question, sort of opening up making new relevance of social determinants of health to our work, focusing on data.

Next slide.

So, in research on extant health data, a step that really stood out for both of these places was Asthma emergency department admissions.

So, the numbers you see here specifically represent emergency department visits due to asthma, reported is the age-adjusted rate for 0,00 people who seek treatment in an emergency department.

So, although this is for all admissions, regardless of age, it's moderately safe to assume that the vast majority of cases are kids, for two reasons; one, the rates in these counties for young and older adults are very low; and, two, Fresno County is second only to San Francisco in the rate of emergency department visits due to asthma in children age 2 to 17.

So, anyway, I thought this was a good opportunity to sort of unpack the determinants of this particular health condition.

Next slide.

So, you'll see, these physical characteristics of our environment that you see here are sort of what are commonly thought of when people think of what contributes to asthma.

But they're really far from the whole story.

To quote from a pediatrics article that I referenced here at the bottom of the slide, the social patterning of asthma reflects different exposures to pathogenic factors in these physical and social environments.

These physical factors represent chronic and acute stress that comes from structural inequalities, experiences of discrimination, uncertainty, lack of control.

Anxiety is a known pathogenic factor for asthma.

So, what issues should this raise for us? It can point to issues related to economic insecurity, housing standards, outdoor recreational opportunities.

Potentially, the experience of stress is related to documentation status for an individual or family member.

You know, issues of food security, access to nutritious food, obesity, and/or inadequate intake of dietary antioxidants are both very good risk factors for asthma.

Asthma is somewhat easily controlled with consistent high-quality primary care, so high rates of emergency department visits suggest a lack of access to this.

Similarly, studies have shown that the strongest predictive factor for a child's hospitalization is the caretakers mental health, which, again, suggests enormous amount of stress and lack of access to mental health care.

So, lastly here, exposure to violence can directly impact morbidity rates of asthma.

To quote again from this article, crime rates of violence and crime within a community in society are not only chronic psychosocial stressors but also indicators of complements, collective wellbeing and nonoptimal social relationships, all which can contribute to the onset of asthma, or the severity of it.

So, for asthma in particular, but also for social determinants of health more generally, the presence of collective efficacy of social capital, which communities have been shown to be a protective against severity of disease.

So, examples of this are facilitation of health education and communication, elimination of environmental health hazards through collective action, or the reduction of the experience of stress through social support networks.

So, next slide.

Getting close to wrapping it up.

To go back to our questions regarding how we're going to use social determinants of health, those three questions that I didn't earlier address, how do we think about public health interacting with school and community readiness? What do we think about the bridge between public health and school safety? What does it mean to be safe? There are a number of ways that we can look at this.

If you think of asthma as a potential indicator of physical, social, emotional trauma that manifests through toxic stress, it can present through disease presence.

It can also co-present through learning and behavioral challenges that tie back to physiological interruptions that I presented earlier in the presentation, so memory function, impulse control, increased anxiety, maybe increase in aggression.

So, negative experiences in school can feed the cycle.

But, also, positive experiences, through collective advocacy, social capital, treating trauma rather than punishing its symptoms can potentially disrupt a cycle.

And, sect, just to speak further on the buffers, their presence or absence are really, potentially, telling the readiness capacity of the school or a community.

So, next slide.

We're gathering data here, so these are through the extant data gaps.

Our work is gathering data on social capital and institutional trust through the work that we're doing directly with, and within communities for school surveys, interviews, focus groups, and naturalistic evidence.

And the second of these two are or the second two mostly available.

With these, we're working on pinpointing exactly how we can use this tool for better understanding school safety.

So this is a really large, potentially, all-encompassing framework.

We're doing various place-specific work in each of the communities to determine which health indicators in social determinants will have the most impact, potentially, in helping us understand readiness dynamics.

[ Distorted audio ] >> Hi. Melissa is frozen on my screen.

Nick, did you want to wrap? >> Yeah.

I know we've packed a lot in, and we need to move on, so would encourage anyone that's interested to check out more.

Let us know if you have any questions.

Happy to share a big study over a long period of time, but this is just one glimpse on how we're contextualizing school safety outside of the school building itself.

>> Well, very big thanks to both Nick and Melissa for that presentation.

Now we're going to move with our second presentation by Lucy Sorensen and Shawn Bushway, and they're going to share their findings on the administration of discipline by principals in and school resource officer, with an emphasis on disparate impacts by race.

So, thank you very much, and welcome to Lucy and Shawn.

>> Thank you so much.

We're excited to be here.

As she said, I am Lucy Sorensen, and my colleague, Shawn Bushway, is also here and happy to answer any questions, although he's allowing me to do most of the talking today.

So, today we'll be presenting findings from two studies that have come out of our NIJ-funded projects through the Comprehensive School Safety Initiative, and the title of our talk is, "Different Disciplinarians in Schools: The Impact of SROs and Principals on School Safety and Student Outcomes." Okay, so for background, this is probably familiar to most of you in the room.

There have been very active debates in the area of school safety and school discipline over the past several decades.

Kind of culminating in this past summer, Minneapolis and many other school districts choosing to terminate their agreements with -- their partnerships with police departments, and a lot of this has come out of concerns that putting police in schools might overly criminalize minor acts of student misconduct, and it might particularly affect black students in schools.

But it's also linked more broadly to concerns is that folks have over school disciplinary practices and the consequences of those practices for students.

So, at the same time that all of these concerns about, potentially, adverse consequences from school discipline or school policing were also having very real fears about violence in schools, and, in particular, fears that, without possibly the disciplinary tool suspension or without having police officers in schools, that this could leave schools more vulnerable to violence.

Okay, so against this backdrop, we're going to motivate our two studies by kind of a simple pu.

le, which is, which do you think would have a larger impact on student outcomes, adding a police officer or school resource officer into a school or switching a principal from one who acts in more lenient ways to a principal who acts in more harsh ways? We're thinking about these not as isolated forms of exposure; right? Not just how is the single exposure to a police officer affecting students, but really thinking about these personnel as actors in a much more complex system; right? So, if you're thinking about the disciplinary process, it's not a single decision that gets made by a single actor, but it's really these multiple different actors -- teachers, principals, administrators, police -- all in schools interacting with students and making different discretionary decisions at different points in time, and we're really interested in what outcomes those lead to and how to improve the system.

Okay, so what we actually do in our study, we're working in the context of North Carolina, and in particular, North Carolina middle schools, and we've been able to link over a decade-worth of administrative data from the state, where we have disciplinary referral data, we have school records for every student in the entire state, and then we've matched these to juvenile justice court record, as well as adult conviction records from the state, so we have this nice linked longitudinal data study some of these questions.

And in particular, our research questions are, we want to study the causal effects of school personnel, with the first being police and the second being principals, and their practices on school safety, on disciplinary consequences, but then also on these more abstract long-term outcomes in education in criminal justice.

And, in line with the session, which is on distinct populations, we want to pay particular attention to whether we observed differential treatment from school personnel by race, and also differential effects of the school personnel by race.

Okay, so looking at the first study, where we are studying the effects of school resource officers, we make use of the fact that in North Carolina, between 2005 and 2009, there was this staggered rollout of school resource officers into middle schools in the state.

And so, because of this, we're able to use within school research design to compare different metrics that we care about, maybe before a school gains an SRO and then after they gain the SRO, or maybe even before they lose an SRO and after they lose an SRO.

So, jumping right into our findings from this study, one of the primary findings is we find that SROs, when they're in schools, they decrease the incidents of serious violence by 38 percent, and so this includes things like assault on school personnel, assault resulting in serious injury, assault involving a weapon or a firearm, and so they're fairly rare events, but, on average, an SRO reduces the incidence of this violence about a third of an incident per school per year.

We found no effects on weapon possession or alcohol and drug possession, no effects on minor attending by students, no effects on the variety of different text score and educational outcomes we looked at, and no kind of lasting effects into the adult criminal justice system.

We did also find that if you control for or if you hold constant the changes that SROs made in student offending, so if you hold constant that change and serious violence for instance, SROs also effect outcomes in the disciplinary system and referrals to juvenile justice.

So you can see here that SROs decreased felony complaints in the juvenile system by about 10 percent -- sorry, they decreased.

Increased misdemeanor complaints by about 50 percent.

They also increased the use of long-term suspensions in the school by over 60 percent, and increased reports to law enforcement by a large amount.

And I do want to say we're very cautious about this last outcome, because report to law enforcement does not necessarily mean report externally.

It could also just be a teacher telling the SRO at their school that has something has happened, and so we can't really determine whether or not that led to, for instance, an arrest.

And so, when we look at differential effects across different groups of students, we did see some differences, so this reduction in serious violence that we observed at school was predominantly being driven by a reduction in violent events committed by white students, and then the increased use of long-material suspensions in role of law enforcement, that was predominantly being driven by increases in black students.

We did not find any disparate outcomes in juvenile justice referrals.

So, overall, for this first study, we conclude the SROs are preventing violent, at least in the context of North Carolina middle schools during this period of time.

This is consistent with some prior research, but inconsistent with other prior research, but at least it's promising evidence that SROs are meeting one of their stated objectives in North Carolina middle schools.

We also found that, even though SROs are not allowed to give suspensions themselves, and that's not a formal part of their job description, just having an SRO in a school does seem to affect disciplinary processes and outcomes in significant ways.

And it looks like it may do so in ways that actually exacerbate racial inequality and punishment.

So, in our study, we discussed quite a bit about this question of, where SROs are in schools, they're often operating a community policing model, where they're interacting regularly with students.

They're trying to take more preventative approaches, more proactive approaches.

And so even if we see worse disciplinary outcomes with SROs in schools, there's this question of whether it might still be better or worse than more reactive policing where they come in from an outside event.

Okay, so our second study that we are going to talk about today, we're using the same form of data from North Carolina, this time from 2008 to 2016, so we have over 2.5 million student-year observations.

And in this study, we make use of the fact that we have not only data on suspensions but also on each of the disciplinary referrals, so we can see both what is referred to principals and then what, ultimately, decision they make.

And so, the first thing we ask is whether the principals in our sample differ or vary in what we call their propensity to remove students for the exact same offense type and the exact same offense history.

And what we find is that they definitely do.

There is wide variance in how principals respond to even the exact same type of student offense.

Some are much more likely to suspend students than others for the same offense type and history.

And it's not only across schools that we see difference, but we see large differences even within schools.

So, when there's principal turnover within a school, we often see that there is a visible change in how they are handling disciplinary matters.

So our second question is, within a single principal, do we see that they're treating students of different races differently, and, again, we see that, yes, this is the case.

Although principals vary quite a bit, the average principal in our sample is six percentage points more likely to suspend a black student for the exact same offense type and the exact same offense history as a white student, but, again, there is quite a bit of variance across schools.

And so our last question is, kind of what are the consequences of these discretionary principal behaviors for students? What does it mean if you're a student in a school under one of these principals that's more harsh or one of these principals that's more lenient in general? And so we'll go through some of those results.

Okay, so the first result is that we find that harsher principals, when a harsher principal is introduced into a school, so we're looking at within school, again, based on principal turnover events, when we're switching from a more lenient principal to a harsher principal, we do see a reduction in minor acts of misbehavior by students.

So these are things like disruptive behavior, disrespect of faculty or staff, being late for class, cutting class, property damage.

And so we see a moderate reduction in these minor offense types, but we don't see any similar reduction in serious offense types.

And then could be a deterrence effect, so the harsher principals are scaring students into not misbehaving, or it could be that, actually, teachers are practicing different reporting behaviors under different principals.

Second, we find that within the school, when it transitions from a more lenient principal to a harsher principal, there are causal effects on the increased juvenile justice complaints and decreased eventual high school graduation rates.

There are very real and significant consequences kind of immediately and more long term.

But within the school student population, we found that there is a subset of students that were particularly affected by principal disciplinary behaviors, and those were the students who commit minor offenses at some point during the school year.

We think it might be because these students who commit minor offenses are kind of on the margin of being suspended or not suspended, and so being under a more harsh principal will have an undue influence on them.

We find for this subset of students who commit minor disciplinary offenses, they have increased absenteeism, increased likelihood of grade retention, reduction in test scores, and, ultimately, a much lower likelihood of graduating high school.

We also looked at the consequences of differences in degree of racial bias that principals exhibit.

We found that if you're in a school with a more racially biased principal, it has an effect on later educational outcomes for students of different races.

So, more racially bias principals lead to a widening of black/white gaps in attendance, in test scores, in grade retention, and in high school graduation.

So there are real educational consequences for this.

And so, in conclusion, for this study, those confirm some other really excellent studies that have been coming out, showing that schools with higher suspension rates do have long-term adverse consequences on students.

And as we were thinking through the implications of this, you know, we wanted to think about what's the counter factual, how do we improve this? One is, we think that it's definitely a need to calibrate this power for removal that principals have, and postulate a question whether there should be so much discretion in this process.

And the second is just to lend support to promising alternatives to suspension that have come up.

Some are being presented in this session, and then a lot of other sessions throughout the NIJ conference, so things like restorative justice, positive behavioral interventions and supports, and other kind of more preventative practices.

Okay, so I want to return to that pu.

le that we started off with, when I asked which would have a larger impact on student outcomes, a police officer or this principal switching? And based on our findings from these two studies, we have to conclude, definitely the principal; right? Not only do they have larger effects on a wider variety of outcomes, but they also have these lasting long-term effects for students inside and outside of the disciplinary system.

And we go back to that map that I created in the beginning about the disciplinary process, it kind of makes sense that they would be so central for student outcomes, because they're really pretty critical, either principal or assistant principals or members of that administrative team are very critical in kind of navigating the outcomes of this disciplinary process.

They also have the power to make direct referrals to the juvenile justice system, and so it's not only police in schools that can really have an impact on not only educational but criminal justice outcomes.

So, I'll end by kind of teasing what we're working on now, which is, when we came up with this result about these large and lasting effects of principal disciplinary decisions on student outcomes, we wanted to learn more about the mechanisms through which this was happening; right? So how does a single decision to suspend a student in sixth grade somehow affect their likelihood of graduating high school many, many years later? And one kind of descriptive fact that we latched onto is the fact that students who have one suspension tend to then go on to have many more suspensions; right? So, even though, in our sample of sixth graders, only ten percent are suspended each year, of those who have one suspension, 40 percent are suspended that same year, and then 60 percent are suspended by the end of the next year.

There does seem to be this accumulating factor, where perhaps -- this is what we're testing -- one suspension begets another suspension, or that there might be some causal effects of actually receiving a suspension on your likelihood of the latest offense, and even holding constant kind of our individual propensity to offend.

And so, in this study, we're asking this question of whether there might be some type of state dependence in suspension or past dependence, and kind of the implications of that.

So, does that mean that if you are at a slight disadvantage of this first point in time when you receive the suspension, does it have a possibility to accumulate over time and lead to more disadvantage.

And so we're working on this, using some models from economics, to try and test out these dynamics over time and really understand better the mechanisms that might build this pipeline that people have talked about quite a bit.

So that's all I have for you today.

Again, we are so grateful to the NIJ's support that made these projects possible, and we look forward to answering all of your questions.

So, thank you again.

>> Thank you very much, Lucy.

And I'm also glad Shawn is here with us, too, in case he might help us with the Q&A time.

Now we're going to move to our third presentation by Anna Yaros and by Cheryl Roberts, and they're going to discuss their evaluation of a comprehensive initiative to reduce discipline disparities between African American males and other groups, and they will be highlighting their key observations about challenges, implementations, and strategy for overcoming these challenges.

So our welcome to Anna and Cheryl.

>> Thanks so much, Barbara.

Can everyone hear me okay? If you can't hear me, just let me know.

My name is Anna Yaros, and I'm presenting, as Barbara said, today, with my co-PI, Cheryl, and we're presenting our study.

It began in 2017 with a partnership with Charlotte-Mecklenburg schools, and we're going to go through four main sections today a bit -- can you hear me okay? Okay.

We're going to discuss a bit about disproportionality and school safety, and many of these topics have been alluded to and discussed in the previous two presentations, so, this is, hopefully, a nice follow on.

We're going to talk about the interventions, and just as Lucy said in her presentation, we are going to be discussing positive behavior interventions and supports and restorative practices as to three interventions.

We present the study design and then highlights from our first two years of implementation.

So, just as a little bit of background, as folks have alluded to, we started our study, really, with the thought that zero-tolerance policies have not made schools safer, and this is backed up by literature.

Literature also indicates that African American students, especially males -- but this specific statistic is for all students -- are 3.5 times more likely to be suspended.

This has been confirmed even when studies have controlled for SES, social emotion -- sorry, not social-emotional learning, socioeconomic status.

They have also confirmed it even after controlling for number and severity of offenses.

So, it points of bias.

And so contributing factors include teachers' skill in the classroom management, certainly potential difficulties with cultural competence, implicit bias related to racial stereotypes.

And so, using this literature, we developed an overall model to inform our study.

In the center, you can see our concerns about disproportionality with discipline referrals are really driven by the light blue items on the left that I just mentioned, as well as harsh and inconsistent discipline policies.

They inevitably contribute to climate and safety issues, which are in the green, and then there are long-term outcomes, which folks have also mentioned today, which are decreased connectedness or engagement, dropout, and involvement in the criminal justice system.

So this model is consistent with the one a lot of folks have been talking about today.

So, in response to that, we proposed a combination of three interventions, which are the first three bullets in blue here, as well as support through a coaching staff.

So, positive behavior intervention supports is one of the most common interventions in school safety and does seem to reduce office discipline referrals, ODRs, according to the literature.

But it does not, by itself, significantly improved proportionality.

So, a lot of times, you'll see African Americans suspensions go down, but white suspensions will go down even more.

So the proportionality remains a challenge, even with PBIS in most studies.

Restorative practices is gaining some initial evidence of effectiveness in reducing disproportionality, but there are mixed findings.

There are even findings that it can actually reduce -- can actually increase disproportional effects.

However, there are findings that it can actually reduce disproportional effects, so that's why we included it in to study in our study.

But then a key component, specifically tailored for African American males, was this culturally responsive practices or culturally responsive instruction.

And you see the slash, because culturally responsive practices involves specifically honing in PBIS classroom management practices for African Americans and making sure that they're equitably used.

But it also involves culture responsive instruction, which involves things like lesson plans and cultural competence training for teachers so that they are actually changing what they teach, in addition to how they teach, in an effort to engage and maintain engagement with youth.

We specifically targeted the ninth-grade classes in our high schools that we engaged with.

And then the sort of overarching support system that we tried to frame these all in was a coaching staff, so a team that provided direct coaching with a one-to-one coach school ratio to start with, and was really able to become vetted in the school building and provide support so that these three pretty intensive practices could proceed.

So, our design is a randomized control trial.

We are in year three, as I mentioned, so we don't have final outcomes for you today.

You can see on the left that we took all of the schools in Charlotte, some of which -- I'm sorry, I didn't mention yet, Charlotte-Mecklenburg Schools is our partner.

They partnered very closely with us to develop the intervention and to think about how to implement it in their district.

So, we had 16 schools that participated in our sample, half of which we randomized to intervention, two of whom declined, were not interested in this, and, unfortunately, we have had struggles with disproportionality continue, and those schools added to the other ones and became nine control schools.

So, our outcomes really are owing to this previous theoretical model.

Our outcomes really are the green and the yellow, so they really are disproportional discipline, attendance dropout and connectedness.

We're using a staff survey to get a sense of how staff are changing their multicultural self-efficacy, as well as their sense of their school and classroom climate.

And then we're also doing a cost-benefit analysis to try to understand the benefit to society of a program like this by reducing dropout and promoting engagement in school.

So, we're in our final year of our study.

As you can imagine, things have changed significantly since COVID hit about a year ago.

Some practices have continued virtually and some have not.

We were able to complete student focus - well, staff focus groups virtually.

And we're going to continue to try to do that with student focus groups this year, and process data collection to really understand what is happening virtually.

And then our process will be to analyze both pre- and post-COVID implementation in concert with outcomes.

So, large, multi-level models to try to understand the difference between the two treatment groups, both before COVID hit and since COVID has hit, in our outcomes that we've discussed.

So, Cheryl is going to take us through some preliminary findings so far.

>> Thanks, Anna.

Next slide, please.

So, the process evaluation assessed implementation fidelity.

We investigated barriers and facilitators, and we also looked at levels of teacher self-efficacy for implementing these different interventions.

In terms of fidelity, we gathered data on the dose of training, how many staff got trained out of how many were supposed to, the fidelity of the PBIS Schoolwide Implementation, using observations through the Schoolwide Evaluation Tool, the SET, and that was done by our team at RTI.

The school district has an internal standardized PBIS measure, the Tiered Fidelity Inventory, that they measure quarterly, and that measures the different levels of PBIS, the different tiers.

We also used the RP Observe tool, which is an observation tool developed by Anne Gregory, that looked at the quality of the RP community-building circles the teachers do in class.

The coaches measured that with training from Anne Gregory and us.

We have the staff survey and we looked at frequency of RP circles, use of core PBIS classroom practices.

We had coach surveys and interviews as well, to get data on fidelity.

As far as implementation, it varies.

In facilitators, we did focus groups with school PBIS leadership teams and focus groups with teachers.

We had survey interviews with coaches for that, as well as training after each - surveys after each training, professional help training.

And then for teacher self-efficacy, as Anna mentioned, we had staff surveys and focus groups to kind of understand their confidence and their ability to do each of the three interventions.

Next slide, please.

So, the sequence was that the district started with PBIS, and that same year they added in restorative practices, and then moved on to add in last culturally-responsive instruction.

They did have a brief training on implicit bias at the very beginning, but the bulk of the culturally- responsive intervention was this instruction component, which came last.

Next slide, please.

So, where did they get in the first two years in terms of fidelity? For PBIS, six out of the seven treatment schools completed training through the schoolwide PBIS first module, mostly beginning in the summer before that first year.

One school got to fidelity in that first year.

Four schools total reached fidelity by spring of the second year.

Then, in terms of restorative practices, that was harder, adding in that training after the first one.

It was sort of spread out a little more.

Five schools completed the training through the circles training in circles.

And then three schools, in terms of outcomes in our staff survey, a majority of staff reported facilitating circles at least monthly, and many did so on a weekly basis, which was the goal, in three schools.

Culturally-responsive instruction was last, and that suffered from the COVID interruption the most.

Four schools completed the training.

Only one school got I would say far along in adopting the practices of doing this culturally-responsive instruction, although more than one got started.

Okay.

Next slide, please.

So, now I'm going to talk briefly about challenges.

So, just overall, implementing behavioral interventions in high schools has some challenges that apply to all of them.

You know, the typical high school focus is on academic content; right? That's their primary mission as they see it.

And given the age of students, it's a common perception that it's not the school's role or the teacher's role to teach behavior necessarily, or reinforce behavior.

They should know that by this age.

But, you know, as we know from the literature, it's still relevant for students to learn social and emotional behaviors and to have positive reinforcement, and to have good structures in school to create a safe school environment.

The large size of schools is also an impediment.

It makes it take longer to implement these interventions, and research shows that it takes longer to implement PBIS in high schools than it does in elementary schools.

They need more staff and more complex organization.

Next slide, please.

So, these were some overall issues with the initiative not specific to high schools but just true of implementing any initiative in schools.

We found the same kind of challenges that you would see in general and expect to see.

Time, people don't have time.

They have other competing priorities.

Sometimes the staff might not feel that there was clear enough expectations, we would sometimes hear in focus groups.

And then staff buy-in, nothing is going to happen or get implemented until you have some buy-in from the stakeholders.

So, that's always a critical part of school interventions.

And then what's distinct about this, what we're trying to do with this project is, of course, we're implementing multiple interventions.

Ideally, at some point, having them feel like they're integrated but you're asking the teachers to do multiple interventions.

So, that's definitely a bigger challenge.

Next slide, please.

And then from the teacher perspective, you know, they have specific concerns, too, to the interventions, especially, like, do they have the skills.

You know, they can go to a training and have some follow-up training, but it really takes time to get to the point of mastery, some practice, some guidance, some feedback.

In the case of restorative practices, some teachers had some of those facilitation skills, some didn't.

Some were also not so comfortable sort of sharing control with students, you know, in the circles, where students sort of have a bigger voice.

Culturally-responsive instruction, you know, they didn't necessarily know how to develop culturally- responsive lesson plans, and lesson plans are hard for teachers to change.

Schools do them differently.

It takes time and kind of reinforcement.

And they also need to maybe understand more about student cultures and the specific social, cultural, historic background of the groups they're trying to be responsive to.

This intervention tried to do that, but it takes time.

And then cross-cutting, it's just teachers are not necessarily certain how to apply these in all their day-to-day practices, whether it's their planning, their teaching, their collaboration with other schools and how to make these changes.

Next slide, please.

So, moving on to just quickly some of the lessons that we have so far in the first two years.

So, you know, the big challenge is kind of doing all three things, and it certainly came out in talking with teachers in the groups.

Sometimes they didn't see the connection between these things, you know, what's the relationship among these things, why are we doing all three.

So, that's something that's important to clarify in communication, and I think it also takes time to get that integration.

That takes more time.

In terms of the content, all three interventions overlap in having strategies for student engagement.

They're trying to engage student and increase instructional engagement.

And they have overlapping goals in the way they were implemented here.

Increasing equity, especially focused on African American males, creating a positive school climate and positive relationships.

Next slide, please.

And also, I should just add that they're also trying to promote positive responses to behaviors and reduce exclusionary, disciplinary practices, you know, use of suspension and such, they're all trying to do that.

Okay.

Other system lessons are, in terms of timing and sequencing, that turned out to be really important.

Doing PBIS first was important.

PBIS provided this overall framework that was data-driven in a multi-tiered system of support, laid a foundation for the schools, and we found that the schools who got farther in getting that in place were better able to then add in restorative practices.

We also found that it was helpful to do the training before, when possible, and it wasn't always possible, for teachers before the school year started so that they could kind of get it into their plans.

It's harder for them to change plans midstream, not surprisingly.

Coaching was definitely a critical piece of this.

The coaches worked with the school leadership teams and professional learning communities and really were critical in implementing.

And I think a lesson learned by our partners and us also was, you know, it really takes some time to get the coaches themselves trained and up to speed when you're asking them to coach on three different interventions.

You know, one should plan on at least three to six months for them to get those competencies.

And, of course, they do need the skills and professional development in helping staff overcome resistance and increasing in their buy-in, since that's really the biggest challenge in these initiatives is getting the buy-in.

As far as leadership, of course, nothing is going to happen without administrator support, but they also need to be involved.

You need an administrator on that PBIS team, and teacher involvement on the team, and teacher champions.

It's critical to have good support structures in place, including the communication, those clear roles, regular meeting times.

Just specific to these interventions, we noticed that schools did well when the leaders instituted a set day and time for their weekly restorative practices circles.

Then you had some accountability.

You had clear expectations.

Students expected it, too.

That worked much better.

Also having supports such as creating - having school teams or the PLCs created repositories of restorative practices circles topics; right? The big barrier is, how do I get started? Well, here's some topics you can use.

Same with culturally-responsive lesson plans, "We're developing these, here's a repository, teachers can use that." So, with the PLCs dedicated, consistent time, that's where we really saw movement.

And the school that went far in culturally-responsive instruction, the PLCs embraced it and made that a focus.

Next slide, please.

So, data was, of course - is part of PBIS and was critical to this project.

CMS hadn't looked at their data by race and gender so much before.

They had some other breakouts but not by race and gender.

And so our partners developed a data dashboard and key indicators by race and gender.

And they worked - the coaches and schools worked together to use new data to get a shared understanding of the problem because they didn't know, you know, what the patterns were in relation to these gaps for African American males.

And then to develop specific strategies that address the specific patterns that they're seeing in the data.

As far as from a research and practice standpoint, for the evaluation, really, we're going forward in terms of schools measuring their own fidelity.

PBIS, of course, has all these standardized tools.

It's been around for a long time.

Restorative practice doesn't have as many standardized tools for measuring fidelity.

And, you know, it would definitely help to have more practitioner-friendly tools there to look at different thresholds of fidelity over time.

Same thing for culturally- responsive instruction.

I'm going to probably skip on for time, but I would just say that we had a nice feedback loop with the research.

We would find out from the focus groups, for example, that they needed more - the teachers needed more - and leaders needed a little more structure and support.

The coaches would then adjust what they were doing, so sort of the feedback loop and having discussions about the data as we saw it was also very helpful for implementation.

Next slide.

That just wraps up some of the early lessons that we've been seeing so far.

And, you know, this leads to thoughts about what would be useful to continue in future research.

And I think more research on the combinations of interventions and integrating interventions for behavioral and academic outcomes that would improve school safety is definitely needed because these are complex problems and one intervention isn't going to just fix it; right? So, the solutions have some complexity as well.

And I know in a previous presentation someone presented on the Interconnected Systems Framework, which is another example of that.

Certainly, more research on restorative practice, which has been very promising in some of the research to reduce discipline disparities.

And then just developing evidence-based models of culturally-responsive instruction in relation to the outcomes of interest, you know, it's research and foreign.

We know it's the right thing to do, but there aren't a lot of RCTs and say, "Oh, here's the models, you can take them and expect these outcomes." Bradshaw and team are doing some research on Double Check, which kind of addresses that.

And then even just measuring teachers' skills and self-efficacy and culturally-responsive instruction is difficult.

There's social desirability.

People don't know what they don't know and it's challenging, and we're working on a manuscript related to measurement of self-efficacy.

Next slide, please.

So, I want to thank NIJ for funding this and supporting this project, and our colleagues at NIJ, and my colleagues at RTI.

And a huge thanks to Charlotte-Mecklenburg Schools for partnering with us on this project.

Thanks for listening.

>> Thank you so much, Anna and Cheryl.

Now we're going to move on to our fourth presentation by Miner Marchbanks III, better known in law as Trey Marchbanks.

Trey will focus on his examination of the relationship between immigration and school violence by probing variation in school discipline and juvenile justice referral rates in schools near the Texas-Mexico border in comparison to other Texas schools.

So, without further ado, welcome, Trey.

>> It helps if you unmute yourself.

I apologize about that.

I would like to, first off, express my thanks to the National Institute of Justice for having me.

Very honored to be a part of such a good group.

If you look at the list of who's speaking in this conference, it's like being included in Woodstock.

So, I appreciate that.

And I appreciate their support for what we've been doing.

This particular paper is a subset of a lot of the work that we've been doing with NIJ funding, looking at the relationships that exist between immigrants, the border, and school discipline and juvenile justice, and how those experiences may be unique for schools on the border and for individuals who are immigrants.

And got two other co-PIs, Jamilia Blake and Anthony Peguero, who, you know, couldn't do this work without them.

So, if we think about the relationships that exist between race and school, there's a long history; right? Nobody can ever forget the picture of Ruby Bridges as she was integrating the school, had to have federal marshals attend school with her, take her into the classroom, take her out.

You might even consider these the first police in schools.

I'm sure there's probably some other incidences, you know, but that's of a more substantive nature.

That's since evolved into the school resource officer model that was discussed already throughout this conference but also by Lucy just a moment ago.

And there's a bunch of different findings, just as Lucy found, some good, some bad, about what that is like, fueling a criminalization of school, in some instances feeling safer conflicting evidence.

But what we also know is that there's this established school-to-prison pipeline, where you start in school discipline, and that leads to a host of negative outcomes, whether that be grade retention, failing the class, or dropout, or ultimately, involvement in the juvenile justice system and the adult justice system.

While most of the research has focused upon African Americans and Latinx individuals, not much has been done on the immigration side, and how does the experience differ for those individuals who are immigrants.

And we want to base a lot of this paper on minority threat literature, and that's the idea that as a minority group grows in size, the dominant group begins to view that as a threat, either an economic threat, a political threat, or both.

And so they take action to try and maintain their dominant status.

And there's a list of different things that people have studied, everything from voting regulations to increased law enforcement, penalties and things along those lines.

Ken Meier, in political science, has even shown that simply changing the school board elections from being at large for the district to have single member districts elections, that that dramatically increases the presence of Latinx members on the board.

So, we also then want to be cognizant of this idea of the threat that exists within schools.

Over the last couple decades, we've seen a big push toward zero tolerance and the presence of security personnel, school resource officers, metal detectors, several different things, to try and protect the overall - protect students and make students feel safe.

Now, unfortunately, not all the research shows that they do feel safe in those instances.

We've also seen dramatically high levels of discipline.

A recent study showed that 60 percent of students have experienced some sort of suspension from school between seventh and 12th grade.

So, you're actually in the minority now if you make it through school without being removed from the classroom.

So, when we think of place, we know that urbanicity plays a role in what happens in the education system.

There's resource differences.

There's different practices.

You know, we've shown that urban areas have higher discipline rates than suburban and rural areas.

Suburban schools are much more wealthy, as a general rule.

But, also, the campus that a child attends becomes very important, the climate that's there.

And so you look at some of what Cheryl and Anna were just talking about and different things, those are important factors.

But, also, if you think about socialization, the education system is our first real interaction with persons of authority outside of our family.

And so how we learn to deal with persons of authority, a lot of that has to do with how your education experience goes.

If we look at the border, across the border areas, they're viewed as areas of crime and violence.

A lot of that has to do with rhetoric in the media and politics and the like.

And there's also, related to that, this "immigrant criminal" myth, which the myth goes that individuals who immigrate are more likely to be violent offenders, drug offenders, things along those lines.

And those have largely been shown to be inaccurate.

So, one thing that we've noticed is that racial threat has been studied quite a bit, especially as it deals with school.

What has not been studied is whether there's an immigrant threat aspect and whether or not the border relates to the extent that there is an immigrant threat, if there is one.

And we pick school discipline and juvenile justice as a lens for us to look at it because we think it is informative, especially in a state that's on a border like Texas.

So, we have three basic research questions.

The first is simply, is there variability in the likelihood of being disciplined or referred to juvenile justice based upon location? So, if you're on the border, are you more likely or less likely? Second, after we look at just the raw numbers, we want to go through and say, okay, we're going to control for a variety of factors and see if that discrepancy is there or not after control.

And then we want to look and see if the relationship that could exist between punishment and children who are immigrants, is that there and is that moderated by - excuse me - mediated by the presence of being on the border? So, in the interest of time, I'm going to run through the data and methods.

Basically, we combined the individual-level education data with the individual-level juvenile justice data.

Had 16 years to follow these individuals.

We classify first-generation immigrants.

Luckily, Texas education agency simply records, was this child born in another country? So, we classify somebody as a first-generation immigrant if they have that flag.

Second-generation immigrants, we use some of the literature that other people have used that says, okay, a child who is labeled as limited English proficient when they start school but is not labeled as a first-generation immigrant we're labelling as second-generation immigrant.

So, if we look at just the first question, is there variation that exists between discipline and juvenile justice referrals on the border versus not on the border, there really aren't.

They look amazingly similar to each other.

Even first-generation immigrant, it is actually pretty close, significantly different, but from a substantive perspective, it's not an overwhelming difference between the border and non-border.

What we do see, though, is second-generation immigrants are much more likely to be on the border, as are Latinx students.

In fact, African American students, there's only about one percent of the kids in a school in the border are classified as African American.

So, definitely African Americans tend to be in your non-border communities.

As a point of context, in Texas, with the cohort that we have, Hispanics were the plurality.

But in Texas today, the majority of students are Hispanic.

So, if we go and we look at school discipline rates and what predicts it, if we look at the first model two, we see that actually first-generation immigrants are protective.

Your discipline rate at a campus is lower the more immigrant children you have of first-generation immigrants.

Being in a border county has no significant effect when you control for these factors.

Model three, we interact the presence of being on the border with immigrant status.

And then border becomes protective overall but what we see is a unique situation with second-generation immigrants.

It's protective to have more second-generation immigrants in your school if you're not on the border.

However, if you are on the border, it leads to higher discipline rates.

So, there's this complicated relationship where the context of the border matters when we're trying to discuss what do we see as far as discipline.

Now, if we look at the juvenile justice referral rates, this is how many - what percentage of kids are referred to the juvenile justice system from a campus in a given year.

Everything looks almost exactly the same as it did with the referral rates.

We do see that there's a protective factor having first-generation immigrants, however, what we do see is a dramatic difference between border students that are second-generation immigrants versus non-border.

So, it goes from being protective to actually being predictive of having juvenile justice referrals, which is supportive in large part of the immigrant threat hypothesis.

So, if we go through our research questions, there's not any aggregate difference in school punishment rates or juvenile justice rates on the border and non-border.

They're roughly the same.

Statistically, they are the same.

When we control for known factors, we do see that it's still not a significant difference when we don't separate out the effect of immigration.

But once we do separate out immigration, then there's actually less discipline occurring on the border and juvenile justice contact.

Most importantly for this particular study, though, is the border plays a very strong mediating role in determining the effect of having second-generation immigrants in your school.

So, we definitely see support for that.

So, some of the implications, I'm going to jump through these pretty quick, in the interest of time, so that we can have Q&A.

But the "myth" of the immigrant criminal as being delinquent and problematic really isn't that accurate.

If you look at the first-generation immigrants, it's actually protective for a campus to have those students.

We're not the first study to show that.

That's been shown several times.

On the second-generation students, we need to be concerned about whether or not there's going to be what some authors refer to as the downward assimilation, where they become more likely to be involved in discipline, juvenile justice, criminal justice and the like.

And then Aaron Kupchik's and Rios' ideas of "wrong lessons," "youth control complex," where people are utilizing the lessons that they're learning in the school to generalize outwards.

So, they're not questioning authority, to just go along.

And then hyper-criminalizing individuals to where that becomes part of the stigma.

I could definitely talk a lot more about that if we had time.

So, some of the future questions that we want to make sure that we can consider is the complexities that exist between immigrants and their history and identity and how that varies.

In this particular case, we are likely dealing with - we are dealing with more Hispanic individuals as far as immigrants are concerned, but that encompasses a wide variety of backgrounds.

Some individuals are coming over for economic reasons.

Some are coming for safety reasons, any number of things, and that's not just unique to the Latinx community.

You also have, within the Asian community, there's a variety of different nationalities, a variety of different reasons for coming, and those things need to be considered.

Also, to what extent does the message that's coming out from the media, society, and politicians, what degree does that influence this immigrant threat hypothesis, and how does that play out? We also want to make sure that we're - we feel that we need to make sure that federal and state policies consider the unique aspect of immigration but also the unique nature of education.

So, I thank you for your time and look forward to questions.

>> Thank you so much, Trey.

It's really a very opportune time for these panelists to be presenting on disparities and how immigration and racial disparity, and socioeconomic, urban, rural, suburban all impact the lives of kids going through school disciplinary proceedings as well as trying to protect their school safety.

I don't see any questions posted in the Q&A.

So, once again, to all the attendees as well as the participants, feel free to enter a question in the Q&A.

I did have one question that came to mind, listening to the various presentations, and that had to do - initially, I think it was Nick, you were talking about various readiness of different groups to proceed with implementation.

And then I believe later on, Anna and Cheryl were talking more about sequencing of multiple interventions or trying to have everything happen at the same time.

So, one of the implementation barriers I heard across your two presentations was when are individuals or types of individuals ready, and then also whether or not trying to take everything on simultaneously, especially with teachers that are already feeling really under pressure to make sure their performance measurement is going to be achieved against their standards of learning and their educational testing.

So, when you try to lay on three new programs across so many different people, maybe think about sequencing.

I didn't know if we might discuss that.

>> I'll just quickly jump in before letting the others chime in about sequencing, but the nice thing, I guess, for us about our study is that we're not studying any particular school safety intervention or the implementation of that intervention.

We're really kind of taking a step back and looking at what it would look like for a school to be ready, or a community or students, but it's a great point.

I mean, we talk about all the time, you know, you can have an intervention brought on by a district that says no more student suspensions.

You know, the district was ready, they imposed it on the schools, and then, all of a sudden, the principals are saying, "Wait, what do you mean I can't get these kids out of the school?" Teachers are saying, "What do you mean I can't remove that kid from class?" So, yeah, readiness is huge, but I'll let others chime in on the sequencing of particular interventions.

>> Yeah, I think it is really interesting.

We were able to - we had set up PBIS as that first one because of the framework sort of model that it provides.

And so I think that works but getting folks to PBIS fidelity is a challenge sort of in and of itself.

So, it's a process.

And any grant is probably not long enough, even if it were many, many years.

So, Cheryl, did you have anything to add? >> Yeah, I guess it's important to think about how the things fit together.

So, there might be more logical sequences, depending on what the interventions are, you know, and how they might relate to each other and potentially build on each other, as well as what the capacity is to absorb different things, and even when you can get in the different training.

Teachers have limited professional development time.

Then you have snow days and they're gone, you know.

So, getting the training in is hard You definitely have to prioritize what has to come first.

>> I think it's really important for there to be the types of studies that we're seeing because I get asked questions about, "So what should a school do?" And without good evidence-based measures to be able to point to, I'm at a loss sometimes to say, "Well, do better." And it's kind of like "Let them eat cake." So, having evidence-based things - and there are some literature out there that I can point them to, but when you're doing an RCT on it, you know, that provides a lot more weight.

>> I just thought of one more thing, too.

Some of it maybe relates to readiness.

You know, we did notice variation in which schools embraced which interventions.

So, sometimes you have a more built-in audience for one thing than another.

So, some schools went really far with culturally- responsive instruction and others really embraced PBIS.

So, there is also that local variability of what their priorities are at the moment, and it's important to try to match up what you're doing with what they care about as much as you can along the way.

>> If I could add to that just real briefly, I mean, I think we were thinking hard about how to think about policies, and the level of discretion at the school level is tremendous.

So, on some level, we started thinking about the principal list of policies.

So, when you're talking about the readiness or acceptance, a lot of that's going to be driven by the administration team.

And in a middle school, it's usually a principal, maybe an assistant principal, and they have a lot of discretion.

So, thinking about what that means, because you do end up with a situation where no matter what you say at the district level or anywhere else, it's going to be determined by what's going on there.

And I think that's not as well understood - and I thought I heard undercurrents of that in some of the other presentations.

So, thinking about what that exactly means given they can be very localized disciplinary context.

And I think criminal justice is used to that, thinking about that, but I'd like to see if the lessons we've learned in other contexts might be able to be applied in this context and the same ideas.

>> John, that's a great point.

Principals have a lot of power and authority over their schools, you know, and the district is only going to push so hard.

So, they absolutely have to be leading on something.

>> I think we have the experience where principals have declined our intervention and then sort of realized, okay, disproportionality's a problem, sort of like you're saying, Trey, coming back around to say, "Wait, what can we do?" So, it just - readiness wasn't there and then it comes around.

>> That's true.

It did change over time.

>> Yeah, we saw pretty massive differences when principals switch in terms of suspension policies and attitudes towards suspensions, to the extent to which we didn't find any evidence that suspensions themselves were making anyone any safer.

That's an interesting idea, and maybe that's not surprising given what we know about deterrence and those kinds of things.

So, to think about what's worth giving people discretion over, whereas things like having someone present and, you know, sort of monitoring and those kinds of things might actually be more relevant, especially in the school setting.

>> I think at this juncture I'm going to have to jump in and thank all of you so much.

I'm glad the immigrant risk of criminality is a myth.

Keep working.

It's a pleasure to meet all of you.

And thank you to our attendees for joining today.

Very much appreciate it.

And also, thanks to Jake for all the technical support.

You've been here with us the whole time. Thank you.

>> Thank you all.

It's good to see you again.

>> Thanks everyone.

>> Bye.

>> Join the closing session at 3:45.

>> Great. Thank you.