Children Exposed to Violence

Erica L. Smith: First, I'd like to introduce Kimberly DuMont. She's a research scientist with the New York State's Office of Children and Family Services. She specializes in issues related to child development and child maltreatment, and she's currently working on a randomized control trial to examine the effectiveness of an intensive home visitation program. That's what she's going to speak about this morning.

Then we'll hear from Dr. David Finkelhor. He's director of the Crimes Against Children Research Center and co-director of the Family Research Laboratory at the University of New Hampshire, where he is also a professor of sociology. Much of his work focuses in the area of child victimization, child maltreatment and family violence, and he recently published findings from the National Study of Children Exposed to Violence, which he is going to speak about today.

And then we will hear from Patricia Stern, Tricia Stern. She is founder of Stern Steps, a consulting firm that advises government, private sector and nonprofit organizations on policy and program management in the area of family violence. She provides training on addressing family violence for … across the spectrum of practitioners, community and social groups, and government agencies, focused on identifying … on improving the identification and treatment of children exposed to violence across a number of different violence areas, and she's going to be our discussant for the panel today.

Kimberly DuMont: Hi. I'm Kim, and I am going to talk to you today about … briefly about a home visiting program that's designed to help prevent violence in children's lives — or abuse and neglect, specifically — and spend a bit more time on the randomized trial that's been done to evaluate whether that program is successful or not.

The program or the evaluation team is a combination of workers from New York State Office of Children and Family Services and also from the Center for Human Services Research at the University of Albany. Just as a caveat, the Office of Children and Family Services, we operate completely independently of the program division, which is actually housed in a separate building, so just to maintain our objectivity.

The home visiting is particularly important at the current time. There was recent legislation passed this spring in which $90 million were allotted to fund evidence-based home visiting programs, and these programs have to serve populations that are high need, that the programs have to demonstrate measurable effects in six areas ranging from child abuse and neglect to self-sufficiency outcomes. They have to show effects are sustained beyond the program's intended period, and they have to be cost-effective. So there's a big bill, a big order to fill there.

Why home visiting? Why is this getting a lot of attention? Why are so many dollars being put behind it? Home visiting is seen as … first, there's a bit of evidence, particularly around the nurse-family partnership, saying that home visiting can be effective in preventing some of these outcomes of interest. Also, it's an opportunity to enter a family's home, to break down some of the barriers for … that center-based care faced and to work with families in their natural setting, so that you can help support exchanges as they unfold.

One national model is called a Healthy Families America model, and it's a strength-based approach to home visiting. It extends from prenatal or the birth period up through the target child's fifth birthday. It provides intensive services, and the Healthy Families New York program is based on that national model. Healthy Families New York is a voluntary program. It serves a very, very diverse group of families. Some of the women will have already had children. Some are referrals from child protective services. Others are first-time parents. Some are kids 12 years, 13 years of age who are pregnant for the first time. Some are 34 years old who are pregnant for the sixth time. And, as I said, it uses a strength-based approach.

The program has — and the model, national model — have four specific goals: to prevent child abuse and neglect, to enhance parent-child interactions, to promote optimal health and development for the child, and to prove self-sufficiency outcomes: education, employment.

In New York state, the way that the program works is it targets very high-need communities that have demonstrated through Department of Health indicators that they have low or no prenatal care, they have high rates of teenage pregnancy, high rates of infant mortality, and they're particularly impoverished.

Families are sometimes brought to the attention of the program through their prenatal providers or at the time of the child's birth, or they can come, they are recruited through schools or offer the program that way.

Once they are referred to the program, everyone is assessed using the Kempe Family Stress Checklist, which is an inventory, a semi-structured interview to assess for risk of child maltreatment. These might be prior contact with the CPS system, a history of drug and alcohol abuse, poor or inappropriate expectations about parenting.

If a family scores above a certain cut point, then they're eligible to receive the program, and they are referred to the home visiting component of the program. If you are not eligible, you would be referred to other services in the community that fit your need.

Home visits are delivered by home visitors who are recruited from the communities that they serve. This is unique to the Healthy Families model. They tend to be paraprofessionals. They go through extensive training. They have a week of core training. They are not allowed out into the field until they've shadowed visits; they received one and a half hours of supervision a week in addition to the shadowing. They receive wraparound trainings. Visits begin, if the woman has entered prenatally, during pregnancy and then intensify after the birth of the child and then diminish over the next four years, depending on the need of the family.

In 2000, a randomized controlled trial was designed at the state … getting money from the state legislature to answer questions about the program's effectiveness. Eleven hundred and seventy-three women who had assessed positive for the program, just like I talked, were either referred to receive home visiting or just received referrals to the community. The randomization process was done by computer program at the sites. They were not able to break it. They tried. Some of the sites tried, but everyone was assigned appropriately. And you can see the distribution across the control group, which received the assessment and then referrals, and the treatment group, which received the assessment and referrals and home visitation.

The randomized trial then did a baseline interview about a month after they were randomly assigned, and then they followed up the women, if they had entered prenatally at the time of the child's birth. If they had entered postnatally at the time of the age 1, everybody was followed up at age 1, age 2, a subgroup at age 3, so they could do videotaped observations.

And then we obtained funding from the National Institute of Justice to follow up with families at age 7, and I'm going to talk to you primarily about the age 7 follow-up.

At age 7, we put back in all 1,173 families and said, “Who's eligible?” To be eligible for the follow-up interview, you had to still be … you or the child had to be alive … you and the child had to be alive. You had to not have crossed over into the other group. So, if you were in the control group, you had not to have received the treatment. And you had to have … be living with your child. And that was true for the interviews.

For the administrative record searches that we did, we took all 1,173 women and followed them for the entire time of the follow-up period. So that's the true intention-to-treat component, those analyses.

We did interviews with mothers, interviews with the children. We obtained child protective services' records for confirmed cases from five years prior to random assignment up through the target child's seventh birthday. We also looked at their foster care history. We looked at any services, preventive services or protective services that were initiated to avert a placement. We obtained data from the Office of Temporary Disability Assistance to get benefits for food stamps and public assistance, and we extracted cost data from the three sites at which the people were randomized.

And here's a little flowchart of where we went. So we followed up about 80 percent of the 1,173 that started and 84 percent of those who were still eligible. Eight hundred children were interviewed.

What I want to accomplish with just the baseline — these are showing you the treatment and control group as measured as a baseline — is the two groups are remarkably comparable. There are very few differences, and, actually, if you run 40, 50 variables, these two are the ones that come up, but I'm showing you. So it doesn't look quite as impressive as when you run the 40 variables.

And people in the treatment group had slightly higher levels of risk, but, clinically, I'm not sure it's meaningful. But the biggest challenge was that there were more target children who were female in the control group than the treatment group. And I say that's a challenge because on the kinds of outcomes we were looking at the child's functioning at age 7, if you were looking at externalizing behaviors or internalizing, then gender starts to play a role in those, and having that not be equivalent across the two groups, you know, we needed to address at least statistically.

So, for this presentation, I'm focusing primarily on the question of whether the program is effective in preventing or reducing child maltreatment, and I'm just going to take you back to some earlier results that we published, which found that between the baseline and year two interview, we saw some … or year three, we found some evidence that the moms were more likely to adapt positive parenting strategies. This might be appropriate limit setting. We also conducted videotaped observations that were micro-coded, and they tended to use more responsive, sensitive parenting. This is consistent with other studies on home visiting.

We also looked at — based on conflict reports on the Conflict Tactic Scale, we saw differences primarily in how often mom used strategies that involved physical aggression or psychological aggression and not as much in the rates of those behaviors. There was a trend towards differences in rates, frequency of neglect at both years, although that wasn't significant.

At year seven, we looked at the CPS records, and we could look at whether a mom was confirmed as a subject, and we could also look at whether the target child was confirmed as the victim. And we didn't see differences as with other home visiting programs and in the earlier years between rates or the frequency of CPS reports, with the exception of confirmed sexual abuse, which, actually, there were no cases of confirmed sexual abuse in the treatment group, and there were five in the control group. Statistically, that's significant, but the incidence is so low, I haven't brought that up here.

We did see at year seven, consistent with the patterns we were observing earlier, that moms were more likely to use … more moms used and they were more likely to use alternative strategies of nonviolent discipline. This is consistent with the positive parenting patterns we were seeing earlier.

We also found, similar to year two, that moms in the treatment group were much less likely to use serious physical abuse strategies or serious aggression, such as burning your child or choking them, than the moms in the control group. And year two, we found a 75 percent reduction, here we see an 80 percent, so that's very consistent across the two time periods.

For the first time at year seven, we were also talking to the children. We gave them a picture card version of the Conflict Tactic Scale. It was actually audiotaped, so they could listen to it through the headphones, and they had a touch screen where they could privately select their answers. And we only asked them questions about nonviolent discipline strategies, psychological aggression, and the minor physical assault because, well, A, the IRB would not allow us to ask any more severe questions, and, B, we were also uncomfortable putting the child in that position. So, for them, the minor physical aggression was the most serious that they were allowed to answer questions, this would be, you know, spanking, grabbing, and we saw that the children also reported that their moms from the treatment group were less likely to engage in these behaviors.

At year seven, we saw no differences in self-reported neglect, despite the trends earlier.

What difference does this make for children? Well, we did both direct assessments. We used the Peabody. We asked them questions about their deviant activities, their bullying activities. We also asked the moms to … we completed the child behavior checklist with the mom, and we didn't see any differences for the problem behaviors or socio-emotional difficulties or self-regulatory practices, and, again, that was a direct assessment that we had done.

However, when we looked at indicators that might set people up to be successful at school or put them at risk to have poor school outcomes, we saw a pattern of consistent findings, and so our kids who were in the … offered Healthy Families New York were more likely to participate in gifted programs, were less likely to receive special education services — there was no difference on the Peabody — and were less likely to skip school. That's from their self-reports.

So we are trying to reconcile. I mean, we had no differences on CPS reports. There are no differences in neglect, which is largely defined by poverty in indicators, and we don't see differences in their behavior or their emotional health, but we do see differences in their functioning at school.

We then asked the question if there are groups that are particularly … the program is particularly effective with, and we initially asked this question a few years ago because there was some home visiting programs that weren't finding results, and there were others that were finding results. And you could say, well, it might be who delivers that care, nurses or paraprofessionals, but another question was who's receiving the care.

And so we isolated. We looked at the programs that were particularly successful, and we said, “Well, they tend to serve young first-time moms who enter during pregnancy. This is as close to primary prevention as you can get. They have had no opportunity to parent their child. Their minds are particularly elastic and malleable. They are vulnerable about their new roles. This is a good point at which to intervene.”

So we isolated this subsample within our own group, because we serve a very diverse group, and we found at both year two and year three, using two different methods, both self-report and observations, if you look on the group on the left, that there were big differences in harsh parenting. And we saw this across a number of different outcomes, but we'd always find about 20 percent differences.

So, at year seven, we again looked at this group, and we found that there are no differences in rates of CPS reports, but, again, we see differences in self-reported parenting, and we also see that their children are showing some signs of doing better in school as well. And these rates are fairly high, that they were half as likely to repeat a grade by second grade. So this is very encouraging as being a group that might be ripe for target.

But we also wanted to ask at the other end of the spectrum, for those women who had already had children and who had already been confirmed in a report to CPS prior to entering the program, what is the program's chances of success with this group, and we had hypothesized that it wouldn't be very good. I mean, you're in the house one day a week, and you're working with them, and so we were thinking we're not going to see very much, but we were wrong.

And what you see — and these are truly unprecedented reductions in the rates of child maltreatment. These are confirmed reports. You see if the mom … it is particularly effective with the mom. That fewer family services were initiated to avoid placement for these families, and these are the odds that are about half or effect sizes that are .50.

So we wanted to understand why is that happening, and it turns out that that relationship between the program and those reductions is mediated by the number of subsequent children that the women had, and this, too, is consistent with findings that the nurse-family partnership has, but they have only done it with young first-time moms. And so here we see, again, that working with families to family plan and to reduce the number of births reduces the opportunity to maltreat and also creates greater opportunity for them to be self sufficient.

We also looked at particular program activities to further confirm that this might be the case, and, indeed, when we looked at relation correlations between family planning and other similar activities, we saw that they were inversely related to rates of report, so this is good.

What does this mean for government, because that's the last item on it? It's that it has to be cost-effective. For over the … by age 7, the program … essentially, the cost of the program still stands when you look at the whole group of women. When you look at this last group that I still talk to, you've recovered 442 percent — 442 percent of the cost or a savings of about $12,500 per family by age 7.

When you look at the high prevention opportunity, the young first-time moms, about 40 percent of the costs have been recovered by age 7. These are based only on administrative costs. We did not value any school indicators. We did not value any of the other non-monetizable outcomes such as positive parenting or self-reported. This is based just on benefits received and interaction with the CPS, the cost of the program and revenues generated from the mom, so these are conservative estimates.

Am I done?

Erica L. Smith: You can take a little more.

DuMont: OK. So what do we take away from all this? Healthy Families New York serves a very diverse group of women, and they do that well to some degree, particularly in terms of promoting positive parenting strategies and for … or limiting milder forms of abuse, but there are … and this seems to translate or trickle down to higher success in school.

There are ways to prioritize the services, too, at least in some of the families that you're serving, and these are to groups that are particularly vulnerable and have the greatest opportunity for change, those who have never parented and those who have already been in contact with the system and working that way.

So I just wanted to end on that note, and that's it. Thank you.

[Applause.]

David Finkelhor: Children exposed to violence … I guess I should start by talking about the concept because there's a fair amount of confusion about the concept itself. Does it include the victimization of children in addition to the witnessing of violence? Well, yes, and most of its usage it does. Does it include exposure to media violence, like on television or the movies? Well, no, mostly it doesn't. Well, how about nonviolent sex offenses against children, like sexual abuse? Most people who use this concept want it to include that, even though many of those acts do not involve actual violence.

How different is this exposure to violence concept from child maltreatment, which is another concept that is widely used by people concerned about these same things? Well, child maltreatment has a large component of things like neglect and emotional maltreatment and abuse that most definitionally conscious social scientists have a hard time calling “violence,” if you think of violence as acts that are involving force that are committed with the intention of causing pain. But most policymakers want this covered under the concept of exposure to violence.

I proposed a number of other concepts that I see as somewhat more definitionally defensible. I've used the term “developmental victimology.” I've also used the term “childhood victimization.” That's the title of my book on this subject. But I'm afraid these haven't caught on exactly. They're not commonplace.

So it's too bad, in a way, we don't yet have an agreement on what to refer to all of this as. I have been taken to calling this as much as possible “exposure to violence, crime and abuse” to make sure that some of these other things are well included.

What I think I can say is that almost everybody in the field is in agreement about the idea of moving in the direction of studying and advocating about a broader range of children's experiences, that what we've been doing up until now has been unfortunately fragmented. I sometimes like to use the term “Balkanized.” I have a cute little slide. I'm not even sure it's geographically correct anymore, but it illustrates that we have had a lot of research and advocacy and kind of separate, siloed, unconnected issues related to children's safety and victimization, but the recognition is that there are real benefits to looking at all these things in a more comprehensive and holistic way. Among other things, it really gives us a better perspective on the full burden that kids experience and the way of these harm-generating kinds of experiences.

We get to look at the interrelationships among these kinds of exposures and many people suspecting that they are not independent, that they're occurring to the same kids, and that the risk for one increases the risk for another.

A comprehensive view allows us to identify the kids who are really the highest risk and most victimized. It also allows us to appreciate better the developmental patterns as some of these things morphed into other of these things as kids move up the age scale.

It also promotes better collaboration. One of the key problems in the area is that a lot of the resources, the attention of policymakers or even school principals, is divided among people who are advocating for the bullying prevention program or the dating violence prevention program or the sexual abuse prevention program, and by collaboration, there's maybe a chance that more kids will get exposed to more prevention about more of these things. And there may be an opportunity to bootstrap more research if we really collaborate.

So I've been spending a fair amount of my time trying to figure out ways in which we can promote a more comprehensive approach to this, and one of the main efforts I've made is on the issue of measurement because it seemed to me when you have a way of gathering information in a more comprehensive way, then people begin to think about it in the way that they can measure it. And this kind of approach was very effective in the field of delinquency, for example, where the self-reported delinquency measure, I think, was very instrumental in bootstrapping a lot of research about a comprehensive view of juvenile offending.

So the juvenile victimization questionnaire, which we've been developing for the last 10 years, originally consisted of about 34 different kinds of exposures that tried to represent all the different kinds of things that advocates and researchers had been looking at, but also grouped into some of the modules or clusters that identified important subcategories that people might want to look at separately, and this included, you know, topics that had received a lot of attention, like physical abuse, but topics that hadn't gotten much attention, like nonsexual assault to the genitals or burglary of a family household, and includes things that are fairly common and things that are rather rare and unusual, but it is comprehensive.

And we have had a chance now to utilize this in a wide range of both developmental and then national survey situations, and most recently and what I'm going to talk about is the National Survey of Children Exposed to Violence, funded by the Office of Juvenile Justice and Delinquency Prevention, which we have just been reporting some of the findings of and is really our second large national survey with this kind of approach.

These data that I will show you come from the survey that was conducted in 2008. So we're talking about events that occurred to these children in 2007 and 2008, and we were able to gather information on a nationally representative sample of over 4,500 children between the ages of infancy and 17.

The data was collected by telephone interview. When the children were under the age of 10, we interviewed the caregiver with the most knowledge about the child's experiences. When the child was over the age of 9, between 10 and 17, we actually got permission and interviewed the youth themselves. These interviews, we paid 20 bucks. We had an over sample of minority and low-income children, and we did pretty well in terms of our response rate.

Now, in addition to the JVQ, we also expanded the kinds of questions and issues we were asking about on a number of dimensions, as you can see here. So I'm going to be speaking of assault. You're going to get assaulted now by a lot of data. I'll try and pick out some of the important stuff that you might want to pay attention to.

Overall, over 60 percent of the children had been exposed to something in the past year, and we've been talking about exposure in the past year. We also have lifetime exposure. I'm much more favorable to talking about past-year exposure. I think it's more valid and more comparable, but in our reports, you will also find information on lifetime exposure, if that's something of particular interest to you.

Sixty percent is a large number, but I also feel it's important to emphasize. Because there's such a broad range of things that we're looking at here, including peer assaults and sibling assaults, I prefer actually not to focus on it or dwell on it. Of course, one of the big values about what we have done here is that we have aggregate numbers as well as able to drill down and pick out specific topics of interest to people, and I think that's very important, so that people … you get a more complex view of actually what's going on.

But you can see close to half the kids had been exposed to a physical assault in the past year, and then, you know, one in 10, maltreatment episode, 6 percent some sexual victimization.

Another really valuable thing was the ability to really look for the first time at the trajectory of these experiences over the course of the full childhood, and you could see that rates were pretty high for some of them, over pretty much the whole course of childhood, like physical assault, property victimization. For some kinds of exposures, like sexual victimization, the witnessing of violence, and even the rates went up as kids got older.

So I'm going to also just run through some of the specific types but not dwell on these too long. I'll give you the reference for a couple of sources where you can get the specific information if you want it.

So, for example, about one in 10 of the kids was actually injured over the course of the last year as a result of an assault. These bullying kinds of episodes are not included in the aggregate physical assault here; substantial number of assaults by siblings, non-sibling peers.

The summary: Boys have higher rates. Most assault in bullying victimizations. Highest rates among elementary school-age kids, 6 to 9, but certain categories like dating violence, highest among teenagers, and as you can imagine, most of the perpetrators in these assaults are juvenile acquaintances and siblings.

Here are the findings on the sexual victimizations. Of course, sexual harassment, the largest category, but substantial numbers of assaults, attempted and completed rapes, and, of course, highest victimization among female teens; adults responsible about 15 percent of the total of sexual victimizations and 29 percent of the sexual assaults, once again primarily acquaintance perpetrators.

These are the maltreatment rates. I think we probably did our worst job in the estimation of neglect, which as maybe you know from child protective statistics tends to be the largest category, but it's hard to really assess in a … because it takes so many forms, hard to assess in the survey format without really taking a huge amount of time. So I think we certainly underestimated that here, but a substantial amount of physical abuse, psychological.

The rates are similar for boys and girls, and rates are lowest for the preschool-aged kids and actually grew as the kids got older.

Property victimization. Boys have higher rates, and, once again, to the extent that the perpetrators were known, most of them were juvenile acquaintances. And then witnessing, which includes the witnessing of domestic violence, witnessing of physical abuse of siblings, includes witnessing of assaults in the community, and then some categories of things that have rarely been measured before, like kids exposed to warfare; these would be immigrant kids especially; the witnessing your family assaults, again, mostly teenagers.

Now, one of the really interesting things that we were able to do is because we had done a very similar survey back in 2003, the Developmental Victimization Survey. We were actually able to look at some trend information. We used many of the same screening items, although we had a somewhat different sample, 2- to 17 year olds as opposed to infants to 17. So these rates are comparing the sample, 2- to 17 year olds in both the 2003 and the 2008 sample.

And the very interesting and encouraging news is that we saw many more declines than we saw increases, and they were declines in areas like, particularly, peer victimization, where there seems to have been a great deal of prevention effort during recent years, so a suggestion that maybe we're getting some return for that investment. So physical assaults declined. There was actually a pretty big increase in kids reporting robbery, but that was one of the few that increased overall. Theft declined. Burglary of family households — these are the significant changes, the ones that are highlighted here — declines in emotional and psychological maltreatment, pretty substantial; almost 10 percent decline in peer and sibling victimization; bullying very big, almost by a third, bullying; some significant decline in sexual assault and sexual … what would be statutory sex crimes; some increase in witnessing of sibling assault but decrease in overall exposure to community violence.

And this is kind of the summary here, so declines in physical assault, sexual assault, peer and sibling, including physical bullying, psychological abuse, exposure to community violence, and theft. We did not see declines in caretaker physical abuse or neglect, and these two elements were an increase. But overall it seems like a very encouraging development and not inconsistent with what people have found in other surveys of exposures of this sort, like from the National Crime Victimization Survey.

Another thing that the study has allowed us to do and one of the reasons we were eager to do it was to look at the interrelationship among different kinds of exposures, and as you would expect, having the one kind of exposure does put you at a considerably increased risk for other kinds of exposures. So, for example, if you had any physical assault in the course of the past year, you were about five times more likely to also experience a sexual victimization. And there's pretty much complete what I call transitivity among these kinds of exposures in the sense that they almost all … they virtually all lead to increased exposure to other kinds of victimizations.

And, of course, one of the things that this means is that some kids are experiencing a tremendous amount, and this is a slide that shows the number of separate, different kinds of victimization that a child reported having experienced in the last year, which ranged all the way up to about 22. And, in fact, the majority of the kids who did have one or more of these … had one exposure had at least … had two or more, but we also plotted this against some measures of psychological distress that we had in the survey. And not only did it show that, of course, the increasing number of exposures were associated with increasing amounts of distress, but it also showed that there was kind of an inflection point here at which … and this is a cut point that demarcates about 10 percent of the kids in the sample.

And we have developed a term or taken the term “polyvictim” to refer to these kids who experienced a very high level of victimization with a consequent high psychological impact, and the polyvictims, as you might imagine, not only have the highest number of victimizations but the most serious kinds of victimizations. They are substantially more likely to have been injured in the course of victimization, to have faced a weapon-toting assailant, to have had a sexual victimization as part of their package of victimizations or to have had a caregiver perpetrator involved in one of their victimizations.

And they also are — and this is one of the reasons why we think polyvictimization is so damaging is that they have the victimization in a number of different domains, so that there's no real place of safety in their lives. They have these multiple environments where they get victimized.

There is a somewhat larger percentage of polyvictims among the African-American youth and among kids who come from single-parent or step family environments, not that much, not as much association with SES.

I'm going to, in the interest of time, skip over and just make a very important point, though, one that has been insufficiently recognized by people working in this field, and that is to say that most of the studies that we have on the impact of victimization or exposure to violence in childhood has been based on exposure to a single type of victimization or single type of exposure, like sexual abuse or physical abuse or exposure to domestic violence.

And because these studies have not really gotten a comprehensive assessment of all the different kinds of victimizations kids experience, they have tended to overestimate the contribution that a single type of victimization makes.

And to illustrate this, what we did was to show the kinds of regression coefficients or associations you would get when you simply take something like sexual victimization and assault and correlate that or regress that with some trauma measure or psychological distress measure, and you get substantial coefficients here, but if you add the polyvictimization measure — that is, the question of whether the kids are exposed to other kinds of victimization — the individual kind of exposure doesn't seem to make that much difference. So what's really going on with so many of these kids in so many of these findings of associations between, say, sexual victimization and distress is that it's the polyvictim kids in these samples who are carrying the distress.

Another way of illustrating that is with these slides here, which show kids divided up into a number of groups: the nonvictims; the kids who have a single kind of victimizations, so let's say in this case the sexual victimizations, the sexually victimized kids but divided into two groups, the kids with a low, either one or very low number of that kind of victimization, sexual victimization, kids who have a chronic, that is, a high number of those individual kind of victimizations but are not polyvictims; and then the kids who have any of those sexual victimizations but who are also polyvictims. And you can see how in every case while the individual victimization may be associated with some elevation of distress, it's really the polyvictim kids who are the ones who are carrying the high levels of a problem.

We're really trying to train a lot of our attention now on understanding more about this particular subgroup of polyvictimized kids. One of the things that we've done in some of our work now is to try and develop some notion about the pathways by which kids get into this kind of polyvictim condition, and we have a conceptual framework that suggests that maybe there are four major pathways: through living in dangerous neighborhoods, through having violent family environments, through having families which have substantial amounts of diversity, and through having existing mental health problems or disabilities.

I don't have time to go into this in a great deal of detail, but suffice it to say that we have found some support for this in one of the longitudinal studies that we've been able to do. We do find that each of these variables associated with each of these pathways do seem to make an independent contribution to predicting the onset of polyvictimization in the longitudinal perspective.

The other thing that turned out to be very interesting when we looked at this onset of polyvictimization was there seemed to be two main spikes, one at age 7 and one at age 15, which seemed to correspond to the first year at which these kids have entered into a new school environment, typically moving into elementary school and moving into high school. And we think that for kids who are vulnerable for developing this polyvictimization pattern, probably having to adapt to a new environment where they don't really … where there's a new social structure, a new status hierarchy, where there are lots of new routines that they have to accommodate to, where they have to make new friendships, where the people who protect them and watch out for them may not be … in the past are not available or the parents themselves may be very anxious about the children's transition to these environments, the parental control may be declining because of the children's graduation to these new environments, these are particularly vulnerable situations, and that we need to be doing a lot better job of picking up these at-risk kids early on in these environments and trying to improve our ability to reduce their descent into this kind of maelstrom of victimization.

Just to summarize, we think there are many implications of this, but one is we need to assess kids for a much broader range of victimizations. We need to put more effort into identifying the most highly victimized kids. There's some differences in terms of early and later onset of this, and we need to sort of focus on that more. We need to look more at the relationship between this polyvictimization notion or the spectrum of victimizations and whether kids are also developing delinquent and aggressive patterns of behavior.

We need to have treatment approaches that are more multifactorial and that take into account the fact that kids have had multiple exposures. We need to help people who work in some of these more Balkanized or siloed areas, like child protection, to also get experience with information about things like bullying and peer assault and dating violence, so that they can direct resources towards kids who are vulnerable to that.

If you want more information, there are two good sources for the overview. There's an article in Pediatrics in October 2009 that summarized a lot of the findings. There's also a Department of Justice publication, and there will be additional … there are many additional publications from the study, including ones focusing on polyvictimization, that will be out in the near future.

And then, if you go to our website, there's a great deal more information about all this. OK.

Patricia Stern: Good morning. I think, as you can see, this is a very exciting time for practitioners working in the field addressing children's exposure to violence, one with the new NatSCEV that Dr. Finkelhor just spoke about, but also Attorney General Holder. This is one of his passions and has devoted a lot of interest and support not only in funding but awareness about the issue as well. And NatSCEV has really started to have practitioners think differently about how are we really going to best serve these children and really sets the stage of almost a starting point of where we need to begin to best reach these children. And Dr. DuMont's work tells us a little bit, it gives us some of the elements that we need to apply in best serving these children.

So how do we apply this research locally? Well, both NIJ and the Department of Health and Human Services have outlined, starting with a needs assessment, to go to the local community and figure out what is already in place, to begin to both plan and develop services and interventions and treatment programs and follow-up services for kids exposed to violence.

The final reports of both NIJ and Department of Health and Human Services report the community should make children exposed to violence a priority and identify the needs of these children, services that are in the community for these children, the availability and accessibility of services, and gaps in services. And they specifically talked about services, but I would include interventions — and we're going to talk a little bit more about what we mean by “interventions” — and then implement a strategic plan based on these findings, so really a blueprint of how do we go forward in each local community of best identifying and treating and preventing children's exposure to violence.

So the Safe Start's demonstration sites came up with several key elements in terms of responding to children's exposure to violence, and I'm going to go through them today. It's early detection and identification, training and community outreach, protocols, policies and procedures, awareness of services, evidence-based interventions, comprehensive and coordinated responses, staff support and supervision, and evaluation.

So let's first start with early detection. We really need to have better case tracking in regards to children's exposure to violence. You know, we have NatSCEV, which is on a federal sample or national sample, but if we're going to start working in local communities, we have to try and better understand what is going on in that local community in terms of children's exposure to violence.

So we want to first see is there any published local data. That could be from organizations focusing … now, again, as Dr. Finkelhor said, really NatSCEV is the first time looking at a broad range of children's exposure to violence. So, unfortunately, we're going to have to piece apart what are the child welfare statistics, what are the domestic violence rates, and we want to see if there's any research studies in our local communities that are implementing the self-reports that Dr. Finkelhor spoke about.

What are our community agencies doing? Are they case tracking? Are they identifying it? Even if it's in their intake or case files, how can we best get that information to better understand what's going on?

Law enforcement records. Now, this gets a little trickier when talking about children's exposure. Law enforcement does a very good job of documenting domestic violence, sexual assault, homicides, but many communities don't have any information about whether children were involved, whether they were present or their proximity to that violence.

Are there any databases? Again, we have databases spread out by types of violence, but one is the NCA track, which is for child advocacy centers to look at children who have come into a child advocacy center where there's been an allegation of child maltreatment or mostly child abuse and then, you know, looking at it to see if there are other types of violence as well.

And co-occurrence information: Particularly around domestic violence and child maltreatment, many states have now started implementing co-occurrence protocols. Well, child welfare agencies are now screening and identifying children exposed to domestic violence. We need to better understand the numbers there. What are they finding? How many cases are they seeing?

So then we move to identification assessment and who should be screening for this. Well, obviously, key professionals involved in children's lives: health care providers, particularly pediatric providers; schools; preschools; Head Starts. Dr. Finkelhor spoke about that schools, you know, in middle school and high school, that's a peak period for polyvictimization. Then we need to, you know, train schools to better be identifying these children, particularly at those time periods. Day cares, after school programs, youth programs, police, faith-based organizations, child welfare and juvenile justice system. So, wherever you would normally think about where you would find children, this is where we need to work to better have them identifying and assessing for children's exposure to violence.

Dr. DuMont spoke about how important prenatal and newborn care is in terms of screening and identifying for children exposed to violence. Holden in 2003 was talking specifically about children's exposure to domestic violence but did state that exposure prenatally is a form of exposure. So we need to start thinking about that, as well, as a type of exposure. And as Dr. DuMont talked about, screen pregnant women in a community rather than only new mothers. This, again, means OB/GYNs, midwives, prenatal and newborn home visits, and we're not only talking about the home visitors that Dr. DuMont's program employs, but also, you know, I've been involved with trainings for doulas, for lactation consultants, you know, people who are getting into the home when this violence may already be occurring.

As Dr. Finkelhor mentioned, polyvictims or children exposed to violence are more likely to have other life adversities. So, therefore, screening should also be done by professionals working to address these adversities, such as those working in unemployment, welfare. You know, many community agencies work to address economic hardships: substance abuse providers, mental health agencies, and then those addressing accidents and illnesses, first responders, emergency rooms, but also, again, to make a coordinated approach, if you are a domestic violence program addressing that injury, that trauma, also to be screening for the other forms of violence as well.

So what do we want to screen for? I think, again, Dr. Finkelhor did a wonderful job explaining this broader range of exposure categories. And, again, what is the child's proximity to the violence? Where are they? Were they victimized? Did they hear about it? Did they know about it? And, again, as Dr. Finkelhor talked about the wide developmental spectrum, we need to know children of all ages what their exposure has been; the single type of exposure and the frequency, but, again, the polyvictimization, the co-occurrence, and not just of the types we're used to putting together, domestic violence and child maltreatment, but all the other types that NatSCEV talks about to really be assessing better for those.

The past-year exposure and the lifetime exposure: I think from a practitioner standpoint, lifetime exposure is important because particularly I work with health care providers and pediatric providers, and when they're screening for domestic violence, they often will tell me, “Well, it was in the past, so I don't worry, but I didn't ask about it.” But oftentimes we know that maybe the violence itself is not currently occurring, but we have custody battles; we have harassment; we have stalking; so there are other things we need to know about.

We want to know the impact of exposure on the child and the family, so emotionally, trauma, anxiety, depression for both the child and the caregiver, behavioral symptoms. Socially, many of these children, in my experience, have very few friendships, and if they do, they're very superficial friendships, so we want to get a better understanding of their social skills. Physically and developmentally, how are they doing? And it's also been shown that children's exposure to violence impacts the attachment relationship with the caregiver, so we want to know how that relationship has been impacted.

We want to know about the other adversities going on for the family, but we also want to try and take a strained-space perspective as well and look at the protective factors and the resiliency both in the child and in the family.

So how to screen, you know, I think it was talked about, the importance of speaking with parents and caretakers directly, children directly. I do assess children as young as 6 and older. Dr. Finkelhor mentioned there were common victimizations and peak risk periods for different age groups. So we want to make sure if we have a child in that age group that we're specifically asking about that type of exposure.

We need to have protocols and procedures on how to assess and respond. It's really unfair to practitioners to just say to them, “Screen.” We need to tell them how to do it, what to do if they find it, because most of them, and rightfully so, feel they're opening a Pandora's box and have a lot of concern about if they get a positive, “What do I do with it?”

We need to have cross-agency policies. It wouldn't be very helpful to have this agency doing this procedure and policy and this one doing that. We need to try and come together and have a cross-agency coordinated community response in how we address it.

We need to give practitioners screening questionnaires and tools that ask about exposure to violence but also are asking about the impact, and these assessments need to be culturally, linguistically and age appropriate. There are some measures out there that are not. We need to either develop them or implement ones that address this.

And we really do need more valid and reliable screening tools. There's been some studies in the health care field, but, again, talking about specific types of violence, much fewer around children's exposure to violence. We really need to come up with some that address that, but also, you know, is the health care screening tool, should that be the same that we use in a preschool or a Head Start? Probably not.

So, again, we can't just tell people to screen. We need to train them on how to do this, and we need to begin by talking about what are the types of violence that are out there, including this new broad definition of children's exposure to violence. We need to tell them about the local data around children's exposure to violence, because most people don't realize it and at least don't know it from the broad definition that NatSCEV is talking about.

We need to provide them with identification and assessment strategies, and Dr. Finkelhor talks about multivictimization assessment. Child welfare agencies are obviously asking about one type of children's exposure but may not be asking about the broad range. We need to be really thinking about this from a multivictimization or polyvictimization perspective.

And, again, back to coordination, we need coordination of trainings. Practitioners tell me all the time that there is the domestic violence training here and then there's the sexual assault training there, and then three years later, I took this, and there may be a component here and there on children. We really need to have a better … you know, a more coordinated, comprehensive training that someone can take that covers this broad range of children's exposure to violence.

And training has to include referral resources. Again, to have people screen, identify, to be able to begin to address this, they need to feel there's somebody who can pick it up if they identify it.

Similarly, we need more standardized and evaluated trainings for children exposed to violence. There are many, many trainings throughout the country, again, on different types of violence, some on children's exposure to violence, but very few have been standardized or evaluated. So not only are we reinventing the wheel, we don't really know what works.

We need, again, trainings for different settings. A lot of times, we do a one-size-fits-all and just open it up to the community, but a training for a child welfare worker is going to be very different, again, than a preschool worker or law enforcement personnel.

I say training for parents, but we need to also be thinking about how we talk to parents about this information, how they can identify. We know that whether their own child is experiencing it or friends of their children, we need to include them in thinking about how we identify it and screen.

We need to have a greater understanding of how information is best received. It's been started to talk about, again, health care settings that the big lecture with 200 people is not the best way to get the information across, that it needs to be in smaller formats, but a lot of that is anecdotally. We need more research of what works best, what type of training works best for people to really be able to receive the information. Do we do role plays? Do we do it in a lecture format? Do we show videos? What works best?

And there have been studies that have shown that trainings for violence, a lot of this is around domestic violence and sexual assault, but that there is a change in knowledge and attitudes over time but very little change in practice. So, while it's good that we're changing knowledge and attitudes, our real goal is to change practice. It's to have people really be screening for this. So we have to think about how do we do these trainings where really people do change how they're screening.

OK. I want to just move to interventions. So interventions should also be evidence-based, and for a lot of providers, this sometimes makes them nervous. There are evidence-based interventions out there for children's exposure to violence, and there are also a lot of theories that you can base interventions on to explain why you're using a particular intervention.

Should have an evaluation component, like Dr. DuMont's work. It really is important to show that what you're doing works.

We have a lot of different, wonderful programs going on that we know very little about if they're effective, and few people know about it. We really need to have more of a research component.

Intervention should have a clear goal, and that goal should be to prevent or reduce the impact of the exposure, and that seems like a simple statement and, of course, but a lot of times we get these families, they're in crisis, and we just implement a program without really thinking of what is the larger goal.

We want to focus on prevention and treatment. There's, I think, been more focus on treatment, but we need to really begin to think about implementing programs focusing on prevention.

We also need to have programs that address the individual but also the family, community, and also systems, you know, changing the way both systems address and respond to children's exposure to violence.

Again, a coordinated approach, again, Dr. Finkelhor talked about, you know, with the map. You know, we have the child abuse intervention here, the exposure to domestic violence here, and we're not working together, and especially given the polyvictimization, we're really doing children a disservice. And interventions as well should be culturally, linguistically and age appropriate.

Both Dr. Finkelhor and Dr. DuMont talk about addressing adversities, that that was shown that families have a lot of adversities, and Dr. DuMont's work showed that addressing these adversities really helps. So interventions need to do that.

They need to have community input. I used to work at a health care setting, and people in the community used to call hospital personnel “the clipboard people,” you know, that we would come, we would ask them a lot of questions, and then we would tell them the programs that they needed in their community. We need to have their input about what they feel they need, what's going on.

And these programs should be situated directly in the community with a dedicated leader and staff. They need to be accessible and available. If they're not and only certain people are able to get there, can we really say that it's a program that benefits many, or is it only benefiting the people that are able to somehow get to this service?

It should have staff support and supervision and, again, begin as early as possible in the child's life, including the prenatal period.

So questions we have in regard to the field, again, there are many, so NatSCEV is a great first start about better understanding what's going on for these children, but we need to know what are the long-term effects of children's exposure to violence on these children, and is the impact different, and should we be having different interventions for the different types of children's exposure to violence. Should a bullying program be very different than an exposure to domestic violence program? What about what do we do, do we separate kids if they've had an individual episode of one type of violence or a repeated exposure? And, again, those polyvictims, what do we do about them, and, again, is the impact different for these children?

What about child witnesses as opposed to direct victims? Can we serve them together? Should we be planning things differently for them? And how should the interventions for polyvictimization be designed? I think NatSCEV really highlights the importance that if most of these kids have … or a lot of these kids have had multiple victimizations, well, then we really need to be designing interventions for that and how do we do that.

A big question in the field is, “Are interventions effective when the child is still living with the violence?” People feel strongly on either side, particularly, again, around domestic violence. Should we be serving kids when they're still living in the home with the violence, or do we need to say no, it's not going to be effective, we need to wait until the family is not living with the violence?

And for these evidence-based interventions, are outcomes improved over time? I think it's wonderful, Dr. DuMont's study, when the children turn 7. We need more studies like that, that show what is happening to these kids over time.

Again, how accessible and available are these interventions? What are the barriers, and how can we overcome them? Many practitioners say there are a lot of barriers to reaching interventions: waiting lists, transportation, not culturally or linguistically appropriate. What do we do about that? How can we address that, and then does that make a difference for the type of interventions that have been considered evidence-based?

Now that we've broadened the definition of children's exposure to violence, we have many different abusers. We now have sibling abusers. We have school, peer abusers. We have parents. We have adults. So that means we have many abusers. What do we do for them? Do we create interventions for them? Is it worthwhile? Is there a way to prevent it?

And then which intervention should we use and at what time? We have a lot of interventions focused on the child, on the family, on the school, community, systems. Do we do them together? Do we do them one at a time? Which ones?

We need to have more public awareness and resource materials, again, about the broad definition of children's exposure to violence. I think there is — you know, the recent case in Massachusetts of bullying has made bullying, I think, take on a greater role, but most people would not include that in the definition of children's exposure to violence. So we need to really be better about publicizing that.

The NatSCEV data, we need to be better about getting out there, as well as local CEV data.

We need to tell people, get out there more about the real impact of this violence on children. We need to have people know that there are both prevention programs and treatment services.

We also need to find a way to disseminate best practices, what has been shown to work in the field. Again, many practitioners and providers are not aware of that.

And referral procedures, how do people get help, again, both if you're a family — how do I get help if this is going on for me — as well as providers if you're working with a family.

As I said, many states have implemented policies, particularly in child welfare agencies where they're addressing the co-occurrence of domestic violence and child maltreatment, and they're not removing. If there has been a case of domestic violence, we need to get that word out there to families that if they disclose domestic violence, you know, their children, there's a new way of responding to the family.

We need a greater distribution of research and knowledge in the field. We need more coordinated efforts around research and practice. Again, in many areas, there isn't really a good connection. As I said, we need more coordinated efforts around the different types of violence and exposure and not be working separately. Again, I love that map, you know, not be working in our different countries.

We need to have a strategic plan not only for local communities, but as a country — where are we going, what are we trying to accomplish in regards to children's exposure to violence, how are we going to get there, what are the ways we're going to implement — because the goal is to replicate programs, policies and interventions that have shown to be effective in preventing and lessening children's exposure to violence across the U.S.

Thank you.

[Applause.]