 It's 1146 so I think we should make a start with session two. So we've got the first paper in this session is by Diana Batchelor. So Dr Diana Batchelor is an ESRC postdoctoral research fellow at the Centre for Criminology at the University of Oxford. She conducts research on victim survivor perspectives of crime and justice. She focuses particularly on victims' expectations and experiences of communication with the people who committed offenses against them. So Diana if you're ready to share your screen. I can see you're there. I am here. I'll just briefly say hello. I quite like everybody else introducing themselves so you know who's talking but I will disappear once I share my screen. I would like to tell you about the answer to this question from the crime survey. Some of you might be surprised it's possible to even look at whether victims are willing to meet the offender from the crime survey but there's some interesting questions in there which I will tell you about in detail. And as per the introduction I'm looking really in terms of victim perspectives on justice generally and particularly on whether and when they choose to meet the offender usually under the umbrella of restorative justice which means different things to different people. But I am talking here specifically about meetings dialogue between victims and offenders after the offence and at different stages during the justice process. So you may or may not be familiar with the kind of process but it's taking place within the criminal justice system at different stages in the UK. Most often it's which vary in success and funding. So why do we why is the specific relationship between the severity of the crime and victims choice to meet the offender important? Well this general support for restorative justice tends to depend on the severity of the offence with most people assuming that it might be only appropriate for victims of minor offences basically. That's the general public opinion and in fact translates into policy and practice as well. So you often find criminal justice agencies offering restorative justice more often to victims of minor crimes and less frequently to victims of serious crimes and that's an issue if at least some of the benefits that there have been shown to be from restorative justice you know. So there's been lots of studies showing that there are benefits for victims and if even a proportion small proportion of these benefits are in fact the case then denying victims of serious offences the opportunity to get involved is potentially denying them the opportunity to really recover and build resilience and increase their well-being. There are some studies on the relationship between the severity of the crime and victims willingness to meet the offender. Most of them suggest that this relationship is negative. As you might imagine the measuring severity is difficult and so there are different ways in which this has been done. Quite a few different studies show that both in hypothetical situations and in real life victims of more serious crimes are less likely to want to participate. However, there's also quite a few studies finding no relationship at all including analysis of the British Crime Survey as it was then in 2000 just that there was no difference between those who had experienced minor and serious offences regarding whether they would want to meet the offender. Also in hypothetical scenarios which can obviously delve often quite in more depth into the specifics of the crime but then they're perhaps less ecologically valid and there's even a few studies suggesting that the relationship might be positive. In particular this Zabel and his colleagues in 2017 there's probably the most robust quantitative study on the issue. They had three different measures of severity and one of them two had no effect at all but one of them which was the harmfulness of the offence did appear to have a positive relationship on whether the willingness victims were willing to take part. So this quite strange landscape of previous literature led me to hypothesize that perhaps it is a non-linear relationship because it seemed that maybe at the lower end the relationship was positive and at the upper end of severity the relationship might be negative and there are some indications from qualitative studies that might be the case for example victims who didn't want to meet the offender some saying it's not serious enough another saying it's too serious. So this was kind of my idea of what might be in fact going on and I set out to test it using the crime survey. So I used a method which I actually would be interested to hear people's thoughts on because as my impression of the crime survey when I first started it there's so many different variables you can plug them in anyway you want and you can almost always find something positive that looks interesting and obviously there's a move within social sciences in general to have more open and transparent methods and so I was keen to pre-register a study because I found found some effects using the data from 2015 to 2017 and then wanted to pre-register that effectively a replication of the study first part of the study so that I couldn't change it and tweak things to find positive results just because I liked the look of them and I think that was quite successful because what I basically found was that some of the effects I identified in the first study replicated and others didn't and I didn't have the ability to fiddle around and change that so but I'm yeah I think pre-registration of secondary data analysis is fairly rare and I'm interested to hear your thoughts so the question in the crime survey you may know victims who people who are victimised are then asked in the victim module it's explained this way victims are sometimes offered the chance by the police or other criminal justice agencies to meet the offender in the presence of someone else to ask the offender why they committed the offence and how to say how it made them feel and then there's a series of questions about whether they were given the opportunity at the time and if they weren't they were asked the hypothetical question really would you have accepted this offer if it had been made to you at the time and that is the dependent variable on all of my studies and there's three separate elements to each so this was all of these were included in both study one and two so the influence of overall severity so you will probably know the one to 20 severity measure in the study and that's effect on willingness to meet the offender then I looked in more depth at some potential sub dimensions of severity other measures of severity so the impact and whether or not the offence was violent and then I just dip my toe in in terms of psychological explanations for these relationships because in actual fact the measures in the survey don't really allow you to go in much depth on that but I still think there's some interesting findings so the first exploratory study I did using two years of the data suggested that indeed my hypothesis that there was a curvilinear relationship was correct and if you do enter the quadratic so sorry so I should say what I've done this is a general additive model which looks at all different types of curvilinear relationships I hypothesized that it was a curve an n-shaped curve if you add a quadratic term that was in fact significant but there's lots of advice that actually that over estimates whether there is indeed a rise at the beginning and a fall at the end and so a two lines analysis which is a simple simply a piecewise logistic regression that estimates a breakpoint in this this slope and tests whether there's a difference in the slope before and after the breakpoint and indeed in study one all of that indicated that it rises in the lower end of seriousness scale and falls off at the end however when I repeated this using other three years of the survey I did not find that there was no breakpoint the quadratic term was not significant and as you can see this is what it looked like so sort of a straight line but also it was not a linear predictor of willingness to meet so as because as you can see this is a very gentle slope actually so in conclusion the answer was did I find any kind of relationship between the overall severity and willingness to meet no and yet I proceeded to look at some of the other dimensions because from the exploratory study that suggested that violence so just the binary measure of whether the victim said that the fence had included the threat of force or violence reduced people's willingness to participate and impact which I added together this impact scale and I kept it at three increased people's willingness and in both studies this was true this was a pattern across this is just one of so I entered into a logistic regression with a lot of control variables that could be relevant and found indeed that that was the case basically that yeah violent offences were associated with fewer people wanting to meet the offender and the greater the impact the more likely people were to want to meet the offender so this is an interesting finding because obviously usually violence and impact increased together so the fact that they have an opposite effect on people's willingness to participate tells us a few things about what's going on which I'll talk about later so I then wanted to just explore these particular relationships in more detail there are kind of lay theories that haven't been tested either way that perhaps you know this the negative relationship from violence could be because people are more afraid of the offender and then obviously less likely to want to meet them or it could be the exact other way around that when the victim is afraid they do want to meet the offender in order to kind of reduce that negative emotion and so there was I had a kind of bi-directional hypotheses for these three factors and I added them into a path model that looked like this to understand whether they were mediators of the relation the negative relationship from violence to willingness to meet or potentially mediators from the positive relationship between impact and willingness to meet and the the they were but in this way which I shall talk about so the criminal justice response so this was a measure of whether the police had identified the offender and taken action unfortunately there wasn't enough there weren't enough people where the police had taken action to then break that down further and look at the mediation effect but it's just a general indicator of whether justice had been served really from the victim's perspective and this actually did explain so obviously with a cross-sectional study we've got to talk about causality quite carefully but this was a mediator between violence and willingness to meet so the more violent the offence the more likely action had been taken by the police and the less likely the victim therefore was to meet with the offender perhaps surprisingly because you might imagine that when the police have taken action the offender has been identified and secured put in prison so it's all much much more possible and yet it was the opposite way around and then you might have thought on that basis and then the role of emotions seemed to be as positive mediators between impact and willingness to meet so the greater the impact of the crime the more likely the victim was to say they had felt anger and I picked to lose to lose confidence as the kind of mildest measure of fear because that was most all-encompassing and the victims who said they lost confidence tended to also say they felt fear or anxiety and some of the other measures in the survey and both of those were positive predictors of willingness to meet so again somewhat counter-intuitively I think most of the theories would assume that anger wouldn't mean hostility towards the offender and that someone wouldn't want to meet with them but in fact victims perhaps were seeing the opportunity to meet with the offender as a way to express their anger and reduce it and similarly with fear that was making people more likely to want to meet the offender and just finally I've rushed through that a bit but I'd be interested if yeah people have comments and questions so the important thing for where I started at the beginning is that there's certainly no evidence of a negative relationship between severity of the crime and willingness to meet so the idea that we could be not offering restorative justice to victims of serious crimes on the basis that they wouldn't want to anyway just has no basis whatsoever of course people you may want to exclude victims of certain crimes for other reasons but this kind of lazy excuse which is I'm not I'm not going to talk to them about it because they probably wouldn't want to is not acceptable and there are other studies so this Van Kampen women's study from 2016 talks about that the restorative justice offer should be proactive and that victims who are offered it say they appreciate being offered it even if they don't want to take part they weren't any negative effects associated with being offered it and so this finding really aligns with that and then obviously from my look at the overall severity I didn't find a positive relationship either but I think the counter-intuitive findings that impact specifically was likely to make people more willing to meet the offender and that anger and fear actually again increase people's willingness to make to meet the offender what although all of that suggests is that firstly somebody from the outside can't predict what victims will want to do and won't want to do so the idea that justice professionals friends family well-meaning people can predict whether a victim wants to meet the offender or not is not true because we're making what turn out to be quite biased assumptions I think and all the victim is too afraid they'll never want to meet the offender in fact from my qualitative research I've also interviewed victims who do meet the offender and for many of them overcoming a sense of fear was really important motivator so that fits with this in terms of the findings and yeah I'm going to end there because the basic unfortunately I would have liked to say oh well these are the very clear predictors and that means we can create these exclusion criteria for who we shouldn't and shouldn't should and shouldn't offer RJ to but in actual fact I think my findings suggest we can't assume anything and in an ideal world all victims would be offered it and in fact if you do need to create any kind of exclusion criteria for resource constraints probably the people who experience no impact and no emotions as a result of the crime are the people to exclude rather than the people at the serious end of the scale and I will leave it there we will move on now to the next speaker is Lucy Bryant so Lucy speaking on the inequalities in victimisation Lucy is a research and policy officer for the Institute of Alcohol Studies in London currently studying for a PhD in social policy at the London School for Economics so Lucy I'll hand over to you to present your slides thank you okay I'll just share my screen okay thank you so much yeah I am presenting some of the findings from a kind of bigger programme of work that myself and Dr Carly Lighthalla have been undertaking looking at how victimisation from alcohol related antisocial behaviour and and alcohol related violence is distributed across different socio-economic groups in England and Wales so I'm going to focus on our ASB findings today but I will give some of our violence findings for some context as well and obviously I'm happy to chat about either if anybody's interested and I'm really pleased to be able to say that our collaboration was awarded a commendation from the ONS Research Exit Awards last year and following this we've already got plans to take our findings forward and try and look at some of the drivers that might be behind these big inequalities that that we found particularly thinking about off-trade alcohol availability so supermarket and off-licensed sales and kind of where those patterns and how they may be mapped on on these violence and other victimisation patterns we've come across so one of the kind of most common findings in alcohol harm research is one of inequality so there is a particularly prominent finding often referred to as alcohol harm paradox which is the fact that despite lower socio-economic groups drinking less on average than others this is the group that experiences the highest degree of alcohol related mortality and morbidity and this and other findings like it really got us thinking well what about social harms what about things like violence and ASB that we know are really strongly connected to to alcohol so we've used the crime survey to investigate that over the last couple of years and our first findings were around violence and we did find really big inequalities mirroring what's already been presented in health harm so we found that those in the lower socio-economic groups experience higher prevalence of alcohol related violence overall and actually kind of what's behind that is really much higher incidence and prevalence rates for alcohol related domestic and acquaintance violence specifically so this isn't really about violence in the nighttime economy violence committed by strangers and these gaps were really quite big as well so for example we found that the most deprived groups could experience as much as 14 times as many incidents of alcohol related domestic violence every year when compared to the to the least deprived so these findings were really troubling and we went on to try and understand was were there any kind of known risk factors for violence other than this SES driver that could be explaining these patterns so we looked at this and we included risk factors like gender, age, whether or not victims had a disability, how often they were attending nighttime economy spaces as well and generally the patterns held so we did find that SES was a risk factor for alcohol related violence overall and the same was true for the domestic and the acquaintance violence and there wasn't really much of a pattern going on in stranger violence so that they were our first kind of violence findings but we also know that alcohol is associated with some anti-social behaviour and and while anti-social behaviour is maybe more kind of a subjective experience than something that can be kind of defined as a violent crime that doesn't mean it's any less impactful necessarily so there's a lot of survey and focus group work that gives us some insight into this including how anti-social behaviour can actually affect people's sense of safety and shake where they choose to live or spend their time. We also know that there might be ways that these experiences kind of intersect with other inequalities so having your sleep disrupted due to someone's drinking has been linked to deprivation and given what we what we know already about how alcohol harm is uneven what I mentioned about the health harms and then our own violence findings we thought this is another area where it's important and we kind of we kind of look at how these experiences are spread. So we use the crime to everything in Wales as I mentioned and these are the items we drew upon as you can see we use three different SES measures and we actually repeated our analysis three times with a different measure each time we did this to kind of counteract any limitation that one measure might have if we were seeing the same picture with all three measures we could kind of trust what we were finding and not worry that it was necessarily a kind of a hangover from one of the measures. We included the same risk factors as we did when we looked at violence and we used items around ASB that asked people whether they had or hadn't experienced it in the last year alcohol related ASB that is and items that asked how often they'd experienced it if they had and there was a huge range of time frames available and we turned it into kind of a binary weekly or more or less than weekly. So the first thing we did was create prevalence and frequency rates using those two items I just mentioned and the next step which we are currently working on is a similar regression analysis using those extra risk factors. So when I present these patterns I'm going to show you we're then going to kind of look at what else is going on and what else might be explaining them so we know how to kind of move forward from that. We actually used five waves as prime survey data and that was simply because by the time we are kind of cutting everything by SES and then by all these other risk factors you're getting down to quite small numbers of people so that's why we've done that and it worked really well in the violence work so that's the positive. So here are the first results so this is asking people whether they experienced this alcohol related antisocial behavior in the last year and as you can see it's a pretty even spread there and no big differences between the lowest and highest SES groups there's no particular pattern across all the different SES measures you know just under one in ten for almost any any group you want to look at and this is quite different from what we saw with violence and you know we saw this and we're thinking okay maybe this is this is kind of a different picture maybe we haven't got this this kind of big inequality quite the same way we do with violence but then we moved on to look at the frequency so these people who reported ever experiencing this ASB in the last year they then reported how often they had experienced this alcohol related ASB and this is where we see really big differences so amongst those victims of alcohol related antisocial behavior it's those from the lowest SES economic groups that are most likely to experience this with high frequency where high frequency means at least weekly and as you can see the proportions are really quite stark now it's quite different so for measuring your SES by your occupation or your housing those in the lowest SES groups by those measures more than half of the ASB victims here had experienced alcohol related ASB at least weekly throughout the whole year so this is really quite a qualitatively different experience to many of those in the higher SES groups and I think that's something we need to really pay attention to especially when we're talking about effects of ASB that I mentioned at the beginning you know around feelings of personal safety and physical effects like your sleep and particularly as we might think about these these are obviously individuals reporting these but if these are things that are occurring when someone's in their home there might be children that are caught up in in these kind of experiences as well especially if the kind of ASB is like noise outside and things like that so this these gaps were really quite stark and really worrying to see and here it is graphically so this is your when SES is measured by income by home ownership and by occupation so we again we're seeing this same shape that we saw with the violence findings earlier all of these findings together I think have led us to kind of consider some of the policy implications that might come out with them and it's really positive that the violence findings have already been raised in some of the debates around the domestic abuse bill and kind of thinking about the role alcohol might play and thinking about you know approaches to that kind of violence and how that might be might be different and we also are really interested in in the potential that pricing and availability interventions around alcohol might have to kind of close some of these gaps so there's already kind of wealth of research that suggests raising the price of alcohol and reducing both kind of the physical and the temporal availability of alcohol are going to pull levels of violence down and really interestingly we know that minimum unit pricing has been modelled to actually improve health outcomes for the lowest SES groups to a greater degree than any other group so there's potential that the same might be true around these kind of social harms and that's something we think is really pressing to investigate and we also think that this might kind of give us cause to reflect on wider criminal justice approaches to this kind of alcohol-related violence and harm particularly as I think our findings really suggest there's something structural going on here that there's structural drivers behind this violence that may be kind of approaches to policing about alcohol violence that we've seen before like policing in the nighttime economy and and focusing on individual people or premises might be not necessarily the most kind of impactful root and actually thinking about these measures like pricing and availability might yield much better results especially if we're talking about violence in the home that's often not reported and the you know traditional policing approaches might find might struggle to kind of get out and I think although pricing and availability interventions are often thought of as kind of national level policy and they certainly are often implemented that way they can be done at the local level as well and with the support of police forces so there are routes through the licensing system that can get these things in place like the reducing the strength scheme that's been really positive in some areas in Suffolk which kind of took the cheapest most high-strength alcohol off the shelves in off-trade premises. I also wanted to take this opportunity to kind of use the wisdom of the room I suppose we're using these anti-social behaviour survey items for the first time and obviously the format's slightly different to the violence items in the survey that's why instead of making kind of a more traditional incidence rate we made this frequency measure that I presented but I wonder if anyone has more experience with these survey items and whether there's any other considerations we should keep in mind as we finish off our analysis and then finally we've also been reflecting a bit more about the terminology we're really aware that ASB obviously has kind of a political history and it is one tied up with the stigmatisation of lower SES groups and young people and as this work is you know trying to kind of address these socioeconomic inequalities in these really kind of impactful experiences we're just really aware it would be unfortunate for it to kind of feed back into that and contribute to that stigmatisation more even though we're talking about we're talking about victimisation obviously not perpetration but we're aware there's kind of new political interest around ASB incidents so it might be kind of on policymakers minds and that's made us think about whether we ought to kind of frame our findings slightly differently in terms of the terminology we're using obviously when we're talking about survey items we're using the wording that's there but whether kind of these findings should be presented in a way that's kind of separated from that so something around maybe kind of you know environmental wellbeing or something like that I don't know if anyone has any thoughts on that but I'd be really interested to hear it and yeah any insight anyone has would be really appreciated. Yeah so that's everything these are the papers if you'd like to read them and like I said I'm happy to chat about any of this work now or later thank you. So our next presentation our next speaker is Nadia Butler. Nadia is an early career researcher from Public Health Institute Liverpool John Moores University and her research interests are in interpersonal violence and particularly violence against children she undertakes original research including population survey analysis systemic systematic reviews and data synthesis and program evaluation studies and works on local national and international research projects including work conducted as part of the WHO collaborating centre for violence prevention. So Nadia if you want to just say hello and get your slides up I'll hand over to you. Great thank you very much Catherine I'll just see if I can share. Okay thank you very much. Okay so my name is Nadia Butler I'm a public health researcher at Liverpool John Moores University and this is the study I did with my colleagues Professor Zara Quig also at John Moores I'm Professor Mark Ballas at Public Health Wales and we were looking at the association between childhood abuse and risk of re-victimisation as an adult using the current survey for England and Wales. So just briefly just a bit of background so interpersonal violence is a major public health issue it claims over 1.3 million lives globally each year but actually non-fatal forms of violence are even more prevalent than homicide and have serious consequences for lifelong health and well-being so for example 35 percent of women are estimated globally to experience either physical and or sexual intimate partner violence or non-partner sexual violence in their lifetime and addressing the physical and mental health consequences of non-fatal violence can impose substantial burdens and costs on health services and so for the year 2015-16 the estimated cost of non-fatal interpersonal violence to the NHS and other healthcare providers in England and Wales was over 1.8 billion so interpersonal violence is very high individual societal health service costs and represents an important public health issue but an improved understanding of the associated risk factors is actually crucial to informing effective intervention strategies to prevent, reduce and respond to interpersonal violence but using a life course approach or perspective there's actually evidence to support the concept of re-victimisation and that is that individuals who experience victimisation in childhood are at increased risk of a range of adverse outcomes across the life course including violence re-victimisation as an adult but historically actually much of the focus of research was on outcomes associated with child abuse in isolated individual types of child abuse they should say sorry in isolation so a review of studies in 2001 found that only a very small minority of published work up to that point includes a consideration of more than one type of abuse and early studies on re-victimisation were similar so they tended to focus on just one specific type of child abuse and its relationship with a particular often similar form of victimisation in adulthood so for example you had studies looking at child sexual abuse on its own as a risk factor for increased risk of sexual assault in adulthood or similarly looking at exposure to domestic abuse as a child and being associated with being a victim of domestic abuse as an adult and exposing future children to parental conflict also but this type of this body of research really implied the types of abuse occur individually however more recent research has demonstrated that child abuse and other types of adversity in childhood typically co-occur and consequently these type of studies which attribute risk of re-victimisation to just one specific form of abuse may actually be misleading and over the past two decades there's now been increasing attention and is focused on the contribution of co-occurring adverse childhood experiences so looking at lots of different adversity together and their contribution to a host of poor outcomes across the life course so ACEs incorporate a range of stressful and potentially traumatic events during childhood that they directly affect the child so they might be things like verbal physical or sexual abuse or or also might affect the environment in which they live so things like domestic witnessing domestic violence or living with an adult with substance misuse issues or mental illness but critically these studies have demonstrated that adversity in childhood can affect individuals right throughout their life and they've been found to be one of the strongest predictors of core outcomes throughout the life course and so in areas like education, employment, health and well-being but relevant today also a strong predictor of involvement in violence either as a victim or as a perpetrator and the ACE studies have been useful and for the framework that they've kind of provided this shared understanding of the impact of adversity on a wide range of public health outcomes and in a lot of countries this has been instrumental in driving national policy changes around responding to adversity in childhood so one of the most crucial points or contributions of the ACE literature is about establishing that the quantity of stressors during childhood so the ACE count as it's known is an important predictor of many poor behavioral and health outcomes across the life course and these ACE studies demonstrate that adversity operates in an additive manner whereby there's a positive relationship between exposure to higher numbers of ACEs and then increasing risk of a range of poor outcomes so this is findings from a national household survey in England that was done in 2013 and it looked at a range of different outcomes I've just picked the violence related ones here but you can see the incremental increase in the prevalence of both perpetration and victimization as the number of ACEs increased so two percent of people with no ACEs had experienced victimization in the past year and this rose up to 16 percent amongst those who had reported four or more ACEs and these same relationships hold in multivariate analysis so compared to someone with no history of ACEs someone with one ACE was 1.5 times more likely to have experienced violence in the last year four times more likely amongst those with two to three ACEs and seven times more likely amongst those with four or more ACEs but in the UK there's actually been a limited number of these types of studies that looked at the relationship between child abuse and specific types of adult interpersonal violence so they tend to just look at just being generally being a victim of violence in the past year and there are some studies from the US which suggest the same strong graded relationship between number of ACEs experienced and risk of being a victim of intimate partner violence or sexual violence however they are limited because they use a non-representative healthcare sample so the rationale for the current on the aim of the current study so as I already said previous research has tended to focus on violence in general rather than looking at specific types of violence and adulthood or they've been limited by the use of non-representative specific samples like healthcare samples but crucially they've typically tended to focus either on the cumulative burden of ACEs so the ACE core provides this nice simple means of examining the impact of clustered adversities but it risks oversimplification by not considering the impact of individual ACEs so it just talks about the number it doesn't actually talk about whether there's differences between different types of ACEs on outcomes but then on the other hand you have studies considering the individual ACEs but omitting the impact of experiencing multiple forms of victimization and we know from the ACE studies that types of adversity are not only highly likely to co-occur but they also demonstrate the cumulative risk on outcomes so the aim of our study was to the primary aim was to look at the association between being a victim of child abuse and then violence revictimization in adulthood but specifically looking at intimate partner violence sexual violence and physical assault and then using data from a nationally representative general population sample and then given to date that the study literature tends to be split between count and type we considered the impact of both the type of abuse experience and also the number of types experienced on risk of revictimization and so we used data from the 2015-16 wave of the crime survey for England and Wales for that year there was a total of around 35,000 surveys completed with a response rate of 72 percent however our sample for our study is adults age 16 to 59 who completed the self-complete module that year so the total number of individuals in our study is around 22,000 so for that that survey was the first time the crime survey incorporated a module of questions which asked adult respondents whether they'd experienced abuse as a child before the age of 16 years so measures in our study we looked at physical abuse, psychological abuse, sexual abuse and witnessing violence between parents and we also created a derived variable that summed the number of types of abuse someone had experienced and then we categorized that as none having experienced a single type or multiple types. Demographics or covariates we included in our models were age, sex, ethnicity and deprivation and then as I said we looked at three types of violence victimization in adulthood so after 16 years and that was intimate partner violence since age 16 years, sexual violence since age 16 years and physical assault in the past 12 months. We used SPSS, we also coded childhood abuse, anyone with missing data was coded as a negative response sort of in line with previous age literature methodology and then we ran logistic aggressions which with each of the violence victimization variables as the dependent variable and for each of these we ran three models for each violence outcome while controlling for socio-demographics. So in model one we just included the socio-demographics and the child abuse count variable in model two we included all individual types of child abuse and then in model three we included both the child abuse count variable and also all the individual types of child abuse. So in terms of the number of people who are reported experiencing abuse in childhood 18.6 percent per sample reported at least one form of abuse and this was made up of 10.6 percent reporting a single type and 8 percent reporting multiple or two or more types of abuse. Individual types of abuse, 9 percent reported psychological, 6.9 physical and sexual abuse and 8.1 percent reported witnessing domestic violence. In line with previous research and the ACE literature specifically all child abuse types in our study were found to be significantly associated with each other and there was the strongest association between psychological abuse and physical abuse. So in other words there was the highest odds of the two of those types occurring together and the weakest association or lowest odds of sexual abuse and witnessing domestic violence occurring together. Just briefly some of the associated socio-demographics with experiencing childhood abuse. So females were more likely than males to have experienced at least one type and were also more likely to report psychological sexual abuse and witnessing domestic violence. There was also a significant association between each individual type of abuse and abuse count with age and the youngest age group having the lowest prevalence across all and abuse categories. Then interestingly the relationship between deprivation and prevalence of abuse differed by abuse type. So there was a higher prevalence of psychological abuse and witnessing domestic violence amongst the more deprived quintiles. But the prevalence of child sexual abuse was higher amongst the least deprived quintiles compared to the most deprived. Child abuse count was also significantly associated with deprivation with higher levels of deprivation associated with higher prevalence of multiple forms of abuse. I didn't understand that. Sorry. But overall the differences between deprivation levels were relatively small and abuse was prevalent across all quintiles of deprivation. In terms of ethnicity there was also significant associations across all categories with white respondents more likely to report abuse types than other ethnicities. In our sample the proportion of adults who experienced each adult child violence victimization percent is here. So around 19% having experienced into a partner violence in their lifetime, 13% sexual violence and 2.4% physical assault in the past year. So we then looked at the association between the number of types of child abuse experienced and violence victimization and adulthood. So this is the prevalence of each type of violence victimization and adulthood amongst those who experienced no childhood abuse. We then see an increase in prevalence across each type for those who have experienced one type of abuse but experiencing multiple types of abuse associated with the highest prevalence of each form of violence victimization and adulthood. So then we controlled for age, sex, ethnicity and deprivation and compared to individuals with no childhood abuse those who had experienced a single type were 3.3 times more likely to experience into a partner violence since age 16 and 6.4 times more likely to have experienced into a partner violence if they experienced multiple types of child abuse. Similarly for sexual violence so again after controlling for social demographics and compared to individuals with no childhood abuse those who had experienced one type were 3.6 times more likely to experience sexual violence as an adult and 6.9 times more likely to experience sexual violence if they experienced multiple types of abuse in childhood. In terms of physical assault for the past year 2.2 times more likely to experience physical assault if you'd experienced one type of child abuse and up to 3.4 times more likely if you'd experienced multiple types of abuse. So a similar kind of dose response relationship that as was seen in the previous ACE studies. So in my very analysis the prevalence of each adulthood violence outcome was higher amongst individuals who'd experienced each type of abuse compared to those who had not experienced that type but even after controlling for socio-demographics each individual type of abuse was significantly associated with all of the violence outcomes. So if we just think back to that previous literature our body of literature which used to look at each of these types in isolation as a predictor for one of these forms of victimization as an adult that may be wrongly attributing risk to just one type of abuse as all of them are in fact associated with violence the three violence victimization variables. We then looked at what would happen if you entered all of these types into the same regression and control for socio-demographics. So when all individual types were entered simultaneously into the model relationships or odds ratios were slightly lessened but the majority of individual types remained independently associated and would increase odds of the three violence adult outcomes. And then finally in the last model we added child abuse count into the model along with all of those individual types of abuse to see if individual types of abuse were still predictive of adult violence outcomes after we accounted for the fact whether someone was a victim of multiple types or not. And we actually found that each in adulthood violence type was associated with different types of child abuse. So into my partner violence after accounting for whether someone had experienced multi-form victimization into my partner violence remained associated with psychological and physical abuse sexual violence remained associated with psychological and sexual abuse and physical assault was associated with psychological abuse only. So just consistent with early research examining the types of abuse individually shows that all types are significantly associated with each of the three adult violence outcomes in the current study and crucially but consistent with more recent research our findings also show that correlation and co-occurrence of abuse types and that provides evidence that multiple types of abuse should really be considered in both research and practice. And so studies that just look at a single form of abuse could be kind of erroneously attributing that increased risk of re-victimization as the unique impact of just that one type of abuse under study. We also saw that when all individual types of abuse were considered together in the current study most relationships with adult violence outcomes were slightly lessened but they remained significant and this suggests that child abuse in general regardless of what type of experience does increase the risk of re-victimization as an adult. And then consistent with the eighth literature the current study also demonstrated that cumulative impact of child abuse on the risk of violence re-victimization. So whilst most abuse types were independently related to physical assault into a partner violence and sexual violence the strongest predictor of each adult violence outcome was experiencing multiple types of abuse in childhood. And then while abuse in general regardless of type was associated with increased risk of re-victimization particularly for those individuals as they said who experience multiple types of abuse. The current study did expand on previous research on re-victimization by demonstrating that when you control for the effects of whether you've experienced multiple types and the type of abuse is still predictive of adult violence outcomes. And so it suggests that the effects of particular combinations of different types of abuse may be obscured in the sort of eighth literature when each type is given an equal weight and they just use that count approach. So kind of crucially both the type of abuse and the number of types experienced seem to be important and should be something that is considered in future research. And just to finish to say that interpersonal violence is really one of the most preventable causes of premature morbidity and mortality and obviously a key target of sustainable development goals. And evidence from our study suggests that preventing child abuse will also have that downstream effect of preventing interpersonal violence across the life course. And adult violence victimization is likely to kind of add to the trauma or compound effects of childhood abuse. And so this kind of makes the case for focusing public health efforts on preventing childhood abuse and disrupting those intergenerational cycles of violence. Thank you very much.