 And I would like to just say that I am really looking forward to this session and from hearing from both, to hearing from both Bev and Melanie. I've worked with them both and I can say they are incredibly helpful. They're insightful and they are more than willing to share their time, their knowledge and their experience. And I personally found that incredibly beneficial over the years. And I think that what's interesting is that they're not just from Victoria. So Melanie has experience over in Adelaide as well in this similar field. And Bev, as for some of you who might have heard our chat earlier, has a bit of an accent. So I'll leave you to guess where she's from, but she brings in a little bit of international knowledge as well and perspectives. So Bev is a very experienced data analyst and a family violence researcher. She is currently the lead data analyst with evidence and insights at the Department of Justice and Community Safety. And she is responsible for what is now a fairly well established Victorian Family Violence Database and the newer COVID-19 Family Data Portal, which is what she'll be talking us through in particular tonight. Melanie is the Assistant Director of Research and Evaluation at the Department of Justice and Community Safety. And she has a background in both crime and criminal justice research and evaluation. And she leads a program of work on data linkage, analytics and forecasting to support both policy and policy program and evaluation. And she's been with the department, I think Melanie for almost seven years now you can correct me if that's wrong, but has some very valuable knowledge, I guess around the intricacies and the challenges of some of the data that is available across government and accessing that and using that to its full potential. And so with that quick summary, I will now hand you over to our two presenters. Thank you very much. Thanks Kaz. Bev is sharing some slides that we've put together hurriedly for this presentation due to busyness, but we, I guess, wanted to start by saying that we're going to talk in detail about a data product that was developed by Bev and her team. And that's going to be kind of the main event, but I'll start off by just giving a bit of background about our business unit and some examples of the kind of work that we've done, particularly with a view to talking about how we've used data to support policy given the topic of conversation. And I formally had an evaluation role, I don't currently have an evaluation role, but a lot of our work, we are very quantitative, quantitatively focused. But a lot of our work sort of feeds into evaluation and can be quite useful for evaluation, which I thought would be worth mentioning given we're sitting in a room full of evaluators so I will touch on a couple of points where I can see there are like some obvious links with evaluation work. But I guess I'm also acknowledging that this is not really a presentation that's solely about evaluation, it's primarily about the data that we have and how we use it. And then having a look at the development of a data product that evolved in response to questions we were getting about the pandemic and the impacts of that pandemic on family violence. So having said all that, we both work in a business unit within the Department of Justice in Victoria called Evidence and Insights. Evidence and Insights contains the Victorian Crime Statistics Agency, which is probably where all of our public facing data and information sits. And that's also wherever we can, we do try to publish our work via that channel. And it also has stacks and stacks of data on it. So if you want to have a look at our website feel free. So the Crime Stats Agency commenced operations back in 2015. It has a legislated function and is responsible for processing, analysing and publish Victorian Crime Statistics. And importantly, that role is independent of Victoria Police. So although we sit within the Department, we do have an independent function, which means we have a mandate to use Victoria Police data and to provide it to the public. So strategic objectives are to improve the accessibility of crime statistics for all Victorians strengthen the integrity and quality of recorded crime data and instill public confidence in crime statistics. And that was particularly important when we first came on board because prior to that, I mean, part of our reason for being is that there was some controversy around whether Victoria Police were accurately reporting crime statistics. And so Victoria moved to align itself with a lot of other states that had a kind of independent reporting function. To build an evidence based to support decision making and policy development and to provide tools that improve the statistical literacy of stakeholders and clients. So the Crime Statistics Agency Act establishes the role of my boss or our boss, the Chief Statistician to publish and release statistical information. And to take research into an analysis of crime and just crime and criminal justice trends and issues, issues and trends. So that's kind of our public facing a public facing functions. However, we also have a bit of an internally facing work program, which I'll run you through in a couple of slides. This is just an overview of some of the public facing stuff we have stacks and stacks of data available via our website if you're doing very sort of high level policy analysis or evaluation, some of that data can be useful. And, you know, you can look at things that might have changed over time in terms of new offenses coming on board and how that might have impacted the crime rate and those kinds of things. We do do a quarterly publication of broad range of statistics. And we also do customized data requests for the public government media and academics. So we think we do about 1000 data requests a year. I think it's slowed down a little bit the more data we put on our website the less request we get which is great. But we still do get a lot of requests for data. So that's available to anyone. Small data requests are free of charge and then I think there's a very small charge if there's sort of an estimate that it will go more than that it will take more than a couple of hours of work. So if we move on to the next one. The other thing that we do is we use the data that we have to do research and publish research on topics of interest for Victoria and topics of policy relevance. So, for example, and we also have done some in that sort of vein we I wouldn't call them full evaluations because I know I'm in the room of evaluators but we have done some what I would call statistical outcome evaluations. We're using the data that we have available to us so we kind of want to maximize the value of it so an example of one of those four would be a study that we did a while back about the impact or the differences between the giving a young person who's committed an case have an option to give them a caution or like a warning, or they have an option to charge them with a crime and they go to court and they go down that route. So we did an evaluation, just a statistical outcome study comparing reoffending outcomes for kids who are given a caution versus reoffending outcomes for kids who are charged with a criminal offense. So in terms of our other publications, I think there's an example there. And it just shows some analysis we did looking at the criminal trajectories over the first seven years of a young person's offending. And we identified several different types of groups, and the vast majority of young people who are offenders are very, very low level offenders and don't sort of go on to be involved in a lot of crime. So that kind of research that we do informs, I suppose, the different types of policy responses that might be required for different subgroups of young people. Next slide, Bev. Oh, I don't think we were going to do that one. Oh, is this the latest version? Sorry. I think it was our shared version. Okay. Unless it's not updated. All right, I'll just talk about it a bit there. James has just asked if you can also just put it into presentation view it's just a bit, I think, small for some. Oh, yeah, I can give it a go. I'm not used to teens so we'll see if, if it still ends up casting all right. Can you see it okay. I know that's the presentation view. You want to watch the full screen view. Yeah, it's because I'm sharing a particular I've got two monitors and I'm sharing a monitor and I want to stop the display. If you just. Oh, there we go. Yes, we can see now. So before Bev moves on to this slide, I'll just briefly say that a couple of other things we do bespoke sort of research and analysis pieces using linked data to inform policy so an interesting kind of example of that that we've done recently is informing the government's response to the repeal of public drunkenness in Victoria. So we use the police data to inform the design of a service that would replace police responding to public drunkenness as a criminal offense and instead would provide a health based response so we essentially looked at all of the data we had on people who'd previously been recorded by police for public drunkenness looked at what their service needs might be what their characteristics might be where they might get arrested by police for public drunkenness and that's been used a lot to inform you know what capacity do we need to have for services in different areas. So what are the likely needs of the people coming into contact with those services. And in that work, unsurprisingly probably we found that from the data to cohorts of people emerged there were a group of people who you could essentially say we all like people who are out on a Saturday night. They're only ever recorded by police for a singular offense they didn't tend to be recorded for other offenses, and they didn't tend to have a lot of service needs. So we found what we called our high intensity cohort and these people were people who were arrested several times a year for public drunkenness were also in contact with mental health services, also in contact with homelessness services. But, and we looked at sort of where police were finding those people across the stage, and how that might inform what they needed to design in terms of an alternative service delivery response so that was just one that I thought was a good example of how our data has been used to inform policy. But having said all that the other thing that we did at the crime stats agency is release the Victorian family violence database and manage that so I will hand over to my colleague Beverly who is the expert in all things family violence data, and let her run you through that. Thanks Melanie. And so as Melanie sort of alluded to the project that I work on full time at the crime statistics agency is the family balance database. I've been working on it for the past four years, and the coven 19 family violence data portal is a component of that work. And that's sort of what I'll be focusing on tonight, but I wanted to give a bit of a background about what the database project is, and sort of how it came to be and came to land at our agency and then we'll talk a bit about the coven 19 portal which is part of that work. So the family violence database actually has been around for a little while, and it's sort of its origins come from a Victorian government into interdepartmental committee on violence against women. And they first identified sort of a need that to create a database which would help facilitate a system wide view of family violence data in Victoria. So way back in 2000, they began to undertake some of this work thanks to some seed funding from the Commonwealth government's partnership against domestic domestic violence program. The very first publication of the database came out in 2002. So we're coming up to our 20th year anniversary of publications. But from that first publication, the subsequent work was a bit sporadic, as you can see, so it took another four years for another publication to come out. In 2007 a third edition came out, but the project was relocated to the then Department of Justice. In 2007 in 2012, a few more publications came out. But after the fifth publication in 2012 it was, you know, then silent for a number of years, including up to the Royal Commission into family violence which took place in Victoria in 2015 and 16. So the Royal Commission noted the value of the database during during their hearings and in fact made a recommendation to continue that work under a more sort of routine and regular funding arrangement and publication cycle. So one of the recommendations of the Royal Commission, I believe it was 206 was for the family violence or for the crime statistics agency to undertake that work. And since 2016 the project has landed in our office and it now exists on the crime statistics agency website as a sort of umbrella term resource for a number of publications that we update annually. So what is the purpose of the database. I've sort of summarized it very briefly on that reliable and accessible data helps to create informed policies. So there was a recognition, including by the Royal Commission that the database helps bring visibility to trends data gaps and system weaknesses through cross sector analysis and because of that it's a valuable tool for developing evidence based policy around family violence. Really, in its rudimentary form it's sort of an aggregation or a collection of family violence data from a number of different sectors across the VPS. And so just having all of that information and one easy to access places a lot, or is a big help to people in the research and evaluation and policy areas to just access data more readily. So the quote that I've taken from the Royal Commission, which is just says our understanding of family violence and the response to it could be improved by making existing data sets and research efforts more accessible, conducting further analysis of existing data, increasing links between data sets and where necessary, augmenting existing efforts. And so those are really some of the goals that we're trying to achieve with the family violence database project with the view that making data accessible in turn will then help a number of the different research and evaluation policy areas that need to access that data. So this is where the database lives on the crime statistics agency website now you can find most of our work under the heading of family violence data portal. The bulk of our annual work will fall under the data dashboards and data tables. So this houses data from about 10 different sources across the Victorian public sector. And it's updated annually, and we report on financial year figures so you'll get sort of aggregate totals by financial year of a number of different measures of family violence. We also have interactive infographics which you can augment to look at different local government areas in Victoria. And so that's her information from a few of our data sources, including Victoria police. We've produced a few research publications which you'll find in the research and evaluation area of our website. And so we've got two that have come out in the past few years, one on adolescent aggressors of family violence, and one on child witnesses of family violence. We've called them data snapshots and they're really just a high level look at some of the characteristics under those topics. And of course today, we've also produced the COVID-19 family violence data portal. And so that's what I'll sort of be focusing on for the rest of the talk. Um, the COVID-19 family violence data portal. It was developed as a component of the family violence database, really just meaning it's the same team and same funding that's responsible for the database that in turn produced this sort of specialized product. So we hope in response to an identified need for accessible, accessible data concerning family violence incidents in service use during COVID-19 and the related restrictions. So if we think way back to March 2020, most of us were sort of being sent to work from home. There was an early recognition that for victim survivors of family violence home isn't necessarily a safe place to be, and all of us were being asked to spend a lot more time in the home. So right at the beginning of the restrictions there was, I think a warranted increased concern about the incidents and service needs for people affected by family violence and whether those needs would be met. So we realized that there was in turn a need for more visibility around the data concerning family violence. And as I've mentioned, our regular publications are only updated once a year and we really focus most of our analysis within financial years. So the tools that were existing in the family violence database weren't really going to be sensitive enough to provide a focused look at the impact that COVID-19 and the related restrictions were having on family violence incidents and service use. So really the process for us to create the portal. We've got a governance committee for the family violence database work. So we created a project proposal and it was accepted by our governance committee. When we got the go ahead to commence this work we then developed a report and access framework through consultation with a number of our stakeholders. So here's the fancy way of saying we went around to a lot of the people that were asking for this information and tried to get a sense of exactly what they'd like to see in a data platform. So a few themes emerged from that exercise. I think there was a really big interest in how priority communities were affected. And when I say priority communities that's really referring to a number of different cohorts or groups of people that we know to be data gaps in in relation to family violence. So a priority very much from a data sense in that we don't know a lot about how family violence affects these groups. And so that includes communities like children and young people, people from culturally and linguistically diverse backgrounds, Aboriginal and Torres Strait Islander communities, people age 55 and older people with a disability and so forth. And so a lot of information in particular about how those groups were affected. And also especially in the early days a lot of interest around motive access to services. So a lot of service delivery of course traditionally in the family violence, some space is face to face. But as the restrictions were amping up, of course, the ability to provide face to face service was impacted. And so there's a lot of interest around how help lines were affected by increased demand on remote based work. And in addition to that how how website use was going up use of web chats and quick exit buttons and things like that. So a lot of interest in some of those sort of non traditional format based services. So an interest in sexual violence and whether or not that was increasing amid some of the restrictions, particularly within the family violence context. One of the things I think we've seen in some of the more severe restrictions is that some of the traditional crimes that might happen in the community were then still taking place in some instances but instead violence directed in the community. So I think that was the basis for some of the interest around sexual violence. And this last one is something that emerged particularly as we were doing analysis on the data we received but there was an interest and even contextually around disruption to traditional referral pathways. So, what I mean by this is within the family violence sector broadly there's typically a few places where victim survivors or perpetrators might first be identified and then moved into the service sector. So there are of course a huge eight keeper, but a really interesting sort of case that submerged, particularly in its coven for children would be a large source of referrals to child protection is schools. And so, one of the disruptions we saw early on as schools closed was the number of reports made to child protective services went down. And this just become interesting because sometimes we might look to administrative data to give us an indication sort of of community prevalence of certain types of abuse, but we also have to contextualize that with what's happening in the background and if some of these referral pathways are disrupted it's of course going to have an impact on some of the administrative data we might look as as an indicator of prevalence. So it's important contextual information to keep in mind and doing analysis on the data. In terms of continuing developing the portal we then had to engage with data providers. In the end about 20 different data sources agreed to contribute to the coven 19 portal, which really was probably the biggest component of the success of this project. I really was quite pleased that so many people were not only happy to share their data but we're really keen and excited to get on board, and it's a product like this wouldn't have been possible without the support of a lot of the different sector workers who were happy to provide us with data to support the project. So this was a stage of development with the actual product build which we did in power BI, and was then published the CSA website. Some of the challenges that emerged while we were building the data portal and perhaps this will be relevant if any of you go on to tackle a similar project. One of the early challenges we faced was a need to keep the information in a public facing platform quite simple and consistent and digestible without being misleading and just keeping in mind that we've got about 20 sources. So, easier said than done we learned, because data of course comes in in all kinds of different formats, some information isn't able to be broken down in the similar way some sources will define family violence differently. Some will have, you know, certain disruptions going on. I think a big takeaway from that is just the need to frame information with a lot of contextual details and explanatory notes as well which we tried to do in the portal. In addition to that, for those of us in Victoria will be well aware that it was a rapidly changing environment. We began this work in March, and it was published in November and a lot of things happened in Victoria between those times. And, you know, changing priorities came about, even just in terms of what we would have found interesting in the data, which was a bit of a challenge and sort of links in with tight timeframes as well. So, there was certainly a demand for very timely data, and while quarterly data under regular service circumstances is, you know, pretty impressive in Victoria at that time when restrictions were changing so fast. By the time we published data, you know, ending in May if we are June if we published it in November it already felt quite still because we experienced just so many changes in between that time. So, certainly some of the challenges that we faced. But I think overall we've been able to achieve, you know, something more robust than what we typically produce on an annual basis. Just to go over some of the general things that we found along the way so the coven portals existed for three cycles now so we've done three quarterly updates, and I'll give a very broad overview of some of the things that we found when looking at the different quarters. And just to refresh everybody's memories. So this is what we found between March to June. And around that time is when the very first restrictions began to emerge in Victoria. So, on the stage one to three restrictions took place rather quickly, first school closures took place around that time. And then venues, we all started working from home in March, or many of us did I should say, operation ribbon which was an operation that Victoria police undertook to increase proactive monitoring of known prolific family violence offenders commenced in April. And then in May restrictions began relaxing so schools reopened in May 26 and restaurants reopened early June. So there's a lot going on in this first quarter, both in terms of amping up and winding down restrictions. And so really trying to look at the quarter as a whole was a bit difficult, but we found some trends more looking at month on month. Many of our sources at that time, particularly emergency sources saw the largest growth in cases when compared to the prior year in June. So there is a bit of a spike, particularly in victory police data, as well as a number of our help lines as well. In a separate paper that the CSA undertook, we did a bit of forecasting analysis and we found that that June spike was significantly different than what we've seen in the past. So certainly interesting changes happening as the restrictions were winding down in June. We also found that a number of breach of border offenses were taking place at that time. And another theme that sort of emerged in this quarter was the growth in our cohort of people aged 55 and older. So a number of our sources recording actually saw an increase in victim survivors aged 55 and older. For a police data that cohort grew by about 19.4%. And actually, the co or the subset of people in that cohort whose perpetrator was a child, including adult child of the victim grew by about 50% in June. So it was a big growth, but just to contextualize it, the growth of that cohort was from about 1100 people to 1700. So it was a growth of about 603 incidents when comparing 2019 to 2020, relatively small numbers still overall in a small subset of our Victoria police data, but we felt some. It was a interesting trend that emerged in that quarter, and, and certainly seem to be something that was consistent across a few of our contributing sources. In July to September to sort of refresh on the restrictions. So we returned to remote learning sort of commenced in August, and restrictions began to be reinstated. So this quarter ended up housing some of our most severe restrictions, and probably the most severe restrictions in the world. At one point I heard that. I think the state of disaster was declared in August, and that's where the stage four restrictions to place as well so those of us in the try and make well remember when we virtually weren't allowed to leave our homes at all. This is the quarter where a lot of disruptions were evident, particularly in court and legal and child protection services, although that's also true for some of the restrictions that took place in the March to June quarter. So finalized family violence intervention orders went down about 42% in August, and duty lawyer services provided by Victoria legal aid went down about 53%. We've sort of received device and I think it's been documented in the paper as well just as the transition to remote based court was taking place. That was having an impact in some of those official in person services, but the other side of that is that we saw an uptake of interim family violence intervention order so it's not necessarily to say that victims weren't being protected during this time. In addition to that, the number of investigations being brought to the attention of child protection went down about 40% in August as well, potentially related to some of the school closures happening around that time. The other services in this time period were a bit varied, unremarkable, difficult to say what the trends were in them. We saw an increase in some services, particularly Victoria police around August, but I think when we've done are some of our forecasting analysis on it some didn't necessarily see any remarkable spikes around that time. And then of course October to December. This quarter was recently released on our website as of last Friday. This is where we saw a lot of the restrictions begin to lift so roadmap to recovery I believe began at the end of September. So space learning resumed, I think the 12th of October, and throughout October to December there was continued easing of restrictions. So at the end of December there was outdoor gathering some and venues reopened with capacity restrictions. And also in this quarter saw evidence of recovery amid sport services and child protection services. So in particular for the first time, finalized family violence intervention orders met about the numbers that they were at in 2019 as of November so evidence that they're catching up to, I think what's been called the backlog in applications. Similar with child protective services and that we've seen an increase in the number of investigations commenced. And then some increased activity noted in October, particularly in Victoria police and again, some of these increases with the family violence related crimes seem to be sort of more clustered around breach of order offenses which which were increasing. So I'll now attempt to transition to show you the live portal if I can get my share screen working. Right, can everybody see this alright. So, as I've mentioned this is the COVID-19 portal that lives on the crime statistics agency website. I won't take you through all of the analysis but I'll just sort of show you the different components and I'd encourage you to take a look and have a play for yourself about everything that's living in here. So we've flustered our analysis sort of under three broad headings. So look at analysis by sector in this section we've sort of broken down all of our 20 odd sources into different sectors or types of services that are provided. We've got a little filter here where you can select which data source you're interested in. And there's a few different ways you can look at the trends. So this is just a month on month trend line for 2019 20 data, you can look also at percentage month a month percentage changes and do comparisons with the other sources that are clustered within the the selected sector. And we've got our key findings and then information about the data and links to related material about that source. We've opened the sectors down into emergency services, help lines, housing services, which includes homelessness data collection data, and request for accommodation. We've also got some data from VCAT, which is only broken down at a financial year level, but relates relates to least changes on the family violence sections of the residential tenancy act. Family violence protection services is just data that we've got about website usage. And then we've also got legal services which is family violence related services performed by Victoria legal aid. We've got information on family violence protection orders. So that includes family violence safety notices, which are specialized order issued by Victoria police, as well as interim orders, finalized orders. We've got the total number of people as at the last date of a month that are on any kind of family violence specific order. And then everything else that sort of doesn't fit into the clusterings above generally falls under specialist services so here we've got child protection data. We've got family and domestic specialist services that are funded through the homelessness data collection. And then we've got some other specialist services so in touch as a subset of our family and domestic violence services, but they're a specific service for people from culturally linguistically diverse backgrounds. We've also got safe steps, which has sort of been the 24 hour response throughout coven and a few other services this year's rights Victoria and victim assistance program. And other sources we've got just some breakdowns for federal circuit courts and family court data, as well as Department of Home Affairs, and we've got a little bit of data about family violence claims made for applicants or holders of temporary or provisional partner pieces. The second sort of umbrella sort of heading that we've clustered our analysis under is analysis by location. So this, you can look at a few different populations of data typically Victoria police or protective order data broken down by LGA if you're interested in LGA measures. You can look at month by month rates, or just the quarter analysis for the fourth quarter of 2020. You can also break the location based information down by a few different ABS structures. So remoteness area looks basically categorizes areas based on their remoteness or their accessibility to services. You can look at data that way. And we've also included a breakdown by the index of relative social economic advantage and disadvantage so that ranks different areas according to how advantaged or disadvantaged an area is. So you can look at sources that way. The third sort of like breakdown of analysis we did was of course for a priority communities. We included everybody, all of the communities that we wanted. In some cases, it's just very difficult to find reliable data on some of the communities were interested in. But we were able to put something together for children and young people, people aged 55 and older culturally and linguistically diverse communities and Aboriginal people. So we've got access to some data on those particular cohorts that you can explore in the portal as well. We've got a few different other pools of information if you're interested in exploring the portal as well. We put together some infographics, just highlighting some of the services and priority groups, some of the key data in those sections. We've also got a directory of Australian publicly available research and data resources that's related to COVID-19 and family violence or related themes. So if you're looking for some papers on a particular topic around COVID-19 that's publicly available in Australia, then you can check this out. We've got sort of research themes that you can check here and it will show you just the different papers we've been able to find on it. So that might be a helpful resource if you're interested. We've got a few other little downloadable resources. We've put together a summary sheet about all of the different restrictions that were in place in Victoria from way back in the beginning in March through until the end of December. We've also got just a fact sheet with some explanatory information on the data. So we've got a resource here to download most of the trend data that exists in our portal. So if you're interested in any of the trend data and you want to download it for yourself, you can do that in the portal in this area.