Webinar presenter Jamie Roush answered a number of your questions after her presentation, You Don’t Need More Data, You Need the RIGHT Data. Here are just a few of her responses.
Audience Question: What are ways to overcome HIPAA barriers when working to collect data to demonstrate the need for increased funding for programs to reduce involvement with the criminal justice system?
Jamie L. Roush: That’s a great question, I think a really critical element that’s here. So, any data that is HIPAA compliant really should reside within the organization that is facilitating that the data collection mechanism. So, in the example that I gave earlier, HIPAA data generally will fit in a mental health context within the mental health intake facility, and so it’ll be shared among the mental health community, and again, those people who have the ability, the right and privilege to see and access that particular data. I don’t recommend taking the elements of PII or personal information, and the things that are under HIPAA, and moving them over into the criminal justice world. I don’t recommend you creating a data-sharing agreement that says, “Hey, mental health facility provider over here, just share your data with us.” Again, you get into too many barriers associated. It is a violation of HIPAA, in many cases, for organizations having access to it. They don’t need to. So that’s where the example that I gave earlier is about stripping the things that are PII-related. f a mental health provider is willing to give you their data associated with it, and they can: don’t take any of the name information, don’t take the date of birth, don’t take the exact address of those particular individuals that are on their case files, don’t take the data that can pinpoint to an individual, which then becomes a violation of HIPAA. But do take data, associated in individual records with an age, for example. So, if, again, you’re looking for somebody who’s in there, you know, 38, as opposed to somebody who has a date of birth, of, you know, 11-14-1982. So, again, those are the kinds of things to consider when you are thinking about that. So, as you aggregate the data as you look at it at the zip code level, as you look at it at the age block range, as you look at it across, male or female. Maybe you’re looking at it against ethnicities. As you have that data, then build your metrics for your programs around the aggregate data. So, we want to improve, this program is designed to improve people who are within the age bracket of 30 to 35, or we want to improve the mental health response to women over men, for example. So, look at your metrics, as you see associated with that aggregate data, and focusing more on the aggregation of the data, which will give you the ability to evaluate the programs effectively, still contribute to that research and the evaluation to use in evidence-based practice. But really protect that data the way it is intended to be protected.
Audience Question: Crime statistics and officer activity data sets are commonly relied on as measures of police effectiveness and productivity, both internally and externally. But this mindset can perpetuate aggressive and discriminatory practices. So, what data can be captured to measure positive outcomes and use to incentivize officers, as well as educate the public? Fantastic question. Let me know if you need me to repeat any part of that. I know it was kind of lengthy.
Jamie L. Roush: I think you’ve hit the nail on the head. You know, part of the problem is, is that we are using false measures to evaluate performance and effectiveness. I think, really, what most organizations need to be focused on is, you know, generally what are your problems, and what are your issues, and then how can we respond to those in a way that is understanding of the community that what you’re responding to. So, for example, looking at things like how much time you spend in a particular hotspot that’s been identified as a particular problem for X topic. How many discussions have you had with social providers about mental health-related issues or homelessness? How many meetings have you attended associated with the coordination or collaboration with homeless providers, or mental health and substance abuse? These are very alternate measures, but they are designed to continue to move these conversations forward and allow for people to see that it’s not just about going and knocking on people’s doors, or doing traffic stops, or writing field investigative reports, or writing field intelligence reports on individuals’ that happened to be walking through our community. But ways that you can measure, how people are really addressing problems differently? And I think when it comes to things like mental health, it is, it’s having, it’s allowing police officers, criminal justice individuals, whether that’s probation/parole or victim services, and looking at, are they having real conversations with mental health providers? Are they attending meetings about these topics? And using those are the type of metrics to say, yes, this person is, from a performance standpoint, they’re doing what we want, which is finding alternate ways to deal with these particular issues, finding ways to be collaborative, and finding ways to shift and share the problems for public safety issues, and it’s not just about those individual actions that that person takes every single day.
Audience Question: What suggestions do you have for outside organizations making recommendations to law enforcement agencies for improved data collection? I suspect you have a little bit of insight on this one.
Jamie L. Roush: Yeah. I think this is a really important question. I think, you know, this is going to require both sets of organizations to understand that there’s a level of expertise about the type of information that’s there. So, I think outside agencies really have to kind of put it in the context for our law enforcement agency. So, right? An outside entity is not going to be able to tell a law enforcement agency necessarily how to document tattoo information at a very high level that exists within a correctional facility, right? A correctional officer is going to be the person who has the most view of that individual and being able to document that from not only what the type of tattoo is, what it actually has, what part of the body it is, all of those types of things. Similarly, law enforcement agencies, on the other side, need to recognize that they don’t have all of the answers when it comes to what data is necessary to help solve some of these problems that these outside providers actually are or the professionals in. Right? So, it’s the sharing of knowledge associated with that. So, I think everybody has to come to the table recognizing that they have their level of expertise. And that, that is where you kind of share data together. Generally speaking, and this is more of a broad brush here, as is that I’ve found, that you have to show people why the information that they’re collecting is not necessarily valuable to solving the problem and why the additional information as necessary. So, I’ll give you a really good example. I was very famous for starting to request for police officers to put in the provider of a cell phone. And a lot of officers on the street don’t understand why that’s important. It’s not necessarily part of what they would think about at the time. And so one of the reasons for that, though, was because they didn’t understand that from an investigative lead standpoint if you know that a cellphone writer starts with AT&T or Verizon if you have a starting place over an investigation that can help you associated to that number. So sometimes you just have to show people the value of what that data represents to you as the outside entity in order to really get them to contribute that information. It is a relentless pursuit on both sides. I will tell you, though. It is not a quick fix or solution. You probably will have to go over why that data is important, and probably do additional training for that hour on a repetitive cycle, so that everybody does understand the value associated with it.
Audience Question: Over the last several years, many organizations have moved away from CompStat type meetings. Can you talk about where agencies are going in terms of operating operationalizing findings from analytics?
Jamie L. Roush: I think this is a great question, and, yes, organizations are moving away from that CompStat model for a lot of different reasons, but I think they are replacing it with a more problem-based model. And what that is, is looking at what are the pervasive issues that those organizations are facing, and then having more constructive conversations that are non-confrontational about how we actually solve those particular problems. This is a great place where evidence-based policing comes into play because it’s, again, the ability to say, this is our problem. What does the research tell us about this particular problem? And then, how do we actually go and implement the strategy that is associated with what that problem is. And then what’s evaluated through the data, as well. And, so, you are going to see the future here is that data and analysis are not going to be about producing statistics as we move forward. It’s not going to be more about, are you up or down in crime? It’s going to be looking at the normalization of crime and looking at where you might be above or below that of the crime levels, which, again, is going to be threshold-based reporting. And you’re going to see data, and analysis, and analytical staff, being used much more about evaluation. And not evaluation is not just going to be a quantitative one. It’s also going to be a qualitative one. What other data measures should we go get from our partners? Maybe it’s survey data. Maybe it’s interview data to be able to actually paint a picture of whether or not what we’re implementing is successful and effective. So, as CompStat rolls out, the more statistical model that’s associated with it is being replaced with a problem model, and data and analysis are taking a larger step and role in the evaluation of addressing those particular problems.
Audience Question: What advice do you have for what training or background someone might need to work on the data and analysis side? In other words, what makes someone successful in looking at data holistically and creatively?
Jamie L. Roush: So that is a great question. I probably spend the majority of my career actually doing that and training people on that particular topic, but I think people who work in data analysis, especially with criminal justice data have to, number one intrinsically be curious people and want to just go dive in and just see. So that’s the first thing, curiosity is, is the very first skill that’s necessary in doing that type of work. But also, really looking at the technology aspect of it. So being able to do querying through Excel and Access, learning SQL, understanding modeling associated to that, you heard me highlight the importance of geospatial data. So, the future of business operations is a critical element in criminal justice data. So, the more you can learn about geospatial technologies, especially when it comes to mapping this data, the nuances associated, and the caveats associated with the mapping of the data, that becomes really important. And then just baseline analytical techniques. There are some great organizations out there that have analytical technique training, from the International Association of Crime Analysts to the International Association of Law Enforcement Intelligence Analysts, IALEA. You will also see, um, there are organizations associated with training and standards in law enforcement that have training as well. But there are some really good just analytical methods at trainings that exist out there as well. I’d encourage you to contact me offline because I can give you a list of some of those resources, as well as some places you can go get some free training associated with it.