Webinar presenters Dr. Andrew Fox, Dan Carew and Karissa Taylor answered a number of your questions after their presentation, Social Network Analysis: An Innovative Tool to Maximize NIBIN Leads. Here are some of their responses.
Audience Question: What software are you using to do the link analysis?
Dr. Andrew Fox: I use a free software called PAJEK. There are a couple of free software that you can use. People are integrating SNA into their different analytic packages but I tend to find it's easier when it's free. There's another program called GEPHI — it's a free download with a good online help community.
There's a program ORA, Organizational Risk Analyzer, that does cost money now, but is a really good tool, you just pay for the license. It's a great tool for smaller networks under 500 or thousand, manages the data really well.
For some people who are getting more advanced in it, there's a ton of options where you can produce a user interface and manage the data. I know they've been doing some of that in Philadelphia.
Those are some packages that I know. All the networks you saw today were produced in PAJEK.
Audience Question: How long did it take to clean the data from the jurisdictions you're working with and conduct that into your analysis? Just approximate timelines?
Dr. Andrew Fox: We cleaned all of the network data first which took about 3 to 4 months collecting and cleaning it. Then we worked on the NIBIN data which probably took 3 or 4 months. A better part of a year, we've been working to build the data and clean it together. We finally got to a point when we had it all probably over the last year. To maintain it, it wouldn't be as challenging but to get it all together in the beginning.
The hardest part too was combining law enforcement jurisdiction data. Name cleaning and making sure people are the same people and not separate dots. So, we've been working on this on a better part of a year.
Audience Question: You mentioned the NIBIN data, was that provided by the ATF, crime lab, or some other option?
Dan Carew: That was primarily coming from both the ATF and the State Crime Lab who actually does the comparisons. A combination of both of those two.
Audience Question: Can you share one critical lesson learned linking people across different systems? That person A correlates to person B… Can you talk about any of the other data combination lessons learned that you had?
Dr. Andrew Fox: If your agencies have any questions feel free to reach out and email me if you have technical questions. I'm happy to help with that. In terms of combining the data across jurisdictions, you just have to come up with a template. Almost all of these is done in Excel — come up with a template with what variables do we want, what parameters do we want to select, time period, crime types, role types. Do we want everyone to work off that same template to extract data and then you can combine it all to one Excel sheet and clean it from there once it's together? It's the approach that we've taken.
You have lots of different challenges within the jurisdictions. There are some that can't extract the data from their RMS or don't know the right fields.
That's just from my experience with lots of different agencies. Coming up with a common template and working from there.
Audience Question: In working with those other agencies, did you end up creating a formal MOU that defines roles, responsibilities and data sharing?
Dan Carew: Yeah. With each of our partner agencies that we're working on as well as with the ATF and the crime labs, we have MOUs in place about accessing the data and storing the data and sharing the data. That was one of the first components that we took on and also was useful in building out the initial relationships and trust and getting to know each other. We work together on an individual, case-by-case basis, with all the different agencies. But with data collection, it was a new relationship between agencies — so that's one of the first things that we did.
Audience Question: Who is actually involved in the analysis? Is it an analyst, an intel analyst, a detective? Who's actually grinding through the data once it sits in the system identifying the links and doing the follow-up?
Dr. Andrew Fox: This is a test analysis and we're able to fund Lexi Gill who was a graduate student through the grant that we have that primarily worked on cleaning the data and doing the analysis. But the Crime Strategies Unit within the Prosecutor's Office also has an analyst they hired as part of the grant as well within the Prosecutor's Office that does a lot of that data cleaning and analysis work. Raphael who works within the Prosecutor's Office. These are analysts that do the work.
Audience Question: Do you see guns with multiple NIBIN hits that get passed on to people outside of the gang? How often do you come across guns that have been stolen or sold after they've been tied to a NIBIN hit?
Dan Carew: One of the important things to keep in mind when talking about NIBIN is that it only links the gun to common incidents, it doesn't link the person to the incidents. Quite often, we see within the gang or group contacts that were referenced, we do see a lot of movement within a group or people or a gang of the firearms. We have seen some examples where the gun seemingly has, just based on our analysis of where the gun was used, who the intended targets were perceived to be. Who the suspects were that the guns were actually moving from, from group to group. It's not uncommon that it does that. It's something that we're always keeping in mind as we do these investigations. In terms of quantifying it, I don't know that we can put a specific number percent on how often we see that but it certainly is something that's always in the back of our minds.
Dr. Andrew Fox: And if there's another database that I would add to this is e-Trace data. Looking at purchasers of those guns, you might be able to track back through some of these networks of how they moved. It would be a very valuable piece of information within these networks that could enlighten how these guns are moving throughout these networks.
Audience Question: Are the seven jurisdictions that you're working with to gather data the most active areas in the county from a crime perspective?
Dan Carew: Yes, as we've mentioned, we have 40 different law enforcement agencies. When we set out to start to collect data, we try to focus on those jurisdictions with the highest amount of firearm violence. We're in the works of doing under our grant, looking at firearm violence from a public health perspective in addition to criminal justice perspective. Some of the works of our Public Health Department here in the county had already done — they already analyzed ten years’ worth of firearm homicide data and found out that 94% of our firearms homicide came from these seven jurisdictions. That's why we started working with these jurisdictions. We've since added a couple other agencies interested in working with us but that's definitely our starting point — trying to figure out where we could have the most bang for our buck in terms of looking at firearm violence.
Audience Question: How much do you rely on debriefs with offenders to provide historical, familial, or hierarchical information to bolster the other data sets?
Dan Carew: It's not something that we typically do here. We looked at a lot of different models around the country on how prosecutors’ offices have set up their systems, and that's a component of what a lot of these units have done. We don't really do that. I would direct whoever who's asking that question to look at the Manhattan DA's Office — they're at the forefront of the crime strategies, data-driven prosecution movement. I know they've been active since around 2011 or something like that. But I know that they do a lot of offender debriefs and have that as a very structured part of the work that they do. It's not something that we typically do although it may be in the future. But I can't imagine that we're going to do a ton of it.
Dr. Andrew Fox: To add on to that, Milwaukee does a great job of it. They have some report where they do debriefs and collect some of that information. I'd be happy to put you in contact with them because they do a good job of it.
Click Here to Watch a Recording of Social Network Analysis: An Innovative Tool to Maximize NIBIN Leads.