Number of words – 782

After studying reams of data about each of the more than 500 shooting incidents we worked with civic leaders to come up with ways to identify at risk children before they got shot. For each shooting, we asked where did it happen? At what time? In what sort of area did it occur? We looked at the profile of each victim: age, gender, race and family situation. We looked at each victim’s attendance record and academic record and whether he or she had faced any disciplinary proceeding. We noted the type of school for each victim (that is whether it was a general school a charter school or a remedial one). We ended up looking at more than 50 factors from a wide variety of sources and then we looked at the same factors for the general population (more than 110,000 high school student). We began to create a new paradigm of violence prevention based on the idea that if you could actually predict which students were most likely to be shot, you could protect them in a much more proactive and effective ways.

Working with CPS, we found the range of factors which highly correlated to a student getting shot and then tried to focus on those factors that there within the control of CPS. We defined four key risk factors: attendance (those who got shot missed an average of 40% of school days compared with 15% for the general population), academic performance (victims were 5 times more likely to be not graduating), in school behaviour, (victims were 8 time more likely to have committed a violent act at school) and the type of school victim (they were more likely to attend an alternative school, where CPS administrators placed students when the traditional neighbourhood schools couldn’t handle them).

With this focus profile one can imagine solutions that would address the issues at the alternative schools and improve attendance, academic achievement and classroom behaviour. But instead as the leaders at CPS began to shift their outlook to a very proactive ban, aimed at early prevention and we all agreed to use the model to project the probability of each student begin shot. Accordingly we developed a model that used the data to assign a specific probability of being shot to each student and thereby identify the small set of students who were most at risk.

The entire population was grouped into four primary categories: first about 200 students who were deemed ultra high risk with a risk of getting shot of more than 20%. Then emerged about 1000 students who were deemed Very high risk, with the risk between 7.5 and 20%. Then about 8,500 deemed high risk with a risk between 1 and 7.5%. The remaining 100,000 students had a risk of getting shot of under 1%. The model suggested that 10% of the shot population would fall into the Ultra high risk category with 20% in the very high risk. All of that meant that we expected 30% of the shootingd to come from a group of 1200 students, about 1% of the CPS population.

Once CPS had this new box of “Prevention,” it could then fill it with strategies to protect the student rather than saying to protect everyone equally. Awareness enhancement programmes were implemented directly in the school curricula. Another experiment included one on one mentoring. Our analysis showed that most of these young people didn’t have any meaningful relationship with adults. They got into trouble even when they were able to show up at school. CPS developed more aggressive interventions, beefed up security in particularly risky neighbourhoods and instituted a safe passage program to help students make it across the gang boundaries. Gun violence studies have shown that students worried about personal security do not perform well academically, continuing the downward spiral. In some instance, CPS helped relocate students to schools, where they would be safer.

In the 2009-2010 school year, there was a 16% drop in shooting. A new ethic of Corporation began to take hold in Chicago 120 High Schools still feeling in the box of prevention. CPS used the data on students to create a safety metric for each school, based on its specific location. This normalising across all neighbourhoods makes it possible to compare best practices across schools and also increases accountability. CPS discovered that schools with poor safety metric ratings generally rely on security guards and surveillance cameras as opposed to social workers and personal mentoring. Top performing schools are the one that help their students cope with the social and emotional problems underlying the violence swelling around them.

Excerpted from ‘Thinking in new boxes’ by Alan Iny and Luc de Brabandere