Video Transcript: LEADS Scholar Spotlight — Predictive Policing Algorithms
Shon Barnes, a deputy police chief with the Salisbury Police Department in North Carolina and a Class of 2015 scholar of NIJ’s Law Enforcement Advancing Data and Science (LEADS) Program, explains predictive policing and details a quasi-experiment that his department performed. He credits the LEADS Scholarship Program with helping him understand data and ask the right questions.
SHON BARNES: Predictive policing algorithms is just simply taking data and putting into a system that will tell us when and where crime is going to be. The purpose behind algorithms are to better assist police managers and deployment of officers. So we want to be where we need to be in order to prevent crime. It’s a little different from “hot spot” policing, whereas the data tells you where crime is happening now. Predictive policing algorithms give police mangers an opportunity to determine where crime will be.
So what we did was we did a predictive policing quasi-experiment based on the way our police department was structured so that we had one group of officers applying the treatment, which was the predictive policing algorithms. We had officers go to where crime was supposed to be. And then we had our other officers doing the normal routine — “hot spot” policing, answering calls for service,
We also looked at how many people were on the street. Turns out those that did the predictive policing algorithms went out with less officers, were more busy, yet they had a greater reduction on crime than those offers who didn’t. So NIJ was very instrumental in helping me understand my data, helping me ask the right questions. The LEADS Scholarship Program is great for me because it’s not about teaching you what the answers are but teaching you how to ask and write the right questions in order to solve a problem.
May 9, 2018