BookmarkSubscribeRSS Feed

Shootings by American Cops: Location Matters

Started ‎07-20-2020 by
Modified ‎07-21-2020 by
Views 1,759

American police officers' use of lethal force is in the headlines after the deaths of George Floyd, Breonna Taylor and Ahmaud Arbery. To pinpoint key issues around deadly force -- and inspire data-driven solutions -- three Oklahoma State University graduate students examined the aftermath of police shootings as their SAS Global Forum Student Symposium project. Are officers indicted? Do underlying biases and prejudices result in letting them off easy? Does the location of shootings happen make indictment more or less likely?


"The goal of this project is to create a predictive model that can identify the variables that lead to a perceived justified or unjustified police shooting and forecast the shootings across the United States for the next 10 months in an effort to provide key insights to those campaigning for reform," wrote Alex Lindsay, AraVind Dhanabal and Mason Kopasz in their paper, What Happens After Police Shootings? The students, pursuing master's degrees in business analytics at Oklahoma State, analyzed data from and the U.S. Census Bureau.


Spoiler alert: "The predictive model did not result in finding a racial bias for perceived justified versus unjustified police shootings," the paper concluded. "However, the State, County, Department Involved proved to be the most important variable in deciding if a police shooting will be perceived justified or unjustified. This model could be used in practical application by an oversight committee to ensure there is standardization across police forces for launching investigations into Grand Jury indictments."




Video highlights

00:48 - Background and context for the study

01:54 - Goal of the study

02:33 - Data sources

03:53 - Target variable: whether officer is indicted

04:12 - More white males shot, but disproportionately more African-Americans and Hispanics represented in shootings

07:05 - How predictive analytical models were built

10:21 - High-risk states: Georgia, Nebraska, Arizona, Maryland and Louisiana

11:01 - Conclusion and possible future studies


Read the Paper


Related resources

Decision Tree Model (documentation)

Decision Tree in Layman's Terms (community article)


Is there a link to their sas em diagram or any of that sort ? 

@b_smsha Some information is in the appendix of the paper. The author contact information (students) is also in the paper. As the authors were students in 2020, be aware they might not be currently reachable at the university. 

Version history
Last update:
‎07-21-2020 08:29 AM
Updated by:



Registration is open! SAS is returning to Vegas for an AI and analytics experience like no other! Whether you're an executive, manager, end user or SAS partner, SAS Innovate is designed for everyone on your team. Register for just $495 by 12/31/2023.

If you are interested in speaking, there is still time to submit a session idea. More details are posted on the website. 

Register now!

Free course: Data Literacy Essentials

Data Literacy is for all, even absolute beginners. Jump on board with this free e-learning  and boost your career prospects.

Get Started

Article Tags