On May 10 and 17, 2024, leading researchers and statisticians gathered for a two-day virtual event as part of the Ingram Olkin Forum: Statistics Serving Society (S3) Series. The event, organized by the National Institute of Statistical Sciences (NISS), brought together experts to explore the latest methodologies and data sources related to gun violence in the United States.
Day 1 Highlights:
The first day, held on May 10, opened with remarks setting the stage for the afternoon’s discussions. The initial session, titled "Emerging and Evolving Data Sources for Studying Gun Violence," featured Susan Parker from Northeastern University, who provided an update on non-fatal firearm injury surveillance in the U.S. Charles Loeffler from the University of Pennsylvania followed, discussing how recovered ammunition can offer insights into the constraints in the gun violence supply chain. Iris Horng, also from the University of Pennsylvania, concluded the session with a presentation on probabilistic record linkage of two gun violence data sets. The session was chaired by Wayne Osgood from Penn State.
Following a brief break, the second session, "Patterns in Police Use of Force," began with Lucas Mentch from the University of Pittsburgh addressing racial disparities in fatal police shootings. Marie Oullet from Georgia State University discussed the role of networks in understanding gun violence involving law enforcement, while Justin Nix from the University of Nebraska critiqued the quality of open-source data on police shootings. Claire Kelling from Carleton College chaired this session.
The final session of the day, "Causal Modeling of Gun Violence Policies," featured collaborative presentations. Avi Feller (UC Berkeley) and Eli Ben-Michel (Carnegie Mellon) shared statistical methods for estimating the impact of gun policy on violence. Bijan Niknam and Liz Stuart from Johns Hopkins University discussed the challenges of making design choices involving small, undefined geographic areas. Dan Lawrence, alongside Eric Piza from CNA, SIUE, and UMD, assessed the impact of gunshot detection technology. Rosanna Smart from RAND led the discussion for this session.
Day 2 Highlights:
The second day, held on May 17, focused on more specialized topics. The first session, "Point Process Modeling of U.S. Gun Violence," featured George Mohler from Boston College, who analyzed changes in the reproduction number of mass shootings following the COVID-19 pandemic. Andrew Holbrook from UCLA discussed the computation of Hawkes processes for gun violence research, and Yao Xie from Georgia Tech presented on non-stationary spatial-temporal self-exciting point processes. Lingzhou Xue from Penn State chaired the session.
Participants then broke into Zoom breakout rooms for in-depth discussions on the themes from the event, including emerging data sources, police use of force, causal modeling, and point process modeling. Each group later reconvened to share the highlights of their discussions.
The event concluded with a session on miscellaneous gun violence papers. David Corliss presented a statistical analysis of school shootings in the context of stochastic terrorism. Jonathan Jay from Boston University demonstrated how machine learning and aerial imagery can enhance research on the built environment and community firearm violence. Max Kapustin discussed efforts to reduce gun violence in Baltimore’s Western District without exacerbating the harms of policing. Claire Kelling returned to chair this final session.
About the Ingram Olkin Forum Series:
The Ingram Olkin Forum (IOF) Series, presented by NISS, is dedicated to addressing pressing societal issues with innovative statistical approaches. The S3 Forums bring together experts to collaborate on emerging challenges, fostering new ideas and methodologies to advance public policy and research.
For more information, visit the IOF page.