Better than Nothing

Timeline: 2023 - 2024

Role: Research Assistant

Supervised by: Tawanna Dillahunt, Alex Lu, Shruti Shannon, Social Innovations Group

TL;DR: Lead data analysis of community driven survey to better quantify Detroit residents’ perceptions of Project Green Light and community safety. Follow up work to ECN Community Photovoice project.

Skills: Pandas, Statsmodels, Data Visualizations (Seaborn)

Publications

[c.3] Lu, A.J., Moy, C., Ackerman, M.S., Dillahunt, T.R. (2024). Perceptions of Policing Surveillance Technologies in Detroit: Moving Beyond "”Better than Nothing”. In Proceedings of ACM Conference on Fairness, Accountability, and Transparency. ACM, New York, NY, USA.

About

Methods: Quantitative Analysis ~ Python, Pandas, Statsmodels, Data Visualizations (Seaborn)

Following the ECN Community Photovoice project, our team developed questions for the Detroit Metro Area Communities Study (DMACS) to understand Detroit residents' perceptions of safety, surveillance, community, and Project Green Light through quantitative approaches.

Project Green Light (PGL) is a public private partnership with the city of Detroit. PGL allows local businesses to purchase and install Green Light cameras that are monitored by the Detroit Police Department with the intention of community policing and preventing crime. This program has recently come under criticism as the Green Light cameras utilize facial recognition and have been prone to misidentifying Detroit residents, who are predominantly Black, in criminal investigations.

Since receiving the DMACS data, I have led the quantitative analysis and statistical modeling. Our current research questions look to examine the role of gender, race, and age on feelings of safety and sentiments of PGL. When combined with our qualitative work during the ECN Community Photovoice project, we broadly understand Detroiters are dissatisfied with PGL, but these surveillance technologies are "better than nothing." This work recently appeared at FAccT 2024.