Peer-Reviewed Publications
Bureaucratic Challenges of Unified Data: Examining Predictive Policing Technology Design Issues Posed by Federal and State U.S. Criminal Legal System Data Sets
A rise in law enforcement and judicial use of predictive policing technologies in the United States has led to questions about the accuracy and validity of the crime data used to train algorithms. In this paper, we introduce preliminary findings from filing public data requests with Departments of Corrections in 50 U.S. states to highlight how using federal data to train predictive policing technologies could pose significant accuracy issues due to cross-state differences in criminal legal system data availability, usability, and comparability. We argue that technologists need to consider how federal data sets combine data that individual states collect. We introduce several avenues for researchers to explore further, including examining how inconsistencies in federal data may affect AI algorithm performance in minority communities.
The Effect of Race and Sex on Contact Sex Crime Arrest Odds
This article investigates how the intersection of race and sex affects contact sex crime arrest odds. Logistic regression analyses of 2012–2016 U.S. National Incident-Based Reporting System data for 391,352 people reported to law enforcement for contact sex crimes reveals that Black people reported for contact sex crimes have higher odds of arrest than those reported who are not Black, and the relationship between race and arrest odds varies by sex—the effect is stronger for those labeled “female” or “other” and weaker for those labeled “male.” I posit that sex strengthens the effect of race on contact sex crime arrest odds for those who are not male because race helps construct non-male bodies as masculine, and in a U.S. cultural context that has viewed sex as a legitimate way of asserting male dominance, sexual violence is masculine. However, race is not necessary to construct most male bodies as masculine, making the strength of race's effect on contact sex crime arrest odds less significant for males.
“Terrifying and Exhausting”: Secondary Victimization in Title IX Proceedings at U.S. Higher Education Institutions
Students at United States colleges and universities increasingly turn to campus Title IX processes to address sexual violence, but little research exists on the emotional fallout of Title IX processes and student perceptions of process bias. This paper presents an analysis of 72 responses to a 2021 survey of Title IX process complainants and respondents at higher education institutions. While students on both sides of Title IX processes experienced negative emotions, complainants often minimized these experiences while respondents emphasized them. This form of comparative secondary victimization risks further marginalizing complainant voices at higher education institutions.
Protections for Marginalised Women in University Sexual Violence Policies,
Higher education institutions in four of the top 20 wealthiest nations globally (measured by GDP per capita) undermine gender equality by failing to address sexual violence perpetrated against women with marginalised identities. By analysing student sexual violence policies from 80 higher education institutions in Australia, Canada, the United Kingdom, and the United States, I argue that these policies fail to account for the ways that race, sexuality, class and disability shape women’s experiences of sexual violence. Further, these deficiencies counteract efforts to achieve gender equality by tacitly denying women who experience violence access to education and health care. The conclusion proposes policy alterations designed to address the complex needs of women with marginalised identities who experience violence, including implementing cultural competency training and increasing institution-sponsored health care services for sexual violence survivors.