The primer addresses what pretrial risk assessment is and what the research supports.
Tools like Compas allegedly help judges predict future criminal activities and eliminate bias. HRDAG and partners showed how the tools recycle bias.
Algorithmic tools like PredPol were supposed to reduce bias. But HRDAG has found that racial bias is baked into the data used to train the tools.
In July 2009, The Human Rights Data Analysis Group (HRDAG) concluded a three-year project with the Liberian Truth and Reconciliation Commission to help clarify Liberia's violent history and hold perpetrators of human rights abuses accountable for their actions. (This work was conducted by HRDAG while with Benetech.)
In the course of this work, HRDAG analyzed more than 17,000 victim and witness statements collected by the Liberian Truth and Reconciliation Commission and compiled the data into a report entitled "Descriptive Statistics From Statements to the Liberian Truth and Reconciliation Commission." The report is included as an annex to the final ...
Identifiers being sequential could make possible estimations of the population of detained children.
If you’d like to stay informed about HRDAG events, blogposts, and news, connect with us on Twitter, Facebook or through our RSS feed. We also have a LinkedIn page.
You may contact us directly via email at info @ hrdag.org.
A note for persons in search of assistance with specific human rights cases: We are very sorry for your troubles and your suffering; however, HRDAG does not take on casework. If you need help with a human rights case, you might consider requesting it from the International Committee of the Red Cross (www.icrc.org).
Photo: U.S. National Archives
I spent the two weeks over Easter working with Patrick and Megan in San Francisco, trying to figure out a strategy of how best to estimate the number of casualties the Syrian civil war has claimed in the past two years. In January, HRDAG published a report on the number of fully identified casualties reported in the Syrian Arab Republic between March 2011 and November 2012. The number of de-duplicated records of killings for this period was 59,648, a number that is likely to be an undercount since we know that many incidences of lethal violence in conflict go unreported, and that the unreported cases are not missing at random. (more…)
The modular nature of the workflow and use of Git allowed us to work on different parts of the project from across the country.
“If you’re not careful, you risk automating the exact same biases these programs are supposed to eliminate,” says Kristian Lum, the lead statistician at the San Francisco-based, non-profit Human Rights Data Analysis Group (HRDAG). Last year, Lum and a co-author showed that PredPol, a program for police departments that predicts hotspots where future crime might occur, could potentially get stuck in a feedback loop of over-policing majority black and brown neighbourhoods. The program was “learning” from previous crime reports. For Samuel Sinyangwe, a justice activist and policy researcher, this kind of approach is “especially nefarious” because police can say: “We’re not being biased, we’re just doing what the math tells us.” And the public perception might be that the algorithms are impartial.
We aim to produce code that is clear, replicatable across machines and operating systems, and that leaves an easy-to-follow audit trail.
Herb led and mentored a generation of statisticians working in human rights.
(This post is co-authored by Patrick Ball and Kristian Lum.)
Today the Bureau of Justice Statistics (BJS) released a report on their effort to document “all deaths that occur during the process of arrest in the United States.” The analysis estimates that the Arrest-Related Deaths (ARD) program covers only 34-49% of these deaths. A parallel program by the FBI (the Supplementary Homicide Reports, SHR) is estimated to cover approximately the same proportion of deaths. Even taking into consideration both programs, 28% of all police homicides remain unreported.
In order to estimate the total number of homicides that appear on neither the ARD or ...
Field Consultant
Carolina Lopéz has worked with the Archivo Histórico de la Policía Nacional (AHPN) in Guatemala for eight years, and is currently a member of the Archive Technical Coordination team. A professional working within the social sciences, she prefers using alternative research on past practices to develop an understanding of the present. Her work consists primarily of monitoring and creating strategies to systematize, track and create process controls. She also has thorough knowledge of management of historical archive documents.
Since 2006, Carolina has worked in quantitative research at the AHPN with HRDAG team members Patrick ...
Wilkerson was speaking at the inaugural Conference on Fairness, Accountability, and Transparency, a gathering of academics and policymakers working to make the algorithms that govern growing swaths of our lives more just. The woman who’d invited him there was Kristian Lum, the 34-year-old lead statistician at the Human Rights Data Analysis Group, a San Francisco-based non-profit that has spent more than two decades applying advanced statistical models to expose human rights violations around the world. For the past three years, Lum has deployed those methods to tackle an issue closer to home: the growing use of machine learning tools in America’s criminal justice system.
Amelia Hoover Green. In Collective Violence and International Criminal Justice: An Interdisciplinary Approach, ed. Alette Smeulers, Antwerp, Belgium. © 2010 Intersentia. All rights reserved. [Link coming soon]
Margot is a professor in the Department of Energy Resources Engineering at Stanford University, interested in computer simulation and mathematical analysis of engineering processes.
