Remember: bias in, bias out.
When data collection fails to include certain communities or locations, any analysis or predictions based on that data will also be infected by bias.
@vm-wylbur.bsky.social
Director of Research @hrdag.org
Remember: bias in, bias out.
When data collection fails to include certain communities or locations, any analysis or predictions based on that data will also be infected by bias.
Session card for an IRE AccessFest panel on "Utilizing new police misconduct and employment databases"
Learn about using police employment and misconduct data housed in the National Police Index and Police Records Access Project, two new databases for use by journalists, attorneys, researchers, and members of the public.
We'll be presenting at @ire.org's AccessFest this weekend.
We used LLMs to analyze and extract information from millions of pages about police abuses in California. Then we used human annotation to check and validate our analysis.
hrdag.substack.com/p/pulling-ba...
Iβm proud of my organization @hrdag.org and my colleagues' courage to speak truth to power. Human rights and evidence must never bend to tyranny. Β hrdag.org/2025/10/21/h...
21.10.2025 19:03 β π 5 π 2 π¬ 0 π 0A cartoon image of 5 people standing shoulder to shoulder, with two holding laptops. There is a green stripe across the front and words in white that say "We Stand Against Tyranny"
Despite these risks, our staff will continue to support civil society in our collective pursuit of justice.
Read our full statement: hrdag.org/in-the-face-...
Today, those rights are under attack. The United States is no longer a place where scientists and civil society can publish and engage in our work without fear of reprisal.
This has a chilling effect on those using data to advance justice and accountability.
We are data scientists using rigorous scientific methods and statistical analysis to quantify war crimes and human rights abuses. We have long relied on the protections of the U.S. law to protect our right to publish our research, even when it is unpopular.
21.10.2025 17:49 β π 2 π 1 π¬ 2 π 0A cartoon image of 5 people standing shoulder to shoulder, with two holding laptops. There is a green stripe across the front and words in white that say "HRDAG Stands Against Tyranny"
HRDAG remains a nonpartisan, nonpolitical organization. Our mission is focused on the Universal Declaration of Human Rights.
Today, we denounce violations of human rights occurring in the United States. hrdag.org/in-the-face-...
this was a lot of fun! thanks βͺ@plutopia.ioβ¬
11.06.2025 01:14 β π 1 π 0 π¬ 0 π 0this was a lot of fun, thanks @plutopia
11.06.2025 01:13 β π 0 π 0 π¬ 0 π 0"Iβm excited to work with a group of researchers that conduct their work with rigor, transparency, and an appreciation for the gravity of the subject matter."
- Will Taylor, doctoral student with @umich.edu and HRDAG's newest Data Science Fellow
hrdag.org/2025/05/28/d...
βTodayβs legal victory sends a crystal-clear message: Americansβ private data stored with the government isnβt the personal playground of unelected billionaires,β Everett Kelley, president of EFF client AFGE, told @bloomberglaw.com.
10.06.2025 20:04 β π 167 π 38 π¬ 3 π 0The vast majority of wrongful convictions in one Louisiana parish are connected to the police failing to share key evidence.
We partnered with Innocence Project New Orleans to track and analyze this type of police misconduct data hrdag.org/2025/02/20/i...
Will Taylor, a doctoral student from @umich.eduβ¬, is joining us for the summer. He'll be using scientific analysis to better understand and interpret human rights violations such as wrongful convictions.
Welcome, Will! hrdag.org/2025/05/28/d...
"Science gives us tools to go beyond sampled data and talk more accurately about the world." - Patrick Ball,
@hrdag.org's director of research.
www.biotechniques.com/computationa...
Research Director Patrick Ball giving a speech and pointing to a chart on a powerpoint slide. The title of the chart is "La Conclusion principal"
Executive Director Megan Price giving a speech in front of a large auditorium. She is standing in front of a screen with a powerpoint slide behind her, but the words are obscured. The heads of the audience are visible and the camera is behind them so their faces are unseen.
4 staff members sitting at a long dining room table in a room with large windows. Each has a laptop computer in front of them and they are working together, and there is sunlight streaming in through the windows behind them.
We are data scientists, statisticians, and computer scientists.
We use rigorous scientific methods to analyze and investigate human rights abuses and support partner orgs seeking justice.
We believe human rights work benefits from scientific accuracy, and we're grateful to do our part.
"I have learned so much from HRDAG about data collection and documenting harm. It is an honor to support their creative and rigorous team by serving on their board.β - Cynthia Conti-Cook
We are so glad to welcome Cynthia.
hrdag.org/2025/05/28/a...
One of the great antidotes to fake news and politicized propaganda is statistical analysis of existing datasets using rigorous scientific methods.
Learn about our approach to bringing science to understanding history.
www.biotechniques.com/computationa...
A few of our recent projects:
β
Consolidating data to track police misconduct
β
Co-creating a new tool to track "wandering police officers" who try to avoid accountability by changing jobs
β
Categorizing evidence of gender-based violence in police interactions
Read on: hrdag.org/report/2024-...
More than 75% of wrongful convictions in one parish in Louisiana are connected to the police failing to share key evidence. That's more than twice the national average.
We partnered with Innocence Project New Orleans to dig deeper into the data. hrdag.org/2025/02/20/i...
A map of the entire world with lots of blue pins scattered over it, including 3 pins in Africa, several in Europe and Asia, and a bunch in North and South America.
Our data scientists, statisticians, and software developers have traveled to 5 continents.
We travel to post-conflict countries, build databases of human rights violations, and conduct rigorous research and analysis to identify the truth.
hrdag.org
One third of all Americans killed by strangers are killed by the police.
We analyzed overlapping databases to show that, in the U.S., police pose a grave safety threat. Learn about our statistical analysis and read our report, "Violence in Blue":
hrdag.org/2020/06/11/v...
We are launching a substack newsletter. Please subscribe and tell your friends.
hrdag.substack.com/about
We believe that mathematics and science are powerful tools for standing up for human rights.
We use statistics and data analysis to hold perpetrators accountable. Learn more at hrdag.org
please change murder to homicide -- PB.
15.05.2025 20:24 β π 0 π 0 π¬ 0 π 0from the US, always curious to hear local experts. Please keep the analysis coming!
19.12.2024 21:59 β π 0 π 0 π¬ 0 π 0This new article in _Chance_ explains how @hrdag.org and our partners use AI to sift massive troves of law enforcement documents to make sense of police reports about use of force. amstat.tandfonline.com/doi/full/10....
19.12.2024 01:36 β π 13 π 3 π¬ 0 π 0Biotechniques published an interview with Patrick Ball, inspired by his John Maddox Prize award.
"Meet the data analyst putting the perpetrators of genocide in prison"
www.biotechniques.com/computationa...
HRDAG's newsletter: Witnessing Syria on Human Rights Day
mailchi.mp/hrdag/2024-yea-5