Computational Policy Lab

We're a team of researchers and data scientists who use computational approaches to address issues in criminal justice, education, voting rights, and beyond.

[Computational Policy Lab @CompPolicyLab]

(2/2) With TV still the dominant news source for Americans, these shifts raise key questions about the role of broadcast news in shaping voter knowledge and political engagement.

📖 Read more here: https://t.co/irHXI4ounA

[Computational Policy Lab @CompPolicyLab]

(1/2) Our own @jgaeb1 & Sharad Goel, w/coauthors @seanjwestwood & Shanto Iyengar, analyzed 50+ years of US TV news using a novel LLM-based classification system. They find a steep decline in in-depth political coverage & substantive reporting, while soft news & commercials rise.

[Harvard Kennedy School @Kennedy_School]

HKS researchers—including @5harad, @Soroush_Saghaf, and @madisoncoots—created an analytical framework that informs the ongoing medical debate over whether to use race-unaware or race-aware risk assessments https://t.co/oX49Chb8lZ

[Alex Chohlas-Wood @LX_CW]

NEW in Management Science!

My coauthors and I came up with a new consequentialist approach to designing equitable algorithms.

Instead of imposing fairness criteria on an algorithm (like equal false negative rates), we aim for good outcomes.

More in the 🧵 below! (1/)

[Computational Policy Lab @CompPolicyLab]

Less than one week to join our study on using AI to improve education! Give your students access to a free virtual tutor & receive a $1,000 honorarium.

Apps open for college instructors (esp. those teaching math, comp science, stats, & econ) until Dec 16: https://t.co/U9EnKYEQC2

[Computational Policy Lab @CompPolicyLab]

Calling instructors at 2- & 4-year colleges! Give your students free access to a virtual tutor optimized for learning by taking part in our study on using AI to improve education. Participating instructors receive a $1,000 honorarium.

Sign up by Dec. 16: https://t.co/U9EnKYForA

[Madison Coots @madisoncoots]

The use of race in clinical risk models is heavily debated. While race-aware models can be more accurate, some are concerned about reinforcing racialized views of medicine. In our paper in @AnnalsofIM, we offer a new perspective on this debate. 🧵👇 https://t.co/1e8tJ75wVg

[Computational Policy Lab @CompPolicyLab]

Our virtual tutor, called PingPong, is based on ChatGPT.
It's designed to help students learn, not to simply give them answers.

We'll help customize PingPong to your course content.

Instructors can view de-identified student interactions.

PingPong is multilingual.

[Computational Policy Lab @CompPolicyLab]

Calling instructors at 2- & 4-year colleges! Give your students free access to a virtual tutor optimized for learning by taking part in our study on using AI to improve education. Participating instructors receive a $1,000 honorarium.

Sign up by Dec. 16: https://t.co/U9EnKYForA

[Computational Policy Lab @CompPolicyLab]

Congratulations to our own @jgaeb1 and @5harad for winning Best Paper at #EAAMO2024!

[EAAMO @EAAMO_ORG]

🏆 Congratulations to @jgaeb1 and @5harad (@CompPolicyLab) for winning the Best Paper at #EAAMO2024! Their work introduces the Robust Outcome Test to improve fairness analysis in discrimination studies by combining benchmark and outcome tests under a new statistical guarantee.

[Computational Policy Lab @CompPolicyLab]

Finally, we’re excited to note that the US Place Sampler was recognized earlier this year with an “Award of Excellence” from the Society for News Design! Congratulations to @JoeNudell and team. https://t.co/cLaJ703eM4

[Computational Policy Lab @CompPolicyLab]

🧵Tool spotlight: Prompted by research by one of our engineers, @JoeNudell, the United States Place Sampler tool by @themarkup and @BigLocalNews provides insights into address-level internet access disparities. Read more here: https://t.co/IhvX8kgmMI

[Computational Policy Lab @CompPolicyLab]

The US Place Sampler is now available to all, making it easy for anyone to sample random US street addresses to dig into disparities. See this video for an overview of how it works: https://t.co/uILEIeYZe6

[Computational Policy Lab @CompPolicyLab]

Our co-director @LX_CW spoke with @Arnold_Ventures about the role AI can play in supporting fair charging decisions. Read more about our race-blind charging project, which aims to mitigate unintended biases in the justice system: https://t.co/18qiticNwF

[Alex Chohlas-Wood @LX_CW]

It was great chatting with @Arnold_Ventures about why this research effort is so important to me!

[Alex Chohlas-Wood @LX_CW]

Are you a prospective PhD student interested in combining statistical, computational, and social science approaches to address pressing societal problems? My dept @NYU_ASH is starting a PhD program in computational social science. Learn more at our final info session this Thu!

[Computational Policy Lab @CompPolicyLab]

Data collected since 2018 under the Racial and Identity Profiling Act help measure disparities in California police traffic stops.@LX_CW contributed data to highlight these disparities for a story in the @sfchronicle: https://t.co/wzQiwPYevW

[Dan Kopf @dkopf]

Yesterday we posted an unusual story for us at the @sfchronicle that was actually about 400 stories. Here is a little thread about how it happened.
https://t.co/yPccQA5MBA

[Dan Levy @danmlevy]

Want to use ChatGPT in your teaching but don't have much time⏱️to learn how? If so, this webinar recording from @HarvardBizEdu might be helpful.

[Alex Chohlas-Wood @LX_CW]

Some personal news: I’m back at @nyuniversity in a new role as an assistant professor of computational social science at @NYU_ASH! I’ll also continue to lead the @comppolicylab as a faculty co-director. New York friends, get in touch and we can grab a coffee!

[Computational Policy Lab @CompPolicyLab]

Looking for a primer on generative AI? This open course, offered by our lab faculty director @5harad and affiliate @tedsvo, provides an introduction to how GenAI works, how to use it, and the larger opportunities and challenges it poses for society. https://t.co/BW7eAYXUcp