Learning Collider

@learningcollider.bsky.social

1714 Followers 711 Following

Building AI that Expands Human Opportunity Research lab founded and led by UT-Austin economist Peter Bergman. peterbergman.bsky.social LearningCollider.org

Building Fair and Equitable AI Systems; Understanding and Reducing Discrimination in AI and Human Decision-Making; Using AI to Address Learner Variability; Creating Pathways to Economic Mobility

Building AI that expands human opportunity

Favorite Findings: "Voucher holders in Seattle and King County, much like the rest of the nation, tend to live in low-opportunity neighborhoods... Do families with vouchers want to live in high-opportunity neighborhoods, but face barriers that limit their access to such areas? If so, how can we reduce these barriers?

In collaboration with the research team, the Seattle and King County Housing Authorities developed a set of services designed to support moves to high-opportunity neighborhoods, building on lessons from prior mobility and housing search assistance programs.
The program included 3 components: search assistance for families, landlord engagement, and short-term financial assistance.
Creating Moves to Opportunity increased the number of families who moved to high-opportunity neighborhoods by 38 percentage points and changed where families chose to move but did not affect overall voucher utilization rates..."

Panel description: AI in Teaching & Learning - How is AI changing the education landscape? What are Texas K-12 and higher education institutions doing to develop responsible and forward-thinking AI use on campuses? What are some use cases for classroom use? How can organizational frameworks help diffuse AI-based innovations and reduce duplication of efforts? This panel will delve into ways to harness responsible AI use to drive teaching and learning outcomes.

Favorite Findings graphic reads: Using granular data from a nationwide randomized controlled trial at Uber, we estimate the effects of switching from a fixed weekly pay schedule to Instant Pay, a system that allows on-demand, within-day withdrawals. We find that flexible pay substantially increased both drivers’ work time and earnings (ITT: 2%; TOT: 18-37%). Authors: M. Keith Chen, Katherine Feinerman, Kareem Haggag

Screenshot of tweet from Aline Lerner, CEO of interviewing.io: At @interviewingio, we ran a study with @learncollider. We asked recruiters to judge resumes and compared how engineers did in interviews. Turns out recruiters are barely better than a coin flip.

Favorite Findings: "Results show consistently large and robust effect sizes of phone call tutorials on learning across contexts, with average effects across all 5 countries between 0.30-0.35 standard deviations... We find the program yields 3.4 LAYS (Learning Adjusted Years of Schooling) per $100, ranking among the top 10 out of 150 education interventions reviewed." Research by: Noam Angrist, Micheal Ainomugisha, Sai Pramod Bathena, Peter Bergman, Colin Crossley, Claire Cullen, Thato Letsomo, Moitshepi Matsheng, Rene Marlon Panti, Shwetlena Sabarwal, and Tim Sullivan

Favorite Finding Friday quote: “...nearly 70% of the judges make override decisions that are no better than random—that is, they could achieve a lower pretrial misconduct rate by flipping a coin...we also find that 10% of the judges outperform the algorithm when they make a discretionary override, suggesting that a human and algorithm working together can potentially outperform automated release decisions.” Research by Victoria Angelova, Will S. Dobbie, Crystal Yang

Favorite Findings Friday quote graphic: “...we find that eviction exacerbates the economic distress experienced by tenants in the lead-up to a court filing. In the two years after a case, eviction increases homelessness, residential mobility, and hospital visits. During this period of disruption, eviction also reduces earnings, with particularly large effects for female and Black tenants. In the longer run, eviction worsens financial health through reduced credit scores and increased indebtedness.” Attributed to Robert Collinson, John Eric Humphries, Nick Mader, Davin Reed, Daniel Tannenbaum & Winnie van Dijk

Graphics representing housing-related sectors and specialties: engineering, construction, architecture, permitting, zoning, supply chain management, listings, and development

Favorite Finding: Compared to the most widely used placement tests, the algorithms in the study placed 49% of students into higher-level English and 15% of students into higher-level Math, increasing overall credits earned without reducing pass rates. Research by Peter Bergman, Elizabeth Kopko, Julio Rodriguez

Quote from working paper by Jonathan Colmer, John Voorheis, and Brennan Williams: "...disparities in air pollution can account for up to 22% of the Black-White earnings gap, 18% of the Hispanic-White earnings gap, & between 13-20% of the average neighborhood-earnings effect."

Colorful social network map on dark background

Photograph of landscape and blue, cloudy sky. Colorado Front Range.