Fairness, Rankings, and Behavioral Biases

Explore a keynote address from FAT* 2019 that delves into the complex interplay between fairness, rankings, and behavioral biases in decision-making processes. Examine how human biases can impact ranking systems and learn about formal models that analyze these effects. Discover potential interventions to mitigate bias and improve overall performance in screening decisions. Gain insights from speaker Jon Kleinburg of Cornell University as he presents joint research with Sendhil Mullainathan and Manish Raghavan. Follow the discussion on topics such as mental models, pipelines, work styles, and mathematical modeling approaches. Understand the concept of the Rooney Rule and its effects on diversity initiatives. Analyze unconstrained optimization, abstraction barriers, and the bias surface in decision-making contexts. Engage with the subsequent discussion led by Jennifer Wortman Vaughan from Microsoft Research to further explore the implications of this research on fairness and equity in algorithmic systems.

Introduction
Mental Model
Pipeline
Work style
Building mathematical models
Interpretive lens
Evaluating outcomes without thinking
Obstacles to diversity
The Rooney Rule
Effect of the Rooney Rule
Unconstrained Optimization
Abstraction Barriers
Building a Model
Results
Bias Surface
Whats Behind the Cliff
When to Reserve a Slot
Intuition
Discussion
Discussion Structure
Formalization