Principles & Performance Metrics Makerspace

Principles and Performance Metrics Makerspace

Responsible data science frameworks suggest general principles such as fairness, accuracy, accountability, etc. but provide scant structure for how to communicate and compare different implementations of these principles. This event convenes interdisciplinary data-centric and data-adjacent researchers into application areas for a two-day exploration.

Day 1: Participants can choose between 3 breakout groups:
1) Public Sector AI / Procurement: Imagine you’re involved in the development or procurement of an AI system designed to detect housing conditions in a city. What principles are most important to you in this AI procurement process? What metrics or evidence would you expect—or be expected—to provide to demonstrate those principles are being upheld?
2) Health and Social Determinants: As a social impact researcher, what are the challenges you face in choosing metrics to measure the well-being of the population you are benefitting? What are ways you measure social and health determinants and what criterion do you use to choose between metrics? How do you think this connects / does not connect with how the population perceives their well-being?
3) Responsible Research: Academia often talks about responsible research in broader principles such as integrity, accountability, or engagement. From your perspective, which of these principles are important, and how are they measured in your field?

Day 2: We will synthesize the principles and metrics collected from each group and map out common understanding and potential misunderstanding, and how this will affect the policy process. We will find areas and subgroups for future collaborations.