More Data, More Problems: What Society Needs from the Statistical and Data Sciences

Seminar | March 13 | 4-5 p.m. | 1011 Evans Hall

 Brent Durbin, Smith College

 Department of Statistics

Despite more than 20 years of increasing reliance on data-intensive digital tools for commerce, governance, and social interaction, society has been slow to respond to both the promise and the perils of the phenomenon we now call Big Data. This lack of adaptation to new ways of collecting, storing, and analyzing data has been especially apparent within universities and the public sector, both of which have a responsibility to understand and regulate how innovative technologies make their way into the world.

This talk outlines some of the biggest challenges for socially adapted data analysis, including biased data sources, proprietary data and analytical tools, antiquated regulatory frameworks in both academia and government, and an “engineering mindset” among developers and analysts that too often asks how to solve a problem but not why or whether to do so. I consider how university-based statistical and data scientific communities can promote an approach to data-driven research and teaching that is both rigorous and responsible.