MAR 11: Dr. Bhramar Mukherjee, Yale School of Public Health

“Unveiling Bias: A Statistician’s Quest for Data Representation in Health Research”

Tuesday, March 11, 2025
Hybrid Event
In-person: Boston University, Duan Family Center for Computing & Data Sciences
665 Commonwealth Ave, Room 1750 (17th floor)
Boston, MA 02215
12:00-1:00pm Seminar
1:00-2:00pm Luncheon

Bio: Professor Bhramar Mukherjee is the Anna M.R. Lauder Professor of Biostatistics and Professor of Chronic Disease Epidemiology at the Yale School of Public Health (YSPH). Professor Mukherjee serves as the inaugural Senior Associate Dean of Public Health Data Science and Data Equity at YSPH. With over 400 publications in statistics, biostatistics, medicine, and public health, Professor Mukherjee is globally recognized for her research contributions in integrating genetic, environmental and health outcome data. Dr. Mukherjee is a fellow of the American Statistical Association and the American Association for the Advancement of Science. Dr. Mukherjee has received numerous awards including the Gertrude Cox award, the Adrienne Cupples Award, the Janet Norwood award, the Sarah Goddard Power award, the Karl E Peace Award, the Jerry Sacks Award, and the Marvin Zelen Statistical Leadership Award. In 2022 she was elected to the US National Academy of Medicine. She is the founding Director of an NIH funded undergraduate summer program on big data and has been directing this program since 2015.

Abstract: Despite several proposed roadmaps to increase representativeness in scientific studies, most of the world’s research data are collected on selected populations. We rely on summary statistics from well-represented groups and then devise clever statistical methods to transfer/transport the inference for another population. In this talk, I would first argue the obvious: for building fair algorithms and models we need representative training datasets. However, till we have reached the dream of representative big data at a global scale, statisticians have an important role to play. In fact, we have the perfect tools to study the “unobserved” through consideration of selection bias, information bias and alike. I will share examples from my personal journey as a statistician when doing good and timely statistical work with imperfect data (such as data from observational patient care databases) quantified important scientific observations. I will conclude the talk with a call to arms for statisticians and computational scientists to not just develop new methods but also lead efforts for creating, curating, collecting data and pioneering new scientific studies, stepping outside their comfort zones. As public health researchers, our job is not just to predict diseases accurately, but to prevent diseases effectively for communities near and far. I hope to weave a narrative around bias that has multiple definitions and layers.