Skip to main content

Faculty Spotlight: Betsy Ogburn, PhD, MS

Betsy Ogburn, PhD, MS, is a professor in the Department of Biostatistics, where she develops methods for causal inference with complex data.

Published

Betsy Ogburn, PhD, MS, is a professor in the Department of Biostatistics, where she works on causal inference, including interference and social networks, measurement error, semiparametric estimation, instrumental variables methods, and mediation analysis. In 2023, she and her colleagues received a Nexus Award for their research on “Evaluating Accuracy and Reproducibility of Forensic Science Methods Used in Criminal Courts.”

Betsy is a senior fellow of the Good Science Project and affiliated faculty of the SNF Agora Center and the University of Pennsylvania Center for Causal Inference.

How long have you been at the Bloomberg School?

Since 2013. Michael Rosenblum, PhD, MS, invited me to visit when I was a postdoc at Harvard and as soon as I arrived, I knew I wanted to get a job here. The halls were full of laughter and people were talking and eating lunch together, the seminar was really vibrant, and I just fell in love with the Department.

Describe your work in one sentence.

I work on methods for causal inference from non-experimental data with complex statistical dependence. 

What motivated you to focus on causal inference? 

I was initially interested in causal inference because it's the perfect marriage between math and philosophy. It is very heady and deep, but also extremely useful; it really marries all of the things that I wanted in a discipline. Causal inference is about learning cause and effect relationships: Does this intervention causally affect this health or societal outcome? In causal inference you want to understand the underlying mechanism that gives rise to a link between an exposure and an outcome or a treatment and an outcome.

Where do you see causal inference going in the next five years? 

The infusion of causal inference concepts and methods and ideas into AI. Causal inference is exploding in interest right now, which is both exciting and disconcerting. When I started working in causal inference, we were a more niche, underappreciated field and now everybody wants to do causal inference. I think people are starting to appreciate the importance of causal inference in machine learning and AI.

What impact do you hope your research will have? 

I hope that my research on causal inference will protect against flawed analyses or unsubstantiated claims. I think that there's a human cognitive tendency to interpret associations causally, and if I could have any impact it would be to spread a cautionary stance towards over-indexing on causality when analyses are messy or don't license causal conclusions.

What do you enjoy most about being at Bloomberg?

I love the environment here. I love how friendly and collaborative everybody is. The Department really cares about its members, both the individuals and in the way the Department is set up. It's really supportive of faculty. I hear stories about other academic environments, and I just feel so grateful to be in such a friendly, supportive place. 

What would you say to a person considering a career in biostatistics? 

Biostatistics is an extremely flexible area to work in. It’s basically as broad as statistics, plus various areas of computational biology, biotech, and epidemiology. I think the types of things that you can do in biostatistics are so wide and varied that anyone who is inclined towards biostatistics would be able to find a happy home somewhere in the wide breadth of things you can do.