Cassandra Pattanayak

Jack and Sandra Polk Guthman '65 Director, Quantitative Analysis Institute, & Senior Lecturer in Quantitative Reasoning and Mathematics

Statistician specializing in causal inference; creating the Quantitative Analysis Institute, to expand the role of statistics at Wellesley.

My research focuses on causal inference, with applications to education, law, and health. Two phenomena may be correlated, but how can we design a study to address whether one causes the other? In particular, I develop tools that allow rigorous causal inference for applied projects that are complicated by practical constraints. I have worked on diagnostics for covariate balance in non-randomized studies, methods for addressing chance imbalances in randomized experiments, and a Bayesian approach for analyzing outcomes from propensity score subclassified designs. My applied work has included measuring the impact of offering free legal assistance to indigent clients and developing best practices for in vitro fertilization. I am also interested in statistics education and course design.

As Guthman Director of the Quantitative Analysis Institute, my goal is to expand the role of statistics in both research and teaching at Wellesley. I collaborate with faculty and student researchers from a variety of fields, provide and coordinate statistical consulting, and run workshops for faculty and students. I believe that students should master fundamental statistical ideas that generalize across disciplines, along with the practical skills necessary to use this knowledge. I teach the Quantitative Analysis Institute Summer Course, designed to introduce advanced statistical skills and support students in their current and future research projects. I also teach a First Year Seminar focused on causal inference, and I look forward to developing additional courses.

I enjoy playing the trombone in Boston-area orchestras, exploring farmers’ markets, and spending time with my husband.

Education

  • B.A., Harvard University
  • M.A., Harvard University
  • Ph.D., Harvard University

Current and upcoming courses

This course introduces the theory of statistical inference: given a data set, how do we estimate the parameters of probabilistic models like those introduced in MATH 220/STAT 220? What is the optimal way to make use of the information in our data? Topics include the theories that underlie traditional hypothesis testing and confidence intervals, such as maximum likelihood inference and sufficiency. The course will also cover Bayesian techniques for point and interval estimation and resampling approaches, such as the bootstrap.