Lee Kennedy-Shaffer, PhD

My research broadly focuses on the role of statistics and data science in answering questions about causality and time trends. In particular, I aim to develop study designs and analytic methods that can handle time-varying and location- or context-specific effects and that can give clear policy-relevant interpretations. Some specific areas of focus are:

Causal Inference Across Time and Space: How do we describe causal effects, like the impact of health policies, that vary over time and by location? How do we estimate these effects using both randomized and observational studies? What claims can we make and what claims can we not make? These questions drive both methodological research in cluster-randomized trials, stepped-wedge designs, and quasi-experimental (difference-in-differences and synthetic control) analyses and conceptual work on the role of statistics in society. This work has been published in journals such as the American Journal of Epidemiology, the American Journal of Public Health, Statistics in Medicine, and Clinical Trials.

Evaluating Infectious Disease Interventions: What makes infectious diseases different than other health outcomes? How does that shape our statistical methods and interpretations? These questions drive methodological work on study design, new uses of data streams from COVID-19 and beyond, and collaborative work with infectious disease surveillance teams. This work has been published in journals such as Science, Lancet Microbe, Epidemiology, the American Journal of Epidemiology, Vaccine, and the American Journal of Public Health.

Statistics History, Communication, and Education: How do social forces affect the development of statistics, and how does statistics in turn affect the development of science, policy, and society? Given this, how can we best communicate statistical ideas and results, teach statistical thinking, and train the next generation of scientific researchers, policy-makers, and communicators? These questions drive educational innovation and research and writing on new ways of teaching, new understandings of the history of statistics, and examples that can communicate key statistical ideas. This work has been published in journals such as the Journal of Statistics and Data Science, The American Statistician, the Food and Drug Law Journal, and Significance.

Collaborative Research: How should clustered data be handled in an observational study? How can a cluster-randomized trial best answer a research question? These questions from collaborators have driven the use of existing and new methodologies in fields such as nutrition, thoracic surgery, and lab studies of vectors of disease. This work has been published in journals such as the Annals of Thoracic Surgery, the Journal of Infectious Diseases, and the Journal of the Academy of Nutrition and Dietetics.