Research
I am an applied statistician who builds statistical tools to enable cool science.
I have deep expertise in causal inference with complex longitudinal data, and apply that expertise to a variety of methodological and application areas, including health policy,substance use, and oncology. Below is a selection of methodological and collaborative projects.
Methods Projects
*
Health policy evaluations often use difference-in-differences to estimate effects. However, interesting practical questions arise when using this approach with multilevel administrative data.
An extention to the target trial emulation framework to include special considerations when designing high-quality non-experimental studies to evaluate health policy.
Methods for the design and analysis of sequential, multiple-assignment randomized trials with continuous longitudinal outcomes.
Applied & Collaborative Projects
Selected projects represent
*
A series of collaborations with cancer researchers at the University of Pennsylvania and the University of Michigan investigating associations between biomarkers, genes, tumor morphology, and outcomes.
HeartSteps was the first-ever micro-randomized trial (MRT) designed to optimize a just-in-time adaptive intervention to increase physical activity among sedentary adults.
The opioid epidemic has claimed countless lives in the United States, and policy solutions may be useful in curbing it. This series of collaborations investigates the effects of health policies on opioid prescribing and treatment in the United States.
Patient-centered care is an important aspect of treating cancer, particularly as a way to inform comparative effectiveness research. This work builds an evidence base for treatment selection based on quality of life and other measures.