I have degrees in Psychology (University of Wisconsin-Madison), Public Health (University of Michigan) and Applied Statistics (New York University). The focus of my work is on developing software and methods that can bridge causal inference methods from the field of statistics and the work of applied researchers. I lead development of thinkCausal, a software that provides a scaffolded platform for fitting Bayesian Additive Regression Trees (BART) for causal inference problems. BART methods have the potential to make fewer assumptions and have been shown to perform better in many settings than popular methods such as propensity score matching. Simultaneous to my software development, I have worked on developing simulation methods to evaluate the effectiveness of novel applications of Bayesian Additive Regression Trees for educational research settings such as stan4bart which extends traditional BART models to multilevel contexts.
My career began as an applied researcher running randomized controlled trials to estimate the causal effects of brief classroom writing interventions on educational disparities. I understand the needs of educational researchers and the focus of my current work is on developing tools and methodologies that fit those needs.
In addition to my methodological work, I have interests in real world evidence, and how people think about causality and process statistical evidence. I have given talks at the American Causal Inference Conference and the New York City R conference.