Biostatistics is a field that requires a precise form of creativity. Enabling students to embark on their own creative journey is the goal of Josh Warren’s teaching.
Warren, an associate professor of biostatistics, teaches Bayesian Statistics (BIS567) to a diverse group of students, including those from biostats, MPH, MS, and PhD programs, as well as individuals from Yale’s computer science and statistics departments. In his classroom, he underscores the widespread applicability of the statistical modeling that students use in this course.
Bayesian inference is a method of statistical inference where prior beliefs for model parameters can be incorporated into an analysis and updated once data are observed. The techniques for fitting models in a Bayesian setting are extremely flexible and allow for dataset-specific models, which may more accurately reflect underlying mechanisms and hypotheses, to be developed. It is extremely flexible, allowing students to adapt and tweak frameworks based on the unique characteristics of their data. Students often reach out to him years after taking his course to tell him about their applications in a vast array of public health scenarios.
“Being able to write down and ultimately make inference for the exact model that your dataset and research project requires is a major strength of working in the Bayesian paradigm,” Warren said. “Teaching students the skillset needed to achieve this freedom is a major goal of the course.”
As a student, Warren appreciated consistency and predictability in his studies. To this end, he focuses on delivering a clear and organized message throughout his course, with each lecture following a similar cadence from week to week. Students face broad variations in teaching styles as they navigate their time at YSPH. Warren’s approach resonates with many of his students, who find comfort in knowing what to expect in each lecture.