Stephens’ research focuses on a wide variety of problems at the interface of statistics and genetics. His lab often tackles problems where novel statistical methods are required, or can learn something new compared with existing approaches. Much of that work involves developing new statistical methodologies, many of which have a non-trivial computational component. As data sets get larger and larger, that work often involves modern methods for high-dimensional statistics, making extensive use of Bayesian hierarchical models to borrow information across data sets or sampling units.
Stephens has helped develop several widely used software tools and statistical models, and is currently working on making his lab’s research more open, reproducible and extensible.