Spring 2024 Bayesian Learning Group
On April 26, we held our last meeting of the Spring 2024 Bayesian Learning Group! The Bayesian Learning Group is an experimental learning community of researchers and students of all career levels from across the University, coming together to discuss and co-develop skills in Bayesian statistics.
The goal of launching the Bayesian Learning Group was to fill a particular niche in the university ecosystem. It can be difficult to acquire new statistical skill sets, especially as one progresses out of formal classroom learning, or if the problems that one needs to tackle are significantly more specialized and complex than the idealized examples that appear in textbooks and homework problems. The BLG offers a community, accountability, and support for anyone trying to learn more about Bayesian statistics. Beyond the content we cover, we hope to connect people from across different disciplines and career stages around a shared interest and mutual support in learning new concepts and tools.
Jessica Guo and Renata Diaz, both from our group, organized the learning group - but the richness of the sessions was really driven by the participants! We met every two weeks to discuss concepts and work through examples drawn from Richard McElreath’s Statistical Rethinking. In between sessions, participants read the book and watched video lectures on their own. Our conversations ranged from the philosophical underpinnings of hypothesis testing, to the effects of baby foxes on the size and territorial success of fox families, to the nuts and bolts of fitting models in Stan, brms, and McElreath’s rethinking package. Beyond our regular offerings, we also had model “show and tell” sessions, where volunteers walked us through models they’d developed in their own work, and for our last session, Dr. Henry Scharf (a real-live, practicing Bayesian!) joined to give a guest lecture.
We were amazed at the level of interest and engagement we saw for the BLG. Our initial signup list had nearly 50 respondents, and attendance in session peaked at more than 60 before dropping off as the semester picked up in intensity. We were excited to see this level of enthusiasm for a voluntary learning collective focused on some robust, and sometimes challenging, topics!
We’re excited to announce that we will be continuing the Bayesian Learning Group into summer of 2024! Taking into account feedback we received from spring participants, we will move at a slower pace, lean less heavily on reading and watching content outside of session, and focus more on working through examples (and less on the philosophy of statistical inference). We will be meeting for shorter periods of time (1 hour blocks), on a schedule yet to be determined. You do not need to have participated in the learning group in the spring in order to join us for the summer. We intend to make the sessions as modular as possible, so that you can join us at any point and still benefit from the experience. To register and express your scheduling preferences, see the signup form here!