Duke vs Michigan vs Chicago Stats PhD

[jMazek]

HI! 

So I got accepted into several programs and now have to choose between Duke, Michigan Ann-Arbor, and UChicago. I attended the online visits for Duke and UofM but Chicago did not do one. I was very impressed by Michigan and the broad areas that its faculty covers. Duke is obviously famous for Bayesian Stats but I don't really wanna do that unless it's computational methods for BS like Variational Inference or Normalizing Flows. In Chicago, I was intrigued by Prof. Vietch's research but he replied that he will not take on any more students for the next few years as he was at "capacity". I am mainly battled between Duke and UofM tbh.

Both departments seem very friendly and collaborative. UofM seems a better fit for research and it's a much bigger university, however, Duke's campus seems very nice and the climate is definitely better for me. Moreover, Duke's stipend makes you enjoy a slightly better standard of living compared to Ann Arbor.

Any advice? 

[Stat Assistant Professor]

It sounds like you may (strongly) prefer University of Michigan. If you find that you are still interested in variational inference, normalizing flows, and the like, then I note that there are some strong researchers at UMich who have expertise in these areas (e.g. Jeff Regier and Yixin Wang). And it sounds like you are open to other areas as well. Duke is world-class for Bayesian statistics, of course, but you seem to have some reservations. 

You should go with your gut!

[bayessays]

You really can't go wrong - both places have many great professors and you'll be set up for success either way.  Only you can decide whether you think research fit or weather is more important to your happiness the next 5 years.  A lot of people change research interests as they go along, and I'd encourage you to keep an open mind because you can always pivot research areas during a post-doc.