Choosing a PhD program: Duke, UT Austin, UMich, Cornell

[PVC00]

Hi everyone, I am an international student and I have been admitted to a few American PhD programs in statistics: Duke, UT Austin, UMich and Cornell. I am having a hard time making up my mind. I feel like my top choices would be Duke and UT Austin, because they have strong faculty members that do research in Bayesian statistics, which is the topic I am interested in at the moment. However, my knowledge of statistics is not very broad, so I would like to have the chance to explore other fields of research. Moreover, I am under the impression that UT Austin could be a better place to live in, because I am afraid that I might get tired of living in a small college town like Durham. Finally, at the moment I don’t see myself pursuing a career in academia, so I would also consider which university offers more opportunities in terms of industry placements.

What is your opinion? Are there any other factors you think I should consider? At the moment I lean towards UT Austin, but I fear that I will regret it, because Duke’s department is more prestigious. Does this make sense?

[bayessays]

All of those departments are prestigious enough where, even if you would want to go into academia, I think you could easily choose any of them.  It sounds like you want to go to UT-Austin.  Read what their recent PhD grads are doing, lots of good data scientist jobs: https://stat.utexas.edu/alumni-and-friends/phd-and-postdoc-alumni

[statman2000]

I think objectively all but Ithaca NY are tremendous places to live. Weather is honestly probably best in Durham. Also, while you're living in a "small college town", Durham is apart of the broader Research Triangle that will provide plentiful opportunities to stay entertained (the other 2 being Chapel Hill and Raleigh). Ann Arbor is consistently ranked as one of the best places to live in the US. That department has tremendous variety in research being conducted if you want to keep your options open. If you're dead set on Bayesian Statistics, there's no better place to go than Duke. 

Seems like you have more research to do on the cities and programs though. All are strong enough that they will never inhibit industry placements. Cornell may arguably be the best bet if you are dead set on industry if you care about prestige (likely that the Ivy League label will carry more weight than individual strength of department to those not in the loop on Stat department strength).  

[ohrnor]

I also got into Austin and talked to a faculty member and some students so far. Everything sounds great to me, and I'm not even that interested in Bayesian statistics. They are expanding (hiring more faculty and admitting more PhDs) and are/will be good at many other areas, in particular machine learning. Had I not have another offer that I can't say no too I would have chosen Austin, over NCSU which I also got into. I didn't get into/apply for other programs so I can't really speak on them but I'd agree that Austin is really good (location, stipend, vibe, research, etc.) and I can easily see their ranking rise. BUT all other schools are great so it really is up to your preferences. 

Edited March 19 by ohrnor