Stat phd program: UChicago or Harvard?

[Vera]

Hi,

I think I am lucky to be accepted by the two programs (stat@Harvard, stat@UChicago). But at the same time, I feel it difficult for me to make the final decision.

1) Harvard. The name is so appealing to me! The night I got the news that I was accepted by that program, I simply couldn't fall asleep. I do have a very strong Harvard complex. The program is ranked 3rd or 4th. And MIT is just next to Harvard, so I may have great opportunities meeting people of great potential. But the size of faculty is so small (only 5 professors) and they mainly focus on Bayes and MCMC. (Of course, the 5 professors are all stars!) And I've heard from some other professors that they feel the reputation of this program is vastly overrated and they seldom consider PhD graduates from that program as their assistant professors. I am really sad to hear that because I really like the name of Harvard. I would really need a lot of courage to turn it down.

2) Chicago. The faculty size is large and there are several excellent professors whose interest is in stochastic processes, as well as other fields. Although the name of the university of less appealing and the program is ranked 6th, I feel the program is better (I got the idea just by comparing what is posted in the webpages of the two programs). The program seems to be a better match for me. I know that the program emphasizes theory very much and many PhD graduates find positions in universities. The placement is good! But, I personally prefer to live in Boston.

I am not sure about the research direction I am mostly interested in, since I major in math and lack deep understanding of all the branches of statistics. Currently, I feel that I am interested in stochastic processes and MCMC. On the one hand, Professor Kou at Harvard is good match, but he seems to be the only professor at Harvard who are interested in stochastic processes. Of course, Harvard is very strong in regard to MCMC. On the other hand, there are many professors at Chicago interested in stochastic processes, and the research in Chicago is broader and deeper (some professors said that to me). I really can't decide which one to go. I might feel regretful to turn down Harvard, but I might have a better future if I attended University of Chicago.

What should I do? Can someone share some opinions? Thanks!

[Pierre]

I don't know how much this will help you, but having considered both programs when I looked at schools here are my two cents:

My background is more in math than in statistics, I have a degree in engineering physics with a master's in engineering mathematics (lots of probability theory and related mathematical topics). From my perspective, the program at U. Chicago is a far better math than Harvard. First of all, if you're not entirely sure on what direction you want to take it seems better to go to a department that gives you more choices rather than one whose faculty have very limited research interests. Also, my impression is that at Harvard there is a lot of focus on applying statistical theory to biology and related topics, something that wasn't very attractive for me. At the same time, with my interests I would choose Chicago and the possibility to work with Prof. Lawler any day of the week so I am of course a bit biased

In the end, I think you just have try and decide on a general direction in which you want to take your studies - should you feel that the few topics that are emphasized at Harvard would suit you, then by all mean go there. If you instead have a hard time deciding and want to "browse" some of the subjects more before you make a commitment, I would suggest Chicago. Don't let the name Harvard get to you too much. I know it's hard, I also had only the "famous" universities on my original list. Being an international, you seldom here about other places than the really famous ones and thus you think that they are to prefer for all subjects. Three months later and lots of consideration and looking at schools, my list was very different and was much more based on how my research interests compared to those of faculty members at the different departments. If you have a strong personal bias towards living in Boston that should of course be taken into account, but remember that you will have to work really hard for 4-5 years and therefore should make sure that you can choose a topic that you really enjoy.

Good luck and remember that this is a quite nice choice to have

[robben]

if you like MCMC, stochastics stuff uchicago is the way to go for sure and it doesn't sound any inferior to harvard. this isn't math it's stat. cornell is #1 but chicago and berkeley aren't far off.

[sciencegal]

Harvard is a top school in my field as well, but has many limitations you describe. While the name is famous to everyone, people in my actual field often like to hire from other schools because they have strengths Harvard does not. Chicago is well known too, especially within the field of statistics and I've seen it ranked higher than Harvard on some lists. I'm somewhat familiar with both of these programs and I personally would choose Chicago over Harvard myself. I assume, as an international, you are unable to visit? I know that really helped me when I chose my program.

[The_Hanged_Man]

Small departments can be problematic when it comes to finding advisors too. What happens when the one guy at Harvard who shares your interests already has too many advisees, has a personality conflict with you, or kicks the bucket? Your interests will probably also change over the course of the program so the bigger program at Chicago may be more accommodating.