Yale (Stats) vs University of Washington (Biostats)

[DMX]

Trying to decide between these two programs. Obviously feel very lucky to be in a position of choice. I won't have the opportunity to visit either program (I am overseas), and need to make a decision this month. I want to stay in academia upon graduation.

 

Some pros and cons:

 

Yale

 

Pros:

- I like the flexbility of statistics, as I have an interest in a wide variety of discplines (CS, biology, economics, political science)

- Good theoretical training

- Good academic placement (recent grads went to Berkeley/Cambridge/UNC)

- Department is expanding rapidly (at least according to the program director)

- Yale brand name--I know one shouldn't emphasize this too much for a PhD program, but there's a good chance I will go back to my country (Asia) and try to get a faculty position there. I believe the Yale brand name will not be insignificant in this case.

 

Cons:

- Maybe too theoretical? There's a couple of faculty working on statistical genetics etc. but most of the faculty's interests seem to be in theory (especially stochastic processes). I am more of an applied guy.

- Less than 10 faculty--less room to explore different research areas.

- USNews ranking places Yale outside of top 30--how accurate is this? Is it due to the smallness of the program?

 

University of Washington

 

Pros:

- Top 3 for biostats (some would say second only to Harvard?).

- Between Biostats/Stats there's ~100 faculty. A lot of room to explore different areas of research

- Heavy emphasis on theory (not as much as Yale, but definitely one of the more theoretical biostats programs)

- Biostatistics is more applied (almost by definition), and I am more interested in applications of statistics rather than theory (I do want a good theoretical foundation however).

- Excellent academic placement (something like 60%+ placement in academia)

 

Cons:

- Don't want to limit myself to biostatistics academic positions

- Back in my home country, biostatistics is not as established as statistics, so may be hard to get a faculty position back home.

 

If I were sure that I wanted to be in biostatistics forever, it would be an obvious choice to go to Washington. But I want to have as many options open as possible. Any input would be appreciated.

Edited March 4, 2013 by DMX

[little blue burry]

If I were you, I would choose Yale. Its stat program doesn't rank high just because it is not a big program. Yale is pretty good at machine learning and data mining, which are not theoretical at all. You can earn a lot of money in those areas. Considering not staying in the U.S., Yale might be better than Washington because of the name.

[cyberwulf]

I would favor Washington, for the following reasons:

 

- Between stat and biostat (which are closely linked there), there are plenty of faculty working on machine learning, likely more than Yale. 

 

- You'll have more options; a department with only 10 faculty is awfully small for someone like you who doesn't have any set research interests.

 

- Students at Washington get enough theoretical training that the better ones can compete for jobs at good stat departments, at least in the U.S. Don't know how graduates fare in Asia.

 

- I just don't buy the "Yale name" argument, particularly if you will be looking for an academic position in Asia. Many faculty at non-U.S. institutions were trained in the U.S., and know quite well that UW is more highly-regarded than Yale in stat/biostat.  There's a reason that UW attracts so many top applicants from China and Korea; their advisors all recommend that they go there!

 

Edited March 5, 2013 by cyberwulf

[Shostakovich]

I also disagree that you would be limiting yourself to only Biostats positions coming out of UW. The theory sequences are identical for Stats and Biostats students at UW for the first two years, so you get a theoretical flavor unseen at most (if not all other) Biostats departments. There are also faculty members with joint appointments in both departments with varying levels of orientation in theory and application. Perhaps you won't be as competitive as a Yale graduate in a Pure Stats department though...

I believe cyberwulf is right about Asia as well. While Yale may be a more prestigious institution overall (I know this stereotype can be huge in Asia from growing up there), most sources suggest that the people who matter (people giving you jobs) know what's going on overseas and regard UW as one of the top Stats departments (perhaps behind Stanford, Berkeley, Harvard, Chicago).

Also for me personally I'd also consider 'soft' factors such as the weather and living environment, which I feel would be better for many people at Seattle (as opposed to New Haven).

Congrats on your acceptances, and good luck on your decisions! Let me know if you have more questions about UW or decide on it, but I think I'm biased since I didn't really apply to Stats departments and I had a large dose of 'Why UW' from their visit days

[echlori]

Amongst other reasons (namely, the significant other being already there), I applied to UW biostats as they seem to have some good work using Machine Learning coming out of there. Also, the faculty at UW seems to cover quite a lot of ground which I find to be a plus.

 

I didn't look at Yale, so I can't say much, but I'll just say how insanely jealous I am that you actually got an offer from UW biostats  Congrats!

Edited March 5, 2013 by echlori

[wine in coffee cups]

If I were you, I would choose Yale. Its stat program doesn't rank high just because it is not a big program. Yale is pretty good at machine learning and data mining, which are not theoretical at all. You can earn a lot of money in those areas. Considering not staying in the U.S., Yale might be better than Washington because of the name.

The OP didn't mention machine learning specifically so this may be entirely irrelevant to his/her interests, but I would dispute ML/data mining as a reason to choose Yale stat over UW biostat. The UW is spending gobs of money in a well-publicized effort to become one of the top universities for ML. Lots of new faculty have been hired in this area just in the past year (see http://news.cs.washington.edu/2012/09/06/uw-cse-makes-game-changing-hires-in-machine-learning-big-data-computer-vision-and-computer-systems/ for just the ones as of last fall -- definitely more hires coming for the next academic year) and a new concentration is being developed for ML/big data to launch this year. I don't know how the ML/big data initiative has been to advertised to current/prospective biostat students, but is certainly a big new thing for UW stat/CS students and something I think the biostat students would have ready access to. It's a very bold claim to say one would get a better experience in doing anything related to ML at Yale given how serious the UW is about beefing up its ML research and the combined size/reputation/breadth of the stat/biostat/CS departments relative to Yale's equivalents.

[DMX]

The OP didn't mention machine learning specifically so this may be entirely irrelevant to his/her interests, but I would dispute ML/data mining as a reason to choose Yale stat over UW biostat. The UW is spending gobs of money in a well-publicized effort to become one of the top universities for ML. Lots of new faculty have been hired in this area just in the past year (see http://news.cs.washington.edu/2012/09/06/uw-cse-makes-game-changing-hires-in-machine-learning-big-data-computer-vision-and-computer-systems/ for just the ones as of last fall -- definitely more hires coming for the next academic year) and a new concentration is being developed for ML/big data to launch this year. I don't know how the ML/big data initiative has been to advertised to current/prospective biostat students, but is certainly a big new thing for UW stat/CS students and something I think the biostat students would have ready access to. It's a very bold claim to say one would get a better experience in doing anything related to ML at Yale given how serious the UW is about beefing up its ML research and the combined size/reputation/breadth of the stat/biostat/CS departments relative to Yale's equivalents.

Interesting--can biostats students work with stat professors? My research interest is actually in ML. Yale has a couple of faculty working on ML but they are definitely not the department you think of when you think of ML. (they're far behind Stanford, Cornell, CMU, Michigan, Wharton).

 

Thanks for the comments so far everyone! Right now I am leaning towards Washington. Still have to do more research on faculty at both places though.

Edited March 5, 2013 by DMX

[JZappa]

Interesting--can biostats students work with stat professors? My research interest is actually in ML. Yale has a couple of faculty working on ML but they are definitely not the department you think of when you think of ML. (they're far behind Stanford, Cornell, CMU, Michigan, Wharton).

 

Thanks for the comments so far everyone! Right now I am leaning towards Washington. Still have to do more research on faculty at both places though.

 

Usually I think the answer is yes. You might have to have an advisor in the biostats department but I know it is at least somewhat typical to have a coadvisor in another department.

[health_quant]

On 3/4/2013 at 9:15 PM, DMX said:

Interesting--can biostats students work with stat professors? My research interest is actually in ML. Yale has a couple of faculty working on ML but they are definitely not the department you think of when you think of ML. (they're far behind Stanford, Cornell, CMU, Michigan, Wharton).

 

Thanks for the comments so far everyone! Right now I am leaning towards Washington. Still have to do more research on faculty at both places though.

I would agree with you regarding Stanford and CMU, but UW versus Cornell, Michigan, or Wharton is definitely debatable (especially the latter).

 

Among the new and/or incoming ML-related hires at UW: Carlos Guestrin (formerly a prof at CMU in ML/CS); Emily Fox (formerly a prof at Wharton); Ben Taskar (CS/Stats at Penn Engineering/Wharton). There's definitely a big push from UW to build up ML in the stats department (while integrating with UW CSE), while biostats and stats have historically been linked quite closely. 

[zoomzoom]

I was admitted to UW-Seattle Math recently. However I'm interested in doing something more applied/statistical for my PhD. Does anyone know how hard it is to switch or get an advisor in the stats/cs department? Any current math/stat students at UW with experience?

[cetaphil]

dollar - Some stat faculty used to be in the Math department. Also, some stat faculty teach advanced probability which is cross listed in the math department, so you would certainly be able to work with them

 

dmx - biostat and stat students can be advised by any prof in biostat/stat without any paper work. you should go to uw. you will certainly be able to do ml work there. 

[Biostat_Assistant_Prof]

Another vote for UW Biostats... One of the top 3 departments in the world and I've heard you'll get a good dose of theory even in their Biostat department. Plus, Seattle vs New Haven should be an obvious choice in terms of climate/location.

[enomomo]

Thanks for the information cetaphil.

[MCMC]

I would choose Yale Stats if I were in your position. I expect you to have a much better chance of achieving your specific career goal – getting a satisfying job in Asia -- after graduating from Yale Stats than UW Biostats. I think you’d have more flexibility during your job hunt in Asia because of Yale’s outstanding placements in non-academic positions in Asia, which may end up suiting your interests better than academia since you say you have interest in a wide variety of fields. Even if you end up seeking a senior faculty member to advise your ML PhD dissertation to enter academia in Asia, I still believe Yale would be the superior option. Although there are more faculty at UW working on ML than at Yale, ML is more ingrained in Yale’s department (I think the Chair of Yale Stats works on ML) and ML has been a historical strength of Yale’s. The reputation of these departments probably lags a few years behind reality, so even if UW continues hiring great ML researchers while you are a PhD student, you may not reap the rewards in time to give you superior job placement in Asia.

[DMX]

I would choose Yale Stats if I were in your position. I expect you to have a much better chance of achieving your specific career goal – getting a satisfying job in Asia -- after graduating from Yale Stats than UW Biostats. I think you’d have more flexibility during your job hunt in Asia because of Yale’s outstanding placements in non-academic positions in Asia, which may end up suiting your interests better than academia since you say you have interest in a wide variety of fields. Even if you end up seeking a senior faculty member to advise your ML PhD dissertation to enter academia in Asia, I still believe Yale would be the superior option. Although there are more faculty at UW working on ML than at Yale, ML is more ingrained in Yale’s department (I think the Chair of Yale Stats works on ML) and ML has been a historical strength of Yale’s. The reputation of these departments probably lags a few years behind reality, so even if UW continues hiring great ML researchers while you are a PhD student, you may not reap the rewards in time to give you superior job placement in Asia.

Hi, thank you for your comments. Not sure I agree--on what basis do you say that ML has historically been a strength of Yale? Chair of Yale stats department does work on ML according to the website, but if you look at his publications you see that it's not really ML.

[MCMC]

Hi, thank you for your comments. Not sure I agree--on what basis do you say that ML has historically been a strength of Yale? Chair of Yale stats department does work on ML according to the website, but if you look at his publications you see that it's not really ML.

 

 

No problem. While working on my own research, I have come across many high-quality papers from Yale faculty and Yale graduates on ML as it relates to high-dimensional data analysis, bioinformatics, and statistical genetics. When I applied to Ph.D. programs a couple years ago, I heard from statistics and computer science professors that I should apply to Yale if I wanted to study ML and apply it to these kinds of areas. These professors mentioned that Yale Stats had a great reputation in ML because the department was actively supporting that type of research while also encouraging collaboration with researchers from Yale’s CS department. 

 

Best of luck in making your decision.