UW or Harvard?

[cabbysaurus]

TLDR: I'd rather go to UW because I love Seattle and it is a really great school for CS, but I don't have any guaranteed RA money or an advisory yet. There are a lot of faculty there I'm interested in working with, but none of them can guarantee me time or money at this point. At Harvard, I have a six year fellowship and an advisor who seems really great and says that I can work on pretty much anything. There are other professors who have work I'm interested in, but not as strong interest as those at UW. Any advice on what to do?

 

I am currently torn between two great options for a Computer Science PhD and would love to get some input please.  Below are the schools and some of my thoughts at this point about them.

School 1: University of Washington

Pros

• arguably one of the top schools for Computer Science
• lots of faculty members doing research that I am interested in
• large program with a lot of resources
• more students doing the type of research I'm interested in, so possibly more help/collaboration
• lots of connections to industry
• nice, new facilities
• I used to live in Seattle and have lots of friends there still

Cons

• even though there are a lot of potential advisors for me, the people I'm interested in working with have all expressed that they may not have time or money to support me
• the program is quite large and I'm worried that it may be hard to have a voice or that I may fall through the cracks
• some of the people I'm interested in working with have not been terribly responsive
• though funding is guaranteed, I don't have guaranteed funding through an advisor so may have to TA a lot 

 

School 2: Harvard University

Pros

• my initial advisor is really great - in fact people I met at other schools even say how great he is. We got along well in person and he is smart, responsive, and willing to let me work on essentially anything I would like
• the funding offer is awesome. I got a university fellowship for five years plus the university has a special fellowship for the final year, so I'm essentially covered for six years. It also pays more than UW and includes relocation
• I can get a co-advisor or choose to switch advisors if I would like, especially since I have funding already
• students there seem to collaborate with other faculty a lot, both in the department and also throughout the university and with MIT
• the program is smaller, so I feel more confident that I will have a voice and not fall through the cracks
• better/more job opportunities for my partner

Cons

• there is really just the one advisor I'd like to work with (at least at this point). I did meet other professors there who were nice and working on interesting projects though, so imagine that may change
• it's not as well-known for Computer Science as UW is
• the program is much smaller and there are very few students doing the type of research I'm interested in so there's less chance of collaboration
• the facilities are kind of old and cramped, but they are building a new building that will open in about two years. But that building is off campus, so I worry about how that will affect relationships across the university
• I don't know if I can handle winter and snow for six years straight

 

I'm really torn. I know that many people say that the advisor is the most important, in which case it seems that Harvard is a better choice since there is a professor there who is a good match for me and who really wants me there. Also the funding package I would get there would make things much less stressful for me, plus I would not have to worry about finding an advisor who has the time and money to support me.

However, my partner and I would both rather live in Seattle. I'm worried about the amount of uncertainty there, though, in terms of both an advisor match and funding. Two of the four professors I spoke to even brought up the option of getting a Master's there then a PhD elsewhere, which makes me feel that maybe they think that I will have trouble finding a match there. A third professor said that she would be concerned about me going there since the professors I most want to work with (herself included) may not be able to support me. I know that there are many different paths to getting a PhD, and that there are nearly a hundred faculty members at UW, so I would likely find someone to work with, but I don't know if that level of uncertainty is something that should make me go to Harvard instead. 

Any thoughts, advice, or words of wisdom would be much appreciated. Thank you!

Edited March 31, 2018 by cabbysaurus

[pterosaur]

I'm a current PhD student in CS at Harvard (so, of course, completely unbiased since I really like it here). I can't really speak to UW, since I didn't apply or consider it.

All of your "Pros" for Harvard are some of the things I really liked about the program, and why I came here. CS and all of SEAS is really flexible, in terms of coursework, advising, etc. It's also a nice collaborative community, which I think the size definitely helps with. I've also taken a couple of classes at MIT, which is nice. For me, the Boston area was a pro rather than a con; I went to undergrad in Boston and already knew I really liked the area. (I think the concern about winter is generally an overblown concern, since the winters are pretty mild by my Wisconsin standards; I can easily bike all but probably a week or two of the year without any special gear.)

As for your "Cons": I've definitely gotten the "not as well known" thing; when I tell people (outside the field) that I'm doing CS at Harvard, I have gotten asked a couple of times "Why not MIT?" But the program is still very strong, and students go on to have good options when they graduate. I even know a few people in my cohort who turned down MIT in favor of Harvard. If you're considering not staying in academia after graduation, the glow of the Harvard name in general is a plus. What facilities felt old and cramped to you? Maxwell Dworkin? That's where I'm based, and I really like it here. It's not as shiny as someplace like the Media Lab, but we're well-supported. Space is definitely at a premium right now, though, as the department keeps expanding while the new building is being built. I'm not a huge fan of them splitting SEAS and moving some stuff to Allston, but having purpose-built facilities for my group's research will be awesome. (I know my group will be moving there, frustratingly 3.5-4 years into my PhD, and I have to deal with moving a ton of robotics hardware, which will eat up a bunch of time...) Putting the new SEAS stuff in Allston, they're also hoping to strengthen the connections between it and the business school, which is already based over there.

But all of that seems pretty secondary to the major issues: funding and advising. It definitely sounds like funding is way more tenuous at UW, and even without a fellowship Harvard SEAS guarantees funding. Those comments at UW also sound kind of concerning -- do they tend to have a lot of people who leave after quals? (Our quals here are very chill; I have mine in less than two weeks and I'm not really stressed about it.)

In terms of advising, the rule of thumb I've heard is that there should be at least 2 or 3 people you could see yourself working with, in case the first option doesn't work out for whatever reason. I have a couple of friends here who have changed advisors for various reasons (change of interest, advisor leaving, not a great personal match) and have pulled it off without serious mishap. One ended up with a primary in Canada and a co-advisor here, but that's sort of the exception. I also know of a few people in CS who ended up with a primary advisor at MIT. For me, there was one person I really clearly wanted to work for, and the others were somewhat tangential in their interests, but I could make it work. That was enough to convince me I had options.

It definitely sounds like a tough decision, and both are strong contenders. Hopefully my thoughts can help a bit, though!

[spamhaus]

@cabbysaurus Congrats on having offers from two amazing schools! I can't compete with @pterosaur in terms of specifics about the programs, but I do have some insights regarding some of your concerns.

First, I went to undergrad in the midwest and hated the cold windy climate. When it would snow, I would walk in flip flops as my way to give a giant middle finger to the weather (a few years ago I hung out with my old roommate from undergrad, where he brought up how crazy I was to walk in flip flops in the snow). I hated the climate so much, that when I got a great offer to work in Chicago, I turned it down citing how I loathe cold windy weather. Fast forward a decade, I've just survived a winter in Alaska (where it's still snow covered), and I'm moving to UMass Amherst for a PhD. That is to say, despite not liking the cold, I bet you'll be able to handle it just fine.

Regarding connections to industry, Boston has plenty. Most of the major players in the field have offices in Boston in addition to their main campuses in Seattle, the Bay Area, etc. That said, you will likely have a bit more of a diversity of choices for internships in Seattle if you aren't interested in conducting research at one of the larger more well known companies.

Finally, I'm of the personal opinion that I'd prefer a smaller more intimate experience than feeling like I'm just a cog in a large academic lab. I've worked in industry for a decade, and I really preferred working at smaller companies rather than in a large corporate environment. Of course, that's personal preference and I'm not sure how/if that translates to academia.

Best of luck in your decision making process (if you haven't already made a decision)!

[TheMostPowerfulApplicant]

Dont come to Cambridge, it is super cold and drivers are terrible.

Stick with West Coast!

[cabbysaurus]

On 4/4/2018 at 1:54 PM, pterosaur said:

I'm a current PhD student in CS at Harvard (so, of course, completely unbiased since I really like it here). I can't really speak to UW, since I didn't apply or consider it.

All of your "Pros" for Harvard are some of the things I really liked about the program, and why I came here. CS and all of SEAS is really flexible, in terms of coursework, advising, etc. It's also a nice collaborative community, which I think the size definitely helps with. I've also taken a couple of classes at MIT, which is nice. For me, the Boston area was a pro rather than a con; I went to undergrad in Boston and already knew I really liked the area. (I think the concern about winter is generally an overblown concern, since the winters are pretty mild by my Wisconsin standards; I can easily bike all but probably a week or two of the year without any special gear.)

As for your "Cons": I've definitely gotten the "not as well known" thing; when I tell people (outside the field) that I'm doing CS at Harvard, I have gotten asked a couple of times "Why not MIT?" But the program is still very strong, and students go on to have good options when they graduate. I even know a few people in my cohort who turned down MIT in favor of Harvard. If you're considering not staying in academia after graduation, the glow of the Harvard name in general is a plus. What facilities felt old and cramped to you? Maxwell Dworkin? That's where I'm based, and I really like it here. It's not as shiny as someplace like the Media Lab, but we're well-supported. Space is definitely at a premium right now, though, as the department keeps expanding while the new building is being built. I'm not a huge fan of them splitting SEAS and moving some stuff to Allston, but having purpose-built facilities for my group's research will be awesome. (I know my group will be moving there, frustratingly 3.5-4 years into my PhD, and I have to deal with moving a ton of robotics hardware, which will eat up a bunch of time...) Putting the new SEAS stuff in Allston, they're also hoping to strengthen the connections between it and the business school, which is already based over there.

But all of that seems pretty secondary to the major issues: funding and advising. It definitely sounds like funding is way more tenuous at UW, and even without a fellowship Harvard SEAS guarantees funding. Those comments at UW also sound kind of concerning -- do they tend to have a lot of people who leave after quals? (Our quals here are very chill; I have mine in less than two weeks and I'm not really stressed about it.)

In terms of advising, the rule of thumb I've heard is that there should be at least 2 or 3 people you could see yourself working with, in case the first option doesn't work out for whatever reason. I have a couple of friends here who have changed advisors for various reasons (change of interest, advisor leaving, not a great personal match) and have pulled it off without serious mishap. One ended up with a primary in Canada and a co-advisor here, but that's sort of the exception. I also know of a few people in CS who ended up with a primary advisor at MIT. For me, there was one person I really clearly wanted to work for, and the others were somewhat tangential in their interests, but I could make it work. That was enough to convince me I had options.

It definitely sounds like a tough decision, and both are strong contenders. Hopefully my thoughts can help a bit, though!

Thanks for the input. I really appreciate having an insider's view. It was a tough choice, but I ended up choosing Harvard so will see you in MD this fall! 

[cabbysaurus]

On 4/7/2018 at 4:12 PM, spamhaus said:

@cabbysaurus Congrats on having offers from two amazing schools! I can't compete with @pterosaur in terms of specifics about the programs, but I do have some insights regarding some of your concerns.

First, I went to undergrad in the midwest and hated the cold windy climate. When it would snow, I would walk in flip flops as my way to give a giant middle finger to the weather (a few years ago I hung out with my old roommate from undergrad, where he brought up how crazy I was to walk in flip flops in the snow). I hated the climate so much, that when I got a great offer to work in Chicago, I turned it down citing how I loathe cold windy weather. Fast forward a decade, I've just survived a winter in Alaska (where it's still snow covered), and I'm moving to UMass Amherst for a PhD. That is to say, despite not liking the cold, I bet you'll be able to handle it just fine.

Regarding connections to industry, Boston has plenty. Most of the major players in the field have offices in Boston in addition to their main campuses in Seattle, the Bay Area, etc. That said, you will likely have a bit more of a diversity of choices for internships in Seattle if you aren't interested in conducting research at one of the larger more well known companies.

Finally, I'm of the personal opinion that I'd prefer a smaller more intimate experience than feeling like I'm just a cog in a large academic lab. I've worked in industry for a decade, and I really preferred working at smaller companies rather than in a large corporate environment. Of course, that's personal preference and I'm not sure how/if that translates to academia.

Best of luck in your decision making process (if you haven't already made a decision)!

Thank you for the input! I ended up choosing Harvard so will maybe see you at some Boston/Mass area hangouts. I know you're interested in ML and NLP - I'm more of HCI but have done a lot of NLP stuff so am happy to get in touch