glori

I think doing thesis-based programs will help you to become more confident in doing research, but if that's not something you plan on focusing on after you graduate and you don't intend on getting a PhD, it's not as much of an issue. If you do a non-thesis based program it won't be the end of the world if you later decide to do the PhD, especially if you're spending time working in between. Another option in non-thesis based programs would be to pursue some research opportunities with professors in your program instead. Getting publications and a broad base of experiences never hurt anyone, and who knows, you might find that you enjoy it!  

I was in a professional masters program (MPPA) that happened to be thesis based.  I am currently in a PhD program in a similar field which required everyone who entered already have a masters of one sort or another.  I can tell you that the thesis better prepares you for a PhD program if you decide to do it later.  Non-thesis masters are fine, but the students I have met who had professional-non-thesis-masters are struggling in the quantitative and qualitative analysis coursework.
If you have no intention of going into a PhD program, then a non-thesis professional masters may be right for you.  It won't shut the door on a PhD, but it will make it easier if you decide to go.
Oh.. one more thing.. finishing a thesis provides a lot more personal satisfaction.. you truly will feel like an expert in whatever it is you research.
IMHO, of course

This one I think is relatively tougher than the UNC vs. Columbia post. I guess the big advantage of Berkeley is that (I think) you don't have to take written qualifying exams. They also (I believe) allow you to pretty much take whatever classes you want. Although I think in a way the second thing can be a disadvantage should you end up not having the same background as your peers or what other schools would call "core" classes. Berkeley also is very tiny. And in my opinion, it is kind of hard to see who belongs to biostats and who belongs to stats (it seems every prof advises almost exclusively stats students). Also only having 2 other people in the cohort would personally make me feel pretty isolated. And while Berkeley has a sizable stats program, in my experience there is a big disconnect between stats / biostats people (I am not sure why). I would also be concerned about the research diversity of the department.
I think Columbia is slightly larger than Berkeley. On their web site, it seems they have some research diversity (at least more than Berkeley). It also is easier to distinguish the program between the stats program since it is housed in the School of Public Health (side note: Berkeley people say their biostats program is also in their public health school--and you gotta apply to SOPHAS. However, their web site more indicates it's a group within statistics). 
I know Berkeley is very strong in causal inference--if you're interested in that for sure, then there's an obvious choice. I think Columbia diversifies their research specialties, so if you're not committed to causal inference, Columbia might be a better choice.
 
 

It's hard to say for certain. To the extent that some firms recruit exclusively or almost exclusively from Ivy League schools (for some firms, brand name > program rank) then there could be an advantage of UPenn. On the other hand, there's a lot of data science recruiting in Seattle (although it might be more regional--pertaining to jobs on the west coast). You probably couldn't make a bad decision either way if you want to go to industry. I suppose I would choose UW for data science just because I know that UW is committed to solving big data problems in their stats and CS departments. 

If you look at a lot of the Ivy League faculty in biostatistics (particularly Yale, Brown, etc.), many of them came from Ivy League PhD programs, despite that UW is a more "highly ranked" program. Just something to consider.

If you are concerned about being limited in career options, UW biostat and stat students take many of the same core theory courses. Biostat students can get an advisor who is in the stat dept as well. Woohoo!
One nice benefit of going to UPenn is you can brag at dinner parties that you went to an Ivy League school!

Okay, I applied to U of T Stats as well and I think I have a good shot of getting in. Maybe it's better to do a Stats PhD and collaborate with the Biostats Department.

Berkeley has had much better academic placements than Columbia in the past 10 years: Kasper Hansen (Hopkins), Sherri Rose (Harvard), Maya Petersen (Berkeley), and Alex Luedtke (Fred Hutch/UW), among others. I can't think of anyone going to a top-tier biostat place recently coming out of Columbia biostat.

From my perspective, UNC has a better program and the cost of living in NC is way more affordable than NYC. I wouldn't worry about living in "the south" as far as Chapel Hill and the surrounding areas are concerned, that's not an issue. I grew up in the south US... the larger cities are generally not very "southern"

Hey Jake,

I am not a Statistics person, so I don't know too much about that area. The only thing I can say is that the official U.S. News and World Report rankings put UNC's statistics program in the top 10, while Columbia's is in the 20s (see here http://grad-schools.usnews.rankingsandreviews.com/best-graduate-schools/top-statistics-schools/rankings). But I do agree that when people hear Columbia, they tend to think that it's more prestigious than UNC, but that probably depends on what group of people you are talking to. If you speak with Statistics professors, I am sure they have high regard for UNC if it's so well ranked.

Also, I graduated from UNC as an undergraduate. So I will agree with you on the atmosphere and the living situation in Chapel Hill. It is quite a bit more quiet than NYC, but I don't think you will get bored. There are plenty of things to do and entertain yourself with if that's what you are concerned about. Also, the living costs are RIDICULOUSLY low. You can rent a very nice and spacious two floor townhouse with a roommate, which is very close to campus, and has all amenities including a W/D in your unit and pool/tennis court/beach volleyball court for $840/mo. I am speaking from first hand experience. That's definitely a plus. Further, UNC is a public university whereas Columbia is a private one. So I could be wrong, but at least in my area of study public universities are much more generous with RA/TA funding for Masters students than private ones.

So if I were you, I would be inclined to choose UNC. But you should definitely get more opinions. I hope this advice helps in your decision.

Cheers,
angusBF

Columbia's program is highly respected in the city - as far as quantitative rigor goes, even more than SIPA. It all depends on what you want to do with the degree. Remember, Columbia has close ties to the UN and other employers in the area. I know a lot of people at the Department of Economic and Social Affairs - Statistics Branch of the UN, who got their PhD in statistics at Columbia. Which brings me to the question, if you want to get a PhD, why do you bother with getting a separate Masters? Almost all PhD students get full funding, AND you will receive a masters on your way to the PhD degree. My advice would be to forget the free standing masters before the PhD and just try for a PhD alone. The separate Masters will be a waste of your money and time.

I'm not sure about Columbia biostats but there's a lot of people talking how their Stats MS is basically a joke and they admit like 250+ students and get taught by adjunct faculty (I would guess their biostats is not too much different). No one is going to argue that Columbia has more "brand name" power than UNC, so if that's what matters to you then go to Columbia. If you want to work in public health, I imagine the "biggest" employers will come to the more highly ranked schools in terms of public health as opposed to overall ranking, so bear that in mind. I'm sure the cost for Columbia will also be astronomical compared to UNC.