bayessays

Not Iowa.

If you want to keep your options as open as possible, you'll be better off attending a higher-ranked program. Florida and Rutgers are definitely the best programs you've gotten into. I think Florida might be a better fit for you, as they definitely have a good amount of people doing more "applied"/methodological work. Rutgers has a lot of very theoretical people, though they have a couple people doing somewhat more applied stuff. I'd recommend reading through the websites of every professor at programs you're interested and counting how many people at each program look interesting. Don't just take my word for it! On the next level I would say is Boston, Virginia, and Rice. These are solid programs with some good researchers, but they don't have superstar professors like Florida and Rutgers do. Boston is associated with the math department, so they have a lot of probability researchers which won't be interesting to you. They have Eric Kolacyzk who does exciting network stuff, but your options for advisors might be slim. Virginia and Rice probably have research that would be more up your alley - it will be more difficult to land a very-high ranking academic job from these places compared to Florida/Rutgers. I would put Northwestern somewhere between Stony Brook and BU/UVA/Rice -- they do have some very applied people like Hedges (whose students get good academic jobs in education departments) and Tipton though, if you want to do that type of work, that might be something to consider but it's not an elite stats program. I don't think Stony Brook is a really serious option unless you have personal reasons to attend there or there is a special researcher you really need to work with.

That is how I would interpret it, but nothing is definite. A lot of smaller departments do some vetting to see if applicants are really interested, but they would only do that for someone they're interested in. I had a weird experience once where the offer didn't pan out after this, but I would suspect that your friend could go to Iowa if they want.

Yeah, this would probably be a good thing to crowdsource next year after the dust settles and it becomes clear which programs aren't adding them back.

I think I remember one applicant here getting into a lower-ranked program with a 158Q once, but I think it's going to be extraordinarily difficult to get into a top 50 PhD with a 156 unless the rest of your profile is extraordinary (elite school, exceptional good research experience).

I think it's too early to tell if any programs that were GRE-optional because of COVID will stay that way, but I can't think of any reputable programs that didn't require the GRE before this year.

Please stop mass-reporting frock's posts because you don't like their content. There are human moderators on this forum who don't want to spend all our time deleting accounts you don't like. If he posts something abusive or starts spamming, he will be banned; otherwise, ignore posts you disagree with and don't respond.

The first-year courses at these programs shouldn't be too theoretical. The really theoretical stuff will be more in the second year. For calculus, I'd make sure to know how to do basic integrals and differentiation, such as with exponential and log functions, basic multivariable stuff, and integration by parts. For linear algebra, the basics will be fine -- personally I'd watch the 3Blue1Brown videos because they're awesome. For probability, Harvard's Statistics 110 course is great, freely available online, and only slightly less advanced than a Casella-Berger class in my opinion. For analysis, you probably won't need much in the first year - basic notions of convergence and uniform convergence, how to do epsilon-delta proofs. An easier analysis book like Abbot would be fine. I'm not sure if those programs use R or SAS, but learning the basics of programming if you don't know them will also save you lots of time.

If you don't feel super strongly about a topic, I'd personally lean towards choosing a school based on location, ranking, environment, etc. One thing I would look at when choosing a department is the level of research and the journals they are publishing in and how that matches with your career goals. If you want to be a professor, you want to work with someone publishing in top stats journals. Some lower-ranked programs don't have many people doing this. Some people I know had a strong passion when they went into grad school (eg spatial statistics, or clinical trials, etc) and chose the advisor that they came to the school specifically for. This is a minority, in my experience. A lot of people don't have strong preferences. For instance, if you want a pretty "standard" job, like being a data scientist or working as a biostatistician at a medical center, it really doesn't matter that much what your specific dissertation was on. Even for an academic job, some people just choose a good professor they feel they will be productive with. And thus some people just sort of fall into their positions with their RAships, or based on taking a class with someone they like, etc. Some people don't have a strong passion for a specific topic, but choose a hot topic that may land them the type of job they want. If you want to be a researcher at Facebook or Google, studying network science/causal inference/deep learning might be a good idea. Or some people might think genetics sounds cool and they start doing research in genetics. Of course the topic you choose for your dissertation has some importance and you have to find something that is interesting enough to you that you enjoy it. But I recommend not stressing too much about this if you don't already *have* a strong preference. You'll never find the "perfect" research topic, and you will learn a lot by working on different topics and can always change directions during a post-doc or later in your career, too. I wish I had spent more time earlier in my career just jumping into research instead of stressing about what I'm going to work about in the future. But going to a department with a variety of options never hurts.

Yes, of course the additional coursework will help, but I don't think you are going to improve your profile by a *drastic* amount. For instance, if you didn't get into any top 30 programs, I certainly think it's unlikely that you'll be getting into a top 10 program in the next application cycle. You may be able to move from the 40s to the 30s or 20s, but I wouldn't expect magical results from taking a few more classes. We have somewhat similar profiles (good undergrad school but mediocre GPA, data scientist at FAANG, and I did submit a nearly perfect GRE score) -- and I didn't get into schools like Ohio State and Purdue after completing an MS with a 3.9 gpa from a top 20 program with all of the PhD coursework and multiple first-author publications. It may be that the department you were admitted to is an appropriate rank for someone with your background.

I think your math background probably has a lot more to do with your results than the lack of a GRE score., as the Berkeley professor told you. I agree with @Stat Assistant Professor that you are unlikely to vastly improve your profile in a year. If you do well at a good but not great PhD program, you can also get a post-doc after to improve your profile, which may be a better use of a year of your life than just waiting for the chance to be admitted to a better program.

I don't think the program is looked down upon at all, even on this forum. There have been multiple applicants over the years who have expressed that the program, at least in their admissions, has some organizational/communication problems. But I don't think anyone doubts that they have some very good people.

That's pretty amazing that your program admits people without having taken linear algebra - I've never heard of such a thing and it does sound sort of ridiculous to expect people to just catch up in that case. The first few years of graduate school can be overwhelming for a lot of people, as you feel there is just *so much* to learn. When I was getting my MS, I felt like I had to learn everything from the most basic level to fully understand it, and make sure I knew all the math that led up to it. But I think a perspective change might be helpful, in that you really only need to do well enough to pass the classes (and exams, which in some departments are awful - if you go to one of these departments, I'm sorry for your terrible luck. Kicking people out of school because of a single 3-hour test is terrible. Maybe you should transfer to a more welcoming program that doesn't weed out domestic students like this -- they exist!). I just returned to schools after 5 years off. I barely remember how to do integration by parts - but when a problem comes up with a weird integral I try it out. I don't know what the the power series for e looks like off the top of my head, but I know enough that when I see a weird series I can look it up and maybe see that it's in a nice form. I certainly don't know any properties of a trace of a matrix. I barely know how to take partial derivatives to do the Jacobian transformations of random variables in Casella Berger. And this time around, I'm not even going to try to trace back my math education to the beginning to re-learn every little skill that would lead me up to this point. Just learn to do the problems in the books, on your homeworks, and I'd highly recommend searching for lecture notes/stack exchange conversations that give clearer explanations than your professors might give.

Is the department admitting students who haven't taken calc 2 and linear algebra? I just can't imagine people don't actually have the background to succeed. You can get an entire undergraduate math education by watching YouTube videos for a couple days.

I think last year there were posts here about how disorganized UCLA's program was too.

I wouldn't worry about what previous alums do - even at top programs, a lot of people just go into data science roles because of the money. Most people who go to lower programs don't want to be academics, so most of this relationship is not causal. I'm not saying program reputation doesn't matter at all, but it matters much less than who you work with. There are plenty of great professors at many programs ranked 25-50. I understand if you don't want to post the program because of privacy concerns, but it is hard to give advice without knowing what your alternative is. If it's a school like OSU/UIUC/UF/UCD/UCLA/UT/UCI (not exhaustive list) that's in the top 50, I'd consider the offer very seriously.

How sure are you of those research interests and how passionate about them are you? Some people can be truly fulfilled by their research and if that'll make you happy, go to Berkeley. But you're not even going to be able to do good research if you're unhappy and wishing you were on the other side of the country. Are you sure that you are that much interested in probability than say, MCMC, where you could work with Xiao Li Meng at Harvard who to me is one of the most interesting people in statistics - just read some of his paper titles and listen to his talks. Are you sure that theoretical machine learning at Berkeley is that much more interesting to you than the reinforcement learning that Susan Murphy is doing? There's plenty of theoretical stuff going on at Harvard that might satisfy you intellectually, and I definitely think that location is extremely important. The facts are that you will be qualified for top stats jobs after working with someone good at Harvard. Maybe Berkeley will offer you a slightly better chance at doing the type of ML that gets a FB research job, but is that extra slight chance worth 5 years? My recommendation would be to download some papers from profs you like at both school. Read the papers from Berkeley and ask yourself if you love reading about that subject so much that you would move across the country to Berkeley to be able to ask the person who wrote it a couple questions every week.

Did they already admit you and you're just waiting about funding? If so, I think that would be totally reasonable. I think it's too early, except in extraordinary cases, for people who have not heard back at all to be emailing admissions. But just tell them you're excited about the offer and looking to see if there are updated on funding.

I was terrified of stuff like this when I was applying, but I very sincerely doubt anyone will care unless it's a core course like calc 3, linear algebra, or real analysis and even then, only if they told you your admission was conditional on those classes -- your admissions letter admits you on the condition that you graduate with your bachelors and that's likely all that matters. I never had anyone mention when I did stuff like this (failed or dropped classes multiple times during last semester of undergrad/masters).

Harvard has Murphy, Imai, Meng and I don't think NCSU has any senior people quite at that level.

UT is newer and smaller so yes it is under ranked. It is also in a location that many many people want to live in, unlike the midwestern schools which many people don't even apply to.

I definitely don't think it would be a step down. Everyone knows McGill is a good school, and though some people might not be familiar with UdeM because it's French only, the quality of their faculty is great. I strongly considered going to UdeM for my PhD, but it was pretty difficult to understand whether it would be possible for a US citizen to get funding equivalent to what we would get in the US, as most of the funding is reserved for Canadians. Yves Atchade went to UdeM and got faculty jobs at top departments Michigan and BU so it's definitely possible to have success from there in the US. Also, obviously your French has to be very good to go to UdeM. Also, not sure if this was a typo, but a 3.9 is a very high GPA. You go to a top 5 school and could just apply to PhD programs, including top US ones.

Do you want to live in Canada or the United States after your PhD? I think they are close enough in ranking where that should be a major consideration.

Are you mostly applying to statistics programs? Biostat programs get back early but often statistics admissions really start going out now at the earliest and til the end of February is really the peak period.

@bayesboi in previous years they have sent out a couple batches where it seemed like the committee would meet at the end of the week and send out a few more decisions. Chicago explicitly says on their website they review applications in groups by potential research interests, which may have something to do with it, but that is just speculation.