StatsG0d

I have enjoyed our Department a lot. This will be controversial, and will likely trigger some current students at other institutions on the forum, but I truly believe UNC is a top-3 biostatistics program. The culture between students is, in my opinion, our greatest strength. All the students are really sociable. We do our problem sets together and friend groups get formed so the current students hang out with one another outside the department. The only department that even comes close to how rigorous UNC is would be Washington, which I would argue (and I think many of our current students would agree) our program is modeled after. That said, we have students from a wide range of math exposure, and almost all students pass the qualifying exam at least by the 2nd time (and the vast majority pass it the first time). The instructors at UNC are very good. The class has printed notes that give the details on all proofs, so you can just sit in class and listen instead of trying to pedantically write notes while the lecture is going on. You may PM me with other questions if you'd like.

Adding onto what my good friend @Stat Postdoc Soon Faculty has said, I have found that being joint advised by one senior professor and one junior professor has been extremely beneficial. If this is possible, I definitely recommend it.

In that case, I would say follow your heart. It sounds like Brown would be a good fit for you, and Ivy League prestige never hurts.

UNC has (by far) the best biostatistics program of the 3, and an MS from there would probably "look better" for PhD programs later on, especially since UNC is renowned for its rigorous theoretical training. I would not put any weight on whether a master's exam is required. The material on a master's exam and a PhD qualifying exam are vastly different. You might want to ask these programs about the possibility of continuing onto the PhD.

This forum is really more oriented for statistics / biostatistics. For math, I suggest you head to https://mathematicsgre.com/

I would not worry about having a weaker stats background. Having the math background is more important. And having the bio background will certainly be helpful if you decide to do genetics, which is Michigan's specialty as you no doubt know. Best of luck!

IMO, it is not wise to choose a department based on one research area without having a backup. I have changed my interests probably at least twice since I have been in graduate school. I came in wanting to do ML, and was sure I did not want to do Bayesian statistics. Now, I am doing Bayesian statistics :). If you have some prior experience (no pun intended) in the research area you like at Brown, then go for it. However, to make a decision based on something you think you'll be interested in is dangerous. At Minnesota, UNC, and NCSU, there's a wide variety of reserach interests among faculty.

I was referring to the fast-track program. From what I have heard, the success rate is not as high as what they indicate. This is hearsay, but I have talked to several former students from the fast-track program who ended up switching departments or just leaving with a master's because they were not accepted to the PhD (I know at least 2). Either way, this is off-topic. I was merely mentioning this to the OP to help in their decision-making process.

Is your Michigan offer a direct admit or one of those conditional master's admits? I've heard some bad stories regarding the latter (lots of students reporting that Michigan says the success rate is ~90% but the actual success rate is ~60-70%). Berkeley would be impossible with 17k. You'd certainly have to take out loans. The COL-adjusted income between Michigan and UCLA are similar according to this adjustment: https://www.bestplaces.net/cost-of-living/ann-arbor-mi/los-angeles-ca/24000

I think a lot of biostatistics students pursue biostats departments because they like it better as opposed to some real constraint. But if you scour through some departments' web sites, you'll find a lot of statistics departments hire biostats trained students (e.g., Florida recently hired two biostatistics-trained PhDs). I would argue it depends more on your research area than your department. If you're doing ML research, you'll be able to get a postdoc at any ML lab provided you're publishing in good journals.

I also majored in economics, also worked for 2 years doing research assistance in economics, but went straight into biostatistics. I would encourage you to apply directly.

Speaking as a biostatistics graduate student, I would say there's really no such thing as a "typical" background. Some of my fellow students came straight from undergrad, some worked for a few years and came back, and yet still some are in their 30's and just starting out as first years. I have peers who have majored in math and stats but also biology and economics. The only common ground we all have is that we've taken Calc I-III, linear algebra, and (most of us anyway) real analysis.

Some programs will not accept you if they think that you wouldn't accept their offer because it will affect their yield.

IMO, you can do ML research at either statistics or CS departments, but there are some differences (others can weigh in if it's different in their experience). In stats, you might take a current ML algorithm or tweak one to prove that, under certain conditions, e.g., the classifier you developed is Fisher consistent or has some sort of nice asymptotic properties. In CS, the focus is more on developing new algorithms or finding new optimization techniques to make current algorithms faster, and they are much less concerned with any probabilistic properties of these algorithms. If you're only interested in ML, my advice would be to choose a CS department, or choose a stats department like CMU, which has ML training embedded in its curriculum. I have found that for the most part, stats programs tend to consider ML an elective, but the stuff you learn in stats departments is only tangentially related to the stuff you need to learn ML, so it'll probably seem like a huge waste of time to deeply learn (no pun intended) linear models, generalized linear models, and measure theory to simply end up doing ML.

I wouldn't waste time applying to Northwestern. The program is more difficult to get into than most other programs with the same prestige. I'll echo @Stat PhD Now Postdoc and say to target some of the larger departments. There seems to be an agreement among many people that domestic students tend to do better at public rather than private schools, likely for funding reasons. You might also want to consider biostatistics programs. Most departments' funding depends on NIH or other US governmental agencies, and this money can only be used to fund US students. I agree with both @statfan and @Stat PhD Now Postdoc... your cumulative GPA is a concern but on the other hand you have shown that you can handle yourself in advanced math classes. I'm at a top-10 institution, and your math background far exceeds what I had when I was applying. On the other hand, my cumulative, major, and math GPAs were higher than yours. It's difficult to know which of these things sends the signal adcoms are looking for as I strongly feel they differ from program to program. You could try to email some graduate program heads and tell them about your profile and your concerns and see what their response is. Worst case scenario is you don't get a response or get the generic "we take a holistic approach...", and best case scenario is they tell you what you can do to help. In my experience, many of them respond and are quite helpful with their responses. There are quite a few people from Utah in my current program. Perhaps you can reach out to people you know to see if they can get someone in their department to talk to you.

This is false. I do not know a single person in the program who did not choose their research adviser. This is true, but I fail to see how this is relevant now. If the admitted student received guaranteed funding, it has nothing to do with whether UNC guaranteed funding to other students or not. Again, this is false. While some students do end up working with their academic or GRA advisor, I would not say it's the norm. Most of the students I know do not end up working with their GRA or academic advisors for their dissertation advisor. Moreover, your GRA advisor may/probably will give you some collaborative work to do. It is true that it is ad-hoc and very informal. I simply sent out emails to some professors I thought about working with. All the professors whom I emailed responded positively. Ibrahim has a rule that you must be co-advised by a more junior faculty member. Not sure about Lin, but I never felt like he was that popular of an advisor. Kosorok does have many students, but my impression is that he is always willing to take on more students.

You probably don't want to spend 5+ years in West Lafayette, IN anyway. Too cold and not much going on.

I would focus more on statistics than biostatistics. The funding for international students in biostatistics is scarce.

Yeah, I think you should apply to more schools. Clearly you have a geographical preference, but I don't think it's worth it to not check out more places. You'd fall off a lot going from Stanford / Berkeley to Davis, and there's a lot of great programs in between.

If you're interested in environmental, I would say biostats is actually better. There's a lot of work being done in environmental health, which includes climate change. The STOR department as far as I know is almost completely concerned with statistical theory and in particular probability theory. There's a few people there working on ML (e.g. Liu), but there's at least as many working on ML at UNCs biostats department (e.g. Kosorok, Zeng). Biostats also has a long history in dimension reduction because for example genetic data infamously is a high dimensional problem. As far as sociology applications I'm not sure. I have had experience in both statistics and biostatistics departments, and I have found that the former doesn't care much for social justice research. If that's what you end up doing, that's great, but you could also just find some fun data set about, say, sports and apply it and that would be just as good to them. Since biostats programs are usually housed in public health schools, the applications have to be related to public health in some way. So the constraint isn't bio so much as public health. But if you want absolute freedom, maybe biostats isn't the best. Still, you won't have to do what you did for your PhD forever.

Yeah, I would apply to at least a couple biostat programs in the top 10 pooled. If you're concerned about rigor, Washington and to a lesser extent UNC have a strong theoretical focus.

Your list is pretty top-heavy. I don't think you'd have a shot at Stanford, Berkeley, Penn, Columbia, Duke, Washington, or JHU. The others are probably OK. As a side note, I went to an undergrad institution (barely) in the top-100 and had about the same grades as you, similar GRE scores, but with less extensive math training, and I got into several institutions in the 10-20 range for statistics and one school in the top-5 for biostats. As the others have mentioned, if you've gone to, say, a state school maybe in the top-70 USNWR rankings (including private schools), then you're probably fine. If you're closer to me, your letters of recommendation will likely make or break your application. I know for a fact that my writers wrote very strong letters for me, so I think that boosted my application dramatically. However, I agree with the others that you'll certainly want to broaden your range. I'd suggest the following breakdown (where the rankings should be interpreted as separate, not the pooled rankings in USNWR--see this thread for the separated rankings). Stats top-10, Biostats top-5: 1-2 programs (just so you can say you don't have any regrets) Stats 11-20, Biostats 6-9: 3 programs (target reaches) Stats 21-30, Biostats 10-14: 3-5 programs (targets) Stats 31-50: 1-2 programs ("safeties") This means sending out 8-12 applications, the costs of which can add up quickly. If this seems too financially burdensome, PM me and I can try to give more specific recommendations. Also, I don't think you need to retake the GRE unless you REALLY want to shoot for those top programs, which will be very difficult given your math background even with perfect scores. Perhaps if you really impress on the subject test, your profile could change dramatically.

Research experience isn't really that valued in admissions unless it's really high profile or very theoretical. I agree with @bayessays. You need that Analysis grade. Get a good grade in Analysis, and you open up a lot of doors.

Don't really have anything to add, just want to second this.

Like @Applied Math to Stat said, I think you have a really good shot at all those schools. Purdue also lets you put down in a form any mathematics courses you've taken. I think you're pretty much set for most schools in the 10-20 range. I definitely think you have a very good chance at the top-10.