Stat Assistant Professor

As a domestic student, if you wish to apply to Statistics PhD programs, I would target schools mainly in the range of 40-80 of the USNWR Best Statistics Graduate Programs. You have a few too many B's, including one in Real Analysis and one in undergrad probability, so it probably doesn't make much sense to aim a lot higher. If you are interested in Biostatistics, it's possible you could aim a bit higher, though even then, I wouldn't apply to many above the rank of 40 overall in USNWR.

OP: Unfortunately, given that you are an international student and your undergrad GPA is on the low side, you will have to aim a lot lower for PhD admissions. Unless the lower GPA is due to poor performance in the first two years but then you managed to get straight A's in your third and fourth years, then it's going to be very tough to break into most of the top 50 Statistics PhD programs. For the record, the list you should be looking at is this one: https://www.usnews.com/best-graduate-schools/top-science-schools/statistics-rankings I would say that all of your target/reach schools are unrealistic, and your list of safety schools is mostly unrealistic as well (not sure about UCR or UCSC, but the other ones on there would be difficult for you to get into). Even schools like UFlorida will be admitting mostly international students from the top universities in their respective home countries (or sometimes if they graduated from a top university in the U.S.A.). If you want a better sense of schools that you could shoot for, you should post the mathematics courses you have taken and the grades that you earned in them. Even without this info, I would say your best shot of getting admitted to a PhD program is to aim for schools in the 60-90 range of USNWR like University of South Carolina, Kansas State University, Virginia Commonwealth University, University of Missouri.

To get into a Biostat PhD program ranked below the top ~10 Biostat programs, the math requirements are lighter than for Statistics PhD programs. In some cases, you only need Calculus I-III and Linear Algebra. Even for Biostat, having real analysis helps though, as do a few other upper division stat/math classes (e.g. the undergrad-level probability and statistics sequence). The more elite Biostat programs might favor deeper math backgrounds. As for "expiration date," I don't think there is such a thing. There are people with PhDs in Statistics who began their studies a decade (or more) after graduating with their Bachelor's. It doesn't really matter when you took the prerequisites as long you have them. However, if you have been out of school for that long, you would have forgotten a lot of details from college-level math and would need to spend a bit of time reviewing Calculus and Linear Algebra before starting the PhD.

What was your GPA in your undergrad math classes, e.g. real analysis? I'll assume that they were decent. With a 3.8 GPA in Math from an MS program from a regional comprehensive, some research experience, and strong letters of recommendation, it is very possible to get into some Statistics PhD programs ranked roughly 40-80 in USNWR (I'd actually be confident in that) without doing a second Masters degree. I would recommend just applying to Statistics PhD programs directly if you're very inclined. The "reach" schools for your profile would likely be schools like University of Minnesota, Texas A&M, Purdue, etc. I know of some folks who have degrees from directional and regional comprehensive universities (e.g. Central Michigan U., Southern Illinois U., etc.), as well as obscure liberal arts colleges who have gotten into schools of that caliber (though I would still consider TAMU, UMN, Purdue "reaches" and would recommend you focus attention on schools in the range of 40-80 of USNWR rankings). *ETA: This is assuming that the OP is a domestic applicant. For domestic applicants to Statistics PhD programs, many programs below the most "elite" schools seem to be okay with admitting students from regional comprehensives. This is not the case for international students, who would have to aim a lot lower for PhD (USNWR 60 and lower) or obtain a Masters degree in Stat/Math at a more "reputable" school first in order to be competitive.

You are right that if you want to have a chance of getting admitted into a PhD program, you need to primarily aim lower (as in the 60-100 range of the USNWR rankings for Statistics). The other option if you want to aim a bit higher is to do a Masters in Statistics where you retake Real Analysis and get an A in that and show that you can get A's in Casella & Berger Masters level mathematical statistics classes. Then you might be able to aim as high as 30 or so of USNWR, though I would not recommend applying to many schools above that (you should still aim for mainly schools in 40-80 or so). With a strong Masters performance, schools like UConn, UF, FSU, UIUC would not be totally out of the question.

It would help to know what your grades in your math/stat/related classes were and who you can ask for letters of recommendation (plus a general feel for how strong you anticipate they would be). That might give a clearer idea of what schools you should apply to. That said, your overall GPA is probably a bit on the lower end for the very top-tier programs -- unless UVA is known for grade deflation and I'm just not aware of that. But let us know your math grades and your intended rec letters, and this will give a better sense of what schools you should target.

As long as you can secure strong letters of recommendation from your online professors, it should be okay. In order for them to say more than just "so-and-so was a good student who got an A in my class," you may need to interact with them somewhat more. As long as you can make sure you get sufficient interaction with them and get to know them well enough for them to write you a strong letter, I don't see any problem with online classes. (Fact is, as a result of the pandemic, students taking upper division classes completely remotely has been the norm rather than the exception).

Community college courses might be acceptable for obtaining the prerequisites, but not necessarily for getting the best letters of recommendation. For PhD admissions, it's better if your letters come from people who hold PhDs that can speak to your "research potential." You'd have a greater chance of getting such letters from professors at 4-year colleges/universities. That said, for PhD admissions, you probably shouldn't ask instructors of Calculus I-III or Linear Algebra for LOR's anyway (Masters admissions are a different story), unless you took more upper division classes with the same instructors. Since you don't have any of the math prereqs, you could take Calc I-III and Linear Algebra at community college. That would be enough for you to get into many Masters programs in Statistics. But then in order to be competitive for PhD programs in Statistics, you need at least a few upper division math classes as well, and it would be better to take those upper division classes at a 4-year school.

Oh, I didn't realize until now that one of your B's was in Stats Theory. That might give some adcoms some pause. I would think that a B in Abstract Algebra would be less frowned upon as it is not that relevant to Statistics (but that is me projecting), but a B in Statistical Theory might be somewhat concerning. On the other hand, if this was during the pandemic, the adcoms may be more forgiving.. you never know. I second @dontoverfit's advice to take two other proof-based classes (upper division proof-based linear algebra and real analysis). If you get A/A-'s in them, that will help your application a lot. I think Columbia, Yale, and Harvard may be far reaches for your profile. These are actually quite competitive to get into (as are all the Ivy League schools), plus they don't accept that many domestic applicants, and they seem to strongly prefer people with deep math backgrounds. I personally am not sure why Yale is ranked in the 30s in USNWR, since they have placed PhD graduates in TT positions at UPenn Wharton, Princeton OFRE, Columbia Stats, UChicago Stats, and just this past year, Cornell S&DS. These academic placements are outstanding. Instead of applying to all three of Columbia, Yale, and Harvard, I would recommend only applying to one of these three schools and substituting in Carnegie Mellon, University of Washiington, or University of Michigan for the other two. Then add a few more schools like Texas A&M, Iowa State, and maybe a few extra schools in the 40-50 range (like Colorado State, UConn, etc.), in addition to schools like UIUC, UF.

Agreed with the above. I think the 'sweet spot' for the OP would probably be schools like UW-Madison, NCSU, and UMinnesota. I would target mainly schools in that general tier, and in addition, apply to a couple of schools ranked above that (e.g. it wouldn't be out of the question to try Duke, Michigan, or Carnegie Mellon) and a handful of schools below that (e.g. in the UF/UIUC tier).

@AliasName Since you already completed one year in your current program and you are presumably fully funded as a PhD student, I am not sure how much sense it makes to jump ship and reapply to (most likely) unfunded Masters programs. Unless you can get your first-year credits in your present program to transfer over to a new program, you would need to repeat the first year sequence in another Biostatistics Masters program anyway. And even one year of paying out-of-pocket for a Masters seems like a poor financial choice when you are already currently funded. If you are really certain than you do not want to do a PhD, are you able to leave with a Masters if you were to complete a second year in your current program? I would be surprised if this is not an option. FWIW, I do know people from my own PhD program who received PhD passes on their qualifying exams (so they could have remained in the PhD program), but they left with a Masters due to lack of interest. I would explore options of finishing with a Masters in your current program, since: 1) you're fully funded, and 2) if you are already in a PhD program and you haven't been dismissed, it's often a better idea to leave it with a Masters if you do decide not to finish. If that is an option, you could consider taking mostly elective classes that are more relevant to your interests rather than required PhD theoretical classes.

Your profile looks pretty strong, and I think you should be very much in the discussion for schools like University of Washington, Duke, and University of Michigan -- provided you can get strong letters of recommendation that can say a bit more than that you are a strong student (i.e. that you have strong research potential). I am not sure if a letter of recommendation from an industry manager will be helpful unless this manager has a PhD in Statistics, Math or a related field and can speak to your research potential. If this is not the case, then I would recommend going with a third professor for your third LOR instead of your industry manager. As long as two of of your letter writers can speak to your research potential, a third "acceptable" one that mainly emphasizes that you are a good math student should be sufficient. I would recommend applying to a bunch of schools in the range of 16-30 in USNWR (e.g. Minnesota, Texas A&M, North Carolina State, University of Wisconsin-Madison... I think you would be able to get into several of those) and then applying to maybe an additional 4-5 schools ranked above those. Schools like Stanford, UChicago, and the Ivies may be tough for you to get into, since you don't have academic research experience or graduate-level coursework (these schools seem to prefer applicants who have done some graduate-level coursework in math/stat and who have already taken the equivalent of the Casella & Berger sequence of Masters-level mathematical statistics). But you might get lucky, and I could see a school like UW, Duke, or Michigan choosing to admit you. At UMich and Washington, the first-year PhD students start out with Casella & Berger.

If your university has a decent track record of getting Masters students admitted to the Statistics PhD programs at OSU, Minnesota, and Florida, then those seem like good places to try. If you are going to continue doing the Masters at your current university, then I would probably not aim much higher than UMN if your school does not have a track record of sending its graduates to those higher ranked schools. For international students, the top 20 Stat PhD programs seem to be heavily skewed towards applicants from a few select Asian universities. You could try your luck at a small number of them if you want, though. Having more math classes is viewed more positively than having more statistics classes. Research experience does seem to be viewed positively, and many international applicants for Statistics PhD programs in the U.S. seem to have it nowadays. Will your research in a neuro lab eventually lead to publications down the road? That would be something to mention in your application, or have one of your LOR writers mention.

Which math classes have you taken? It seems as though the Economics major is more mathematical/rigorous in other countries than in the U.S., so if you've taken the requisite math classes (at the minimum Calculus I-III and Linear Algebra), then I think your chances should be good for your list of schools. UCL is a very prestigious school and your GPA looks solid.

Your profile looks strong, and I think your math background is more than sufficient for Statistics PhD programs in the USA. The bad news is that if your school is ranked 80-90 in the QS Asia University Rankings 2021, then it will likely be very tough for you to get into any of the top 50 or so Statistics PhD programs (without a Masters). For these schools, there are already a number of qualified applicants from top schools in Asia like ISI, IIT, Peking, Tsinghua, Fudan, USTC, University of Hong Kong, NUS, SNU, Yonsei, Sharif, etc., and so programs can afford to be picky about which international applicants they admit. I am not surprised that your classmates were unable to gain admission to Penn State, Iowa State, or NCSU, because there are already so many applicants from more reputable schools in Asia. I know it sucks, but this is the reality. Some programs likely auto-reject international applicants if the admissions committees are not familiar with their undergraduate institution (this is not the case for domestic U.S. applicants where doing very well at an unknown undergrad school can still get you into a Statistics PhD program like NCSU, Iowa State, Penn State, TAMU, etc. -- occasionally a school like Duke). The lower you go down in the USNWR rankings, the more receptive the programs will be to accepting international students like yourself. I'm thinking schools like University of Missouri, Southern Methodist University, University of South Carolina, UMBC. Those would likely be good targets. However, if you want to aim higher, your best bet would be to go to a top school in Asia (or a decent program U.S.) for your Masters degree and to do well there. I've seen this before -- international applicants who went to less prestigious undergrads but then went to either a top program in their home country (like ISI) or a reputable Masters in the U.S. (like Rutgers), and then they were able to get admitted to Statistics PhD programs in the range of University of Minnesota to University of Florida. With a Masters from a top school in your home country or a decent one in the U.S.A., you could probably aim for schools in that general ballpark.

Since you have taken Calculus I-III and Linear Algebra and finished with grades of B- or higher, I would think that you'd be able to get into the Masters programs at your 'match' schools and even some of your reach ones. Your GPA isn't terrible -- not superb, but not low enough to raise any red flags. I think applying to your current list of 10 schools might be sufficient.

If you are in a Masters program and you don't have any prior foundation, then it will probably be tough to do very deep statistics research while taking three classes a semester. It can definitely take awhile to become accustomed to academic jargon, and it will be difficult to juggle trying to do that that while still taking courses. I personally found that even after having taken the graduate-level statistics courses, it was initially very difficult to read/understand statistics papers. It took considerable effort and patience to be able to read academic papers, even with almost full-time effort. But eventually, if you work at it enough, it just sinks in. Your PhD advisor will definitely help you out. As for developing a 'working' understanding of different research areas, it would definitely be helpful to take some Statistics electives like Bayesian statistics, spatial statistics, or high-dimensional data analysis (as these are often useful introductions to specific areas), and to attend departmental seminars. You probably won't understand everything in department research seminars, but you can hopefully pick up bits and pieces here and there and get a flavor about different research areas. If you wanted to do some 'light' research, you could get involved in some interdisciplinary research or applied research where you write some code and/or perform some statistical analysis. That would be seen as a definite plus for PhD applications, though it wouldn't carry as much weight as letters of recommendation and grades. For PhD admissions, I think it is more important to have excellent grades and strong letters of recommendation which can attest to your research potential than it is to have "pure" statistics research experience. I would probably focus more on doing well in your classes and getting good letters above all else. Why study statistics? I suppose the motivation is different for different people. Some students want to get into certain fields (e.g. data science, quantitative finance, etc.), so they pursue advanced degrees in Statistics to help them achieve their goals in industry. Others are motivated by wanting to study specific applications (e.g. environment, demography, public health, etc.), while others are primarily interested in the mathematical foundations of probability/statistics. For me, I personally enjoyed the statistics classes that I took during my Masters program (in Applied Math), especially a class I that I took on Bayesian statistics. After my Masters, I took an industry job but found that I preferred the academic environment, so I pursued a PhD two years later. Over the years, my motivations for continuing to study statistics have also changed a bit. I started out as more of a mathematical statistician during my PhD program (i.e. studying statistics theory rather than applications). But after doing a postdoc, I became a more applied statistician who still does some theory, but it's not always the primary focus of my work. I still enjoy the mathematical challenges of statistics theory, but I find that to stay motivated, it is often useful to have motivating applications for your work -- that is, being motivated by wanting to address real-world scientific problems and wanting to make sense of real data sets (which are often messy and challenging to make sense of!).

OP: You should also check out University of Washington. The PhD program in Statistics there has an optional track for Statistics in the Social Sciences. I know somebody who graduated from UW Statistics with a concentration in Statistics in the Social Sciences, and they ended up as a prof at an R1 school. FWIW, their research was on respondent-driven sampling and social network analysis, and they work a lot on applications specifically to health and public policy.

Then I think your chances for PhD programs in pure math are fairly low. You would have to get a Masters where you take courses like real analysis, abstract algebra, complex analysis, etc. and get all A's to demonstrate to admissions committees that you can succeed in a pure mathematics PhD program. Without having taken *any* of these classes, you might not be able to pass the PhD preliminary exam anyway. Scoring very well on the Math Subject GRE would also help your case a lot, especially since your undergrad was not math. Even if you were to do all these things, I would guess that you still have to apply to mainly mid-to-lower ranked PhD programs, on account of your undergrad GPA (although your publications in a journal and a conference may be definitely seen as a plus, since it demonstrates some research ability). I agree with you that it would be a very long road ahead of you if you wanted to pursue a PhD in pure math. But if that is your passion, then I say go for it. But you need to be realistic about what schools you can get admitted to. The folks over at the MathematicsGRE forum might be able to help you out more. Best of luck.

Unfortunately, I am not able to gauge your chances for math departments -- you could check with people on the mathematicsgre forum for some suggestions. I think your chances are above average for some of the Statistics departments I mentioned (UNC-Chapel Hill, Michigan State University, University of Minnesota, University of Wisconsin), and you could find a PhD supervisor who is more on the probability side of things. The very top Statistics programs also have a lot of probability theory faculty (Stanford, University of Chicago, and UC Berkeley come to mind), but unless you're ranked in the top 10 or so students in your class, these schools might be hard for you to get into.

You have a strong pedigree, so your application should definitely be 'in the discussion' at a lot of schools. Your undergrad performance looks very strong, and your Masters performance at ISI looks quite strong as well. I have seen students with scores of high 80's from ISI graduate "with distinction" and get accepted to top PhD programs in the U.S.A., so it seems like anything over an 80 is considered very good from ISI. I think you should have a shot at schools at least at the level of University of Minnesota and University of Wisconsin (in particular, you are interested in Markov chain Monte Carlo theory -- which tends to be a bit more probability theory-focused than statistics, then there are some good faculty at UMN for this... UWisc also have Timo Seppalainen working on probability theory who can supervise Stats PhD students). The STOR program at University of North Carolina-Chapel Hill also has some good probability theory/stochastic processes faculty. I would definitely recommend applying to UMN, UWisc, and UNC-CH. A lower ranked Statistics program that has some great probability faculty is Michigan State University. You could also apply to UC Berkeley as a (far) "reach" school. In general, though, Statistics PhD programs outside of top programs (Stanford, Berkeley, etc.) won't have as many pure probability faculty. Mathematics departments tend to have more probability faculty. So that is also something for you to consider if you are very sure that you want to be a probabilitist.

I agree that the OP has a very good chance at top 10 PhD Biostatistics programs. Since you have the necessary math background, you would certainly be able to succeed in a Biostatistics program. If you have interest in Epidemiology applications (e.g. health disparities), I will point out Biostats tends to value collaborative research. So you can definitely collaborate with epidemiologists and get a couple of collaborative publications that way. Agreed with @bayessays to look into UPenn Biostatistics. Biostat, Epidemiology, and Medical Informatics are all in the same department, and you can collaborate with people in epi even if you're doing the PhD in Biostatistics.

You may be able to get better advise on MathematicsGRE.com, where you'll find a lot more pure math PhD applicants. However, I do have a few comments. You say that you are interested in pure mathematics, but have you actually taken classes in pure math? Pure math PhD programs will want to see that you have taken courses in abstract algebra, real analysis, proof-based linear algebra, complex analysis, number theory, topology, etc. If you do not have these classes, then you might be totally out of luck, and the immediate course of action is to somehow take a few of these classes (at the MINIMUM, abstract algebra, real analysis, and proof-based linear algebra) OR obtain a Masters degree in math where you take these classes. A lot of Math PhD programs have a preliminary exam that all first year PhD students have to take that tests these subjects at the undergrad level (at least on proof-based linear algebra and real analysis). PhD students who cannot pass this exam at the PhD level are dismissed from the program. So you'd actually need to be really solid in these subjects otherwise you won't even be able to reach the research phase of your PhD program. Scoring decently on the Math Subject GRE is also quite important for getting into a decent program in Pure Math. If you can score well on this exam, that might assuage some adcoms' reservations about your low undergrad GPA. However, a good score on the Subject GRE is no substitute for coursework. You need to demonstrate strong abilities in math through getting good grades in the relevant classes.

Your profile looks very strong. The top students from the top schools in Canada (McGill, Toronto, UBC, McMaster) seem to do very well in admissions to American Statistics PhD programs -- including international Asian students. And your profile looks great. I think you have a shot at all the schools you listed, but I would maybe add a couple more schools, so you have about 10 applications. The competition at the top can be stiff, so you want to make sure you don't get shut out. I would recommend applying to UPenn Wharton and University of Michigan, as these schools have some very good people working on deep learning and functional data (e.g. T. Tony Cai, Weijie Su). And I think you have a solid chance at getting into both UPenn Wharton and UMich. Adding two "safer" options like North Carolina State University and University of Wisconsin Madison may be a good idea as well.

Agreed with @StatsG0d. It would be very beneficial to take upper division, proof-based mathematics courses, especially Real Analysis and maybe upper division proof-based linear algebra. The first two years of coursework in any Statistics PhD program will be quite theoretical, and you will be required to take classes like measure-theoretic probability, linear models, advanced statistical inference, large sample theory, etc. At most Statistics PhD programs, it is possible to do a more "applied" dissertation with little/no theory -- this is true even at heavily theoretical programs like UPenn Wharton or Stanford (usually there will be at least a few faculty members who are more applied and don't really do theory, and you can ask them to supervise you). But to even get to the research stage, you need to get through some pretty theoretical classes and pass qualifying exams on some of this material. The qualifying exams are also challenging, and there are always a few students who do not get a PhD pass on this exam and have to retake it or leave with a Masters. It doesn't really matter what the rank of the program is; the vast majority of ranked Statistics PhD program will be this way. More advanced math classes -- *not* more undergrad statistics classes -- would be the best preparation for a Statistics PhD program.