Stat Assistant Professor

I saw that UCLA Department of Statistics hired an Assistant Professor recently who got his PhD from UT Austin SDS. His PhD advisor was James Scott, I believe. I do not believe there should be a huge disadvantage in the academic job market coming from UT Austin vs. UNC-CH if your publication record and your letters of recommendation are solid -- those are the main thrusts for academic hiring at research universities. Nowadays it also really helps to do a postdoc to make your CV more competitive.

I would talk with the Chair of the graduate admissions committee asking how likely it is for you to get funding after your fourth year. It might be the case that you are very, very likely to get funding for a fifth year, even if not offered it upfront. When I got a fellowship for my graduate program, the length of this fellowship was also only four years. But the Graduate coordinator told me that they would fund me through TA in my fifth (or even sixth) year if it took me longer than four years to finish. I can't imagine that Statistics courses would be short on demand for TA's -- especially the big intro class that thousands of students on campus have to take.

This probably depends on who the postdoc PI is and how well your research jives with theirs. For example, if your research is on causal inference from a Biostat department, you might be able to get a postdoc with someone in a Statistics department who works on causal inference and who is a bit more on the applied side of statistics (e.g. Fan Li at Duke). If you worked on statistical genetics in a Biostat program, you could potentially get a postdoc in Nancy Zhang's lab at UPenn Wharton Statistics. However, unless you've been publishing in top ML conferences and top methodological/theoretical statistics journals, you (probably) can't get a postdoc with say, Michael Jordan or Martin Wainwright of UC Berkeley. It really depends. In general, it seems to be a bit more difficult to switch from Biostatistics to Statistics than the other way around, with a few exceptions (e.g. if you got a Biostatistics PhD from Harvard, JHU, UW, or UNC but you happened to work on some statistical theory for your dissertation -- these schools all have biostatistics faculty who are more theoretical and who serve on the editorial boards of Annals of Statistics, JASA Theory & Methods, etc.). But another reason that it is easier to switch from Stat to Biostat is also due to the fact that there are many more postdocs in Biostat than Stat.

I do not see that you have taken Real Analysis, which is basically a requirement for PhD programs in both Statistics and Biostatistics. And your math background is rather light, especially for an international applicant. Finally, your list was way too top-heavy. For a lot of the schools you listed, you would be competing against a lot of very strong candidates from ISI, Peking, Tsinghua, etc. who have much more math than you. I think you could improve your chances if you enroll in a Statistics or Mathematics Masters program and take two semesters of real analysis and maybe one or two other advanced math classes. Then I would consider applying to schools ranked outside the top 30. I unfortunately do not think that even with a Masters that you would be as competitive as say, a math major who is fresh out of college from a top school in China and who finished near the top of their class.

Attending UC Denver wouldn't put you at a disadvantage (for academia) as long as you have a good PhD advisor who can help you get published and you publish good papers during your studies. Actually, when I was on the academic job market this year, I saw that other job candidates were interviewing on-campus at R1 universities who had PhDs from schools like University of Cincinnati and University of Illinois at Chicago. These people tended to have prestigious postdocs and at least one paper in a top journal, though. I won't deny that the program rankings are correlated with certain things (for example, greater number of professors who can supervise you and who can help you get good postdocs through their networks). But that doesn't mean that the *only* strong academic job candidates are from "the top" schools. An impressive CV is an impressive CV, whether it's from somebody at Stanford or UCD, and I think a lot of hiring committees realize that (in Statistics/Biostatistics anyway).

If there is anything that I learned from being in academia, it is that there is quite a bit of randomness in the whole process. Just as with PhD admissions, if you apply to faculty jobs, you will find that a lot of people with *incredible* CV's are routinely being passed over as well, even at mid-tier institutions. These other schools you mentioned also get tons of applications from very qualified applicants (particularly international ones) who also have legitimate research experience (e.g. http://xumaoran.com/). I wouldn't worry about it too much. Just celebrate your acceptances to amazing schools.

I do know several Statistics PhD graduates who now work as Research Scientists at Facebook, Google Brain, and Amazon (some after spending a summer being a Research Scientist-Machine Learning intern). I don't really know what they do as far as "research" goes and/or if that is any different from "regular" data scientists, though. If you want to do *really* basic research like computing theory or mathematical foundations of statistical learning outside of academia, the opportunities will indeed be very limited (e.g. Microsoft Research, maybe some national labs). Anyway, I would figure out what your priorities in your work/non-work life are and go from there. Even I decided that I was not going to be a postdoc for more than three years and that I would go back to industry if I couldn't find an academic job within 3 years of finishing my PhD (fortunately I found one during the second year of my postdoc). If I had to leave academia, though, I would have definitely missed academic research but I think I would have been fine too -- if I didn't find work sufficiently satisfying/stimulating, I would have redirected that energy into making my non-work life satisfying (maybe even by taking some MOOC's to keep my brain active, likes bayessays suggested).

I would ask the chair of the admissions committee (not individual professors) about funding opportunities, including what possible research assistantships there are. I know some people with Masters in Biostatistics who worked as research assistants to partly fund their studies -- this seems to mainly involves doing some data analysis and programming in R or SAS. Although, I will say, if you got a fully funded offer from University of Wisconsin, I would be inclined to take that. That is an excellent school and should provide ample opportunities for industry jobs -- unless you are really, really interested in getting a PhD and one of the other schools has an easy "Masters-to-PhD" track for MS students.

OP: Since you are already working in the field of data science and you do not seem to have much interest in pursuing a research career, I am not sure that a PhD would really give you much (other than the intellectual stimulation, I suppose, if you are independently wealthy and can afford to take a 4-6 year "break" from industry). Certainly, there are many Statistics PhD graduates who work as data scientists or in non-academic careers, but for the most part, they were not already working in data science prior to starting the PhD -- and a lot of them were international students for whom getting a PhD in STEM actually makes sense if they want to get a U.S. work visa. I would carefully evaluate your reasons for wanting a PhD and whether it is worth the opportunity cost. I wouldn't do it unless you really have a strong drive to create new knowledge and do original research. Even then, this motivation may not be sufficient once you weigh it against your ultimate career goals, opportunity costs, and the sheer difficulty of the whole ordeal. You seem intellectually curious, so I would do as the posters above suggested and take free online classes or study topics in your own free time if you want to learn. But unless you have a drive to create new knowledge (however incremental), it may not be worth your time.

You should go for it, if that's what you really want. A ~3.9 GPA from an Ivy (which most people wouldn't consider "low ranking"), with a minor in math and A/A-'s in abstract algebra and real analysis, are certainly nothing to sneeze at and should put you in great shape to admitted to a top Biostatistics PhD program. I could see you being admitted to UNC and UMich, and it is worth trying UW, JHU, and Harvard as well.

Does the website say what the requirements are for admission to their Masters program? I would think that as long as you meet the minimum requirements in terms of GPA and coursework taken (typically, Calc I-III and Linear Algebra for Masters programs) and as long as you have a decent score on the Quantitative section of the General GRE (~160 or higher), you should have no difficulty being admitted. Most Masters programs are not funded, so the vast majority of them will accept any students that are above threshold.

Assuming all else is roughly equal (e.g. institutional reputation, number of potential PhD advisors, venues where the faculty are publishing, location), I suppose you could delve a bit deeper and look at course requirements and offerings. Some schools have very intensive coursework requirements, including two semesters of measure-theoretic probability and a lot of classical statistics theory, plus two written qualifying exams, while others have fewer coursework requirements and only one written qualifying exam. But like @bayessays said, I don't think you need to worry about the coursework not being rigorous at any ranked PhD program, regardless of whether it is ranked #1 or #90. For the most part, these courses are based on the same or very similar texts (e.g. Casella & Berger for Masters level stat theory, Durrett or Billingsley for advanced probability theory, etc.). As to whether the courses are important: I think they can provide a solid foundation, but their main purpose is to prepare you to (hopefully) pass qualifying exams rather than research. Once you reach the research stage of the program (and if you continue on in academia), you really are teaching yourself for the most part- -- even the PhD advisor isn't typically going to hand-hold you or teach you the subject of your dissertation, but rather guide you at a high level. So that is why I would recommend prioritizing things like potential faculty advisors for your thesis, average PhD completion times, and job placements when selecting a program to attend (especially if you are interested in academia).

Perhaps the most "tactful" way of going about trying to secure more funding is to reiterate one's interest in the program but ask if it is possible for one to be considered for any additional fellowship awards. This seems non-confrontational and non-controversial enough. Plus, I would just leave it brief -- I'm not sure that listing a bunch of reasons why or discussing competing offers would be very compelling, because: a) cost of living varies so widely in different areas and because issues like rent eating up most of the stipend, needing to accommodate a partner, etc. are not unique issues faced by just one student, and b) PhD programs can probably just recruit another exceptional student who has a similar profile from their waiting list (or even fill their desired slots with the others offers they've made) if a student is going to be like, "I would like x, y, and z for me to go to your school." What's that adage? "All men are replaceable."

Good to know! I stand corrected. Please disregard my earlier comment that said this was unlikely.

+1 to this. Also, if you look at UC Berkeley's Statistics PhD requirements (a very top program, if I do say so myself), it seems like students there get a choice of taking two out of three sequences: theoretical statistics, applied statistics, and probability theory. So it seems as though one could actually go through UCB's Statistics PhD program without having learned even measure-theoretic probability theory (which, while good to know, probably isn't super-relevant to everyone's research --if you're not a probabilitist, you can probably learn enough of it to get by without having taken a whole course on it). And UCB also has no written qualifying exams. Doing original research truly is the primary focus of the PhD. For your PhD research, you are mainly teaching yourself the things needed for your research. One cannot reasonably expect to learn *everything* there is to know through classes anyway. I kept a lot of my notes from my classes but I barely look at them, because I can just Google what I need to know if it comes up (e.g. what the expectation of a quadratic form is, various norm inequalities, etc.). Plus, I will say that although my research focuses on high-dimensional Bayesian statistics, it was easy enough for me to pick up on the analogous frequentist methods/theory (like LASSO, elastic net, etc.) once I had enough experience reading and understanding academic papers. I'm sure that will be similarly the case for you.

Very interesting -- this is very useful information! I haven't personally seen it done in Statistics before (negotiating for a higher stipend), but maybe it is possible. In my own case, I and several of my cohort received fellowships that were funded by the graduate school rather than the department (and this fellowship was more money than the TAship). But I didn't specifically negotiate with the Graduate Coordinator or ask to be considered for the fellowship. The department simply nominated me for the graduate school fellowship once I had accepted their offer. I would be interested in hearing if there are any success stories of negotiating a higher stipend from the department in Statistics. Or if it's more like, while the department(s) couldn't offer a higher stipend, the student was able to to successfully secure some fellowships that offered more money than the TAship would have.

I have never heard of prospective grad students being able to negotiate a higher stipend, in any field. In any case, this is almost certainly not possible in Statistics. As @bayessays pointed out, the stipend is determined by the department and pretty much all students get paid the same (except for a few students who may be on a fellowship or some special scholarship, but these students do not get to negotiate amount for those either).

Having been on both the primary (TT) and secondary (postdoc/VAP) academic job markets twice now, I will reiterate that if there is any chance that you are considering an academic job, you should definitely prioritize quality of life for your PhD, because there won't be as many opportunities to be geographically mobile afterwards. The academic job market in Statistics/Biostatistics is crowded and competitive enough these days that many PhD holders/postdocs from the Ivy League, Stanford, Berkeley, etc. are taking jobs at schools whose programs are ranked much lower than their alma mater/postdoc institution. If you can publish extremely prolifically during your PhD and postdoc *in the top journals* (a long publication list without -any- in top venues typically won't cut it either), you can certainly improve your chances of landing at a "dream" school in your most desired location. But that isn't guaranteed. I decided that the academic lifestyle suits me and that the job security and relative freedom in day-to-day job duties were worth sacrificing the geographical flexibility and taking a slightly lower salary than in industry. But I could see that this is (understandably) not the case for everyone.

Agreed with the above that Brown Biostatistics is a great department. If you are interested in industry, Brown University's proximity to Boston is also a plus. Boston have something like a thousand biotech companies, ranging from start-ups to huge pharmaceutical companies. Brown is certainly good enough to land you academic jobs, provided you have a record of achievement. For academic jobs, the CV (namely, the publications) and the recommendation letters are truly the most important factor for landing interviews/jobs. This past application cycle, I saw some folks with PhDs from schools like University of Illinois at Chicago and University of Cincinnati (which are great schools but not "elite") getting interviews for TT jobs at R1's because they had strong CV's with a at least one or two publications in top journals. A strong pedigree is certainly helpful -- I won't deny the correlation between publication record/strong recommendations and institutional reputation. But it's not the *only* thing that matters.

It would be a good idea to ask that during your visit days... though if there are any lurkers on this forum who currently attend those programs, maybe they can answer.

Congratulations on your acceptances! Those are some fine places. I cannot comment so much on the cultures of the departments, but for what it's worth: Columbia has David Blei who works on Bayesian nonparametrics and computational statistics (mainly from an optimization-based approach, e.g. variational inference). His students and postdocs are quite successful at securing jobs in both academia and industry (e.g. Google, Apple, Deep Mind). The same is true of Michael Jordan at UC Berkeley who was Blei's PhD advisor. Students and postdocs of Jordan do extremely well in the job market. Harvard strikes me as a big MCMC department. It seems to me that in the machine learning community (rather than the "pure statistics" community), variational inference methods are currently of greater interest than MCMC. But there are also people working on scalable MCMC so that it can be more attractive to "big data" practitioners (possibly some at Harvard). I know that at Duke, there are also several people working on things like approximate MCMC and "embarrassingly parallel" MCMC (e.g. David Dunson). Duke also seems to have the shortest PhD completion time on average (most students seem to finish in four years), if that is an important factor to you. Overall, I would say that you should definitely consider things like geography and quality of life. If you choose to go the academic route, you may not have much of a choice in where you end up geographically, unless you are a superstar (and even then, it's not a "sure" thing). So the PhD may potentially be the only time that you have an enormous say in where exactly you want to go.

Congratulations on your acceptances! University of Toronto and Carnegie Mellon are both excellent schools. The Department of Statistics & Data Science at CMU is particularly strong in the areas that you listed as interests -- not to mention its proximity with the CMU Machine Learning Department which is also world-class. So this would be an excellent choice for your particular interests. I am not as familiar with UofT, but for the field of Statistics, I don't think it really matters if you go to the same school for your PhD as your undergrad. I know several people who did this, and they are currently in great academic jobs right now (and this is including not just people who went to Stanford, Harvard, or Berkeley for both their undergrad and grad, but also folks who did both their Bachelor's and PhD's at schools like University of Florida). So this isn't really an issue, IMO. If you are interested in academia, you should aim to work with a PhD advisor who can help you publish/submit papers *before* graduation, so you can be competitive for postdocs and TT Assistant professorships. As for the immigration stuff, it should be noted that even in the current political atmosphere, there is no H1-B quota or cap for university workers in the U.S. So usually, American universities have a lot of free reign to hire people irrespective of their immigration status. And if you are reasonably productive in research and your teaching is adequate, then your job is very secure, and there should be very little difficulty transferring from H-1B status to green card. But this pertains mainly to faculty -- international PhD students typically can't become PR's as easily unless they marry an American citizen.

It's definitely possible for somebody with a "non-traditional" background such as yours to transition into Biostatistics. But if you haven't taken math since high school, you would realistically need to take Calculus I-III and Linear Algebra before applying to graduate programs in Biostatistics. Those would be the bare minimum courses required to get into a Masters program. But it can be done. I've seen people with all sorts of undergrad degrees like sociology, biology, etc. pursue Masters in Biostatistics, but they did have to spend some time getting the math requirements done. Some of them were able to do it in less than a year (i.e., they took Calc I in the spring, then Calc II one summer term and Calc III and linear algebra in the second summer term). To get into a Biostatistics PhD program, you would also need to take Real Analysis (an upper division math class). If you decide that Biostatistics is really the career path you want to pursue, then you need to figure out a road map for fulfilling these course requirements. And you need to perform reasonably well in them (B's or higher).

As it currently stands, I think it would be tough for you to be admitted to the PhD programs you listed in your original post, or a school like UMichigan, since your math background is not as strong as that of the strongest applicants from ISI (or students from more closely related majors like CSE from IIT). You have a solid GPA from a strong pedigree, but I'm not sure that is enough when compared with other international applicants -- it would be a different story if you had lots of experience with advanced math, in addition to your B. Tech from a prestigious school. Mathematical preparation/grades and letters of recommendation are the most important part of PhD admissions for Statistics (I'll assume that your GRE Q score would be more than adequate with your background). Unlike other fields, research experience tends to be downweighted for Stats PhD admissions, and there are plenty of successful applicants from pure mathematics backgrounds who have little to no research experience. I don't think online courses are a substitute for grades earned in regular classes. If you could obtain a Masters degree in Statistics first (including one from a institution as reputable as ISI), that would help your profile a lot. Even taking just one year of courses in a Masters program would make your profile much more competitive. I know that this would delay your PhD applications, but I think your chances would improve greatly if you were to enroll in a Masters program.

Oh, I just noticed in the end of your second paragraph that you mentioned you've only taken the minimum math requirements. I think your best bet would be to take a few more advanced, proof-based courses and submit these grades with your application. I am not sure how feasible it would be to take these classes at a reputable institution in your home country and get grades on a transcript for them in time for the December application cycle. But if this is in the realm of possibility, I think that should be your absolute top priority rather than working on economics research or studying for the math subject GRE. Economics research is unlikely to factor heavily in the admissions decisions, and there will be a *lot* of international applicants who scored well on the math subject GRE, so I'm not sure how much that will be make your profile "stand out."