Stat Assistant Professor

I wouldn't upload certificates or awards. It usually isn't necessary to submit any supplemental materials, besides those specifically requested by the program (e.g. some programs like Berkeley want you to submit a list of all the upper division math/stat classes you've taken, textbook, course description, and grade received). However, in your case, since you have publications, it may be useful to upload one of your publications -- although I seriously doubt that most adcoms will have the time to read and scrutinize it very carefully (they may read the abstract and conclusion and skim the body). Remember that there are typically hundreds of PhD applications to read through. I would suggest uploading a publication (as evidence that you have familiarity with academic research). But it is a very good idea to mention in your statement of purpose that you have publications and describe them in a bit of detail. Ideally, you will also have one or a couple of your LOR writers discuss your publications as well, since the letters are the main vehicle through which "research potential" of PhD applicants is judged. Best of luck.

Your list of schools looks very reasonable with your profile. Good luck.

Do not send more than 3 letters of recommendation. If you send a fourth one, the adcoms will probably only look at three of them. I would recommend that you send one letter of recommendation from the graduate-level statistics professor who can discuss your math/stat abilities (so you should probably tell one of your research advisors that their letter is not needed anymore)... you don't really need three letters to all come from research advisors. Having at least one from someone who has taught you and can speak to your classroom performance is a better idea.

I returned to do my PhD in Statistics just over two years after getting a Masters degree (in Applied Math). I think you'll be fine if you do decide to go through with returning to academia. You have already taken the math-stat sequence before, and learning something a second time is *always* easier, even if it's been awhile since you've seen it (you'd be surprised at how quickly it comes back... it's all buried in the back of your head somewhere). You do not need to have specific research in mind or a professor to work with before applying. For Statistics, PhD students are accepted by the *department*, not into a PI's lab or research group. You can mention some topics that you are potentially interested in researching in your statement of purpose, but don't make it too specific. Your application should be given the same consideration as those entering with only a Bachelor's. Actually, the fact that you did a Masters will probably make you more attractive to adcoms and make them more comfortable with admitting you -- since they know you will already know what to expect in terms of rigor. Domestic students who enter Statistics PhD programs straight out of undergrad or without a Masters degree in Stat or Math tend to have higher attrition than those entering with a Masters (or previous graduate-level coursework).

I think the best to target would probably be schools like Emory and Vanderbilt. I would recommend that you also try for a few schools at the level of MD Anderson, in addition to UNC-Chapel Hill.

Adcoms will most likely ignore your volunteer experience. It won't look good or bad.

Yes, you should ask one of your LOR writers to address these circumstances in their letter and say that you ultimately overcame these setbacks, improved considerably after dealing with them, etc. You can also mention it in your SOP. But the personal statement just doesn't carry that much weight.

I don't think your chances are great for Berkeley, Columbia, or Michigan. NCSU won't be easy to crack either. Secondly, your personal statement will not really matter that much, no matter how well-written -- your letters of recommendation count for much more, so you should try to ensure that these are very strong. If you must, I would apply to only one of Berkeley, Columbia, Michigan, and NCSU, and instead, apply to some schools like Texas A&M and Purdue. A few more around the rank of CSU, e.g. Ohio State or Florida State, would be a good idea too.

It's hard to say for Harvard and JHU (as in, I wouldn't classify you as a "shoe-in," but most likely would be "in the discussion"). But I think you have a good chance at being admitted to UNC-Chapel Hill and UPenn Biostats. UVA is a great school, and your GPA is quite good. Physics is a difficult subject too, so I'm sure the adcoms will look favorably at your physics background. You don't need to retake the GRE. A 162 or lower on the Quantitative Section might raise some red flags and warrant a retake, but any Q score 163+ rarely makes any difference in the application.

Your profile looks great, and your list of schools looks very good too. If you are concerned about your chances, it may be a good idea to add one or two schools ranked below UCD and UCLA. But I think you have a good shot at the last 4 on your list, possibly even one or two of the ones you listed before that. Very few applicants to Stat/Biostat PhD programs have publications, so you needn't worry too much about this (a few Biostat PhD applicants who have biology backgrounds might have second or third author pubs in biology journals, but not typically (bio)statistics journals).

Your letters of recommendation will probably make the difference in your case between borderline and acceptance. I would make sure that they are all very strong. It seems as though the two biostatistics letters will be strong. However, I would recommend going with another professor besides your Calc III professor. I would definitely recommend asking a math professor for your third letter, but it may be better to go with the Real Analysis professor or the Probability professor. They can emphasize your ability to handle higher-level math, which would be seen as a huge plus.

I suspect that much of your chances will depend on the prestige of your undergrad and your letters of recommendation (for applicants from "large public universities," there will most definitely be a difference between a Berkeley applicant vs. an Oklahoma State applicant, i.e. the bar will be a lot higher for the latter). If you could give us an idea of the prestige of your undergrad, that may help us to give you a better sense of which schools to target. Many applicants with backgrounds in biology get accepted to Biostatistics PhD programs, provided they have the math prerequisites. And your math GPA is quite excellent, so I think you have a strong chance at many Biostat programs.

I think your mathematics background may be a little lighter than most of the applicants applying to Harvard and U Chicago (especially as an international applicant). UW and CMU have good faculty who work on social science applications, so I think these are both good programs to target as "reach" schools. Still, the programs that you've listed are very competitive, so I would make sure to apply to a much wider range of schools. In fact, the top 20 programs in the combined Stat/Biostat USNWR rankings are all quite competitive, so I would definitely send half of your applications to schools in probably the 21-50 USNWR combined Stat/Biostat range. Your pedigree should help you, but I would recommend taking a few more math courses... numerical analysis, linear algebra with proofs, maybe a second semester of Analysis.

In that case, I would apply to a mix of biostatistics and statistics departments, but it may be a good idea to research the departments more thoroughly to make sure they have enough faculty working on areas that interest you. From what I gather, UNC Statistics is very theoretical and its main strength is in probability theory/stochastic processes. UPenn Wharton is also more on the theoretical side. NCSU and TAMU seem to have more faculty working in applied areas (TAMU in particular has research groups in Spatial Statistics and Astrostatistics). I would recommend you shoot for NCSU, TAMU, and schools like that in statistics. For Biostatistics, you seem to have an excellent shot of getting into a top 10 program... your mathematics background seems to be much more vast than most (I doubt very much that the adcoms will have seen many domestic applicants who have taken measure theory or functional analysis!)

Cornell is also a relatively small department (not quite as small as UPenn Wharton which has <30 PhD students total, but still on the small side, with about 8 new students entering every year), so I would think that this is also quite competitive. Columbia is a bit larger, with over 50 PhD students total, but I imagine it's fairly competitive there too. According to their admissions page, Columbia receives about 300 applications a year for their Statistics PhD program. I think your chances for top Biostatistics programs are pretty good, based on what I know about admissions for Biostat. But I also think you have a very good shot at a top 20 Statistics PhD program. Given your theoretical math background, would you prefer to conduct research in theoretical statistics? There are opportunities to do this in Biostatistics programs, but it seems as though Statistics depts would be a better fit overall if your interests are mostly/purely theoretical. However, it may be worth applying to Washington, JHU, and UNC Biostatistics, which place heavier emphasis on theory than your "typical" biostat programs.

I think you might have a shot at NC State or UNC with your background. UPenn is tough to crack since they aim to have an incoming class of 5-6 students every year, and usually only 1 or 2 domestic students will be in the batch (and the ones I've met were all from schools like Stanford, MIT, and Caltech). I think you have a slightly better chance at Duke or U Washington. A lot of it depends on recommendation letters. I think some "safe" schools to apply to are Ohio State, FSU, UConn. The best range of schools to target in your case is perhaps TAMU, Minnesota, Purdue.

I'm not sure about OR, but I think you can get into most of the Biostatistics MS programs and the LEAP program at Boston University.

You sound like a strong candidate, but I agree with the above: the fact that you are an international student at an unranked university will make admissions much more difficult for you. However, since you wrote a paper at your REU and submitted it to a journal, that may help. Undergraduates submitting research papers to journals seems to be quite rare. The reference letters in this case will be crucial. Based on the info you've provided, I would say that NCSU is a good "reach" school to try. You could also try one or two other reaches. The most appropriate range of schools to target is probably those between TAMU and Virginia Tech. If your interests at in Bayesian/applied stats, I would recommend Ohio State, Missouri, and Baylor. I would also strongly recommend that you apply to some Masters programs at well-regarded Statistics institutions (e.g. UWisc, Rutgers, etc.). Having a Masters seems to improve a lot of students in your situation's chances. However, I do realize that it may be costly.

It would probably be viewed as a minor positive point in your application. But the adcom's primary concern is that you can complete the coursework in their program, so your grades in math classes matter more.

Your profile looks pretty strong. I would say you have a very good shot at a school like Penn State, NCSU, or a school of that caliber. It may be worthwhile to try a couple of top-tier programs like Carnegie Mellon or Duke. Are you also interested in Biostatistics PhD programs? I think you have a very good shot at good Biostatistics PhD programs on the east coast (e.g. UPenn Perelman, UNC Chapel Hill, Emory... maybe even Harvard).

Keep your personal statement generic. Talk about your mathematical preparation in your previous coursework and list some topics that you are potentially interested in (e.g. machine learning?). Most applicants to Stat/Biostat PhD programs cannot credibly discuss their research interest at an "expert" level, so the SOP probably won't carry much weight anyway. The letters of recommendation matter much more.

I think the associate professor who can attest to your quantitative skills is the better choice. Go with your instinct.

I think you can probably get into many Statistics MS programs with your current profile. There are a few MS programs that are selective, e.g. Yale, Stanford, and Duke. But beyond those, I'm under the impression that admissions are not super competitive. Even with a Masters, I don't think your application will be competitive for the very top PhD programs (just because of the competition). So if you apply to PhD programs in the second year of your Masters, it would be a good idea to focus mainly on mid-tier schools then as well (you can try one or two top 10 schools though to see if you get lucky).

Unfortunately, I do not think your chances of getting admitted to a PhD program in the U.S. are very good, given the competition among international applicants and the fact that your university is not well-known. Many adcoms -- including those in mid-tier programs -- automatically reject applications from foreign universities that they have not heard of (they will consider domestic applications from lesser-known colleges/universities though, on account of the less fierce competition and the fact that they may have more difficulty recruiting American students). Plus, your cumulative GPA is on the low side. Your best bet is to do a Masters first.

LORs are somewhat important for MS programs but do not carry as much weight as they do for PhD applications. Most Masters programs are cash cows for the department/university, since they are typically unfunded. As a result, many people are admitted as long as they have a decent ugrad GPA and have taken the minimum math requirements (Calc I-Calc III and linear algebra), and admissions is not as competitive. For PhD programs, the department is making a long-term investment in you and wants to see that you have strong research potential, so the letters will matter more (the letters are really the main component through which the adcoms assess research potential).