Stat Assistant Professor

Agreed with wine in coffee cups. Quality of life is something you should strongly consider (I mean, you'll be living somewhere for 4-6 years, so you may as well not be somewhere that you loathe). I'm currently at a small department, which works for me in terms of my ability to interact with other students and faculty. On the other hand, I readily admit that we don't have as broad a selection of research areas (e.g. we have no faculty working in the area of functional data analysis) or as many faculty working in specific areas. So if you wanted a lot of broad options, breadth would certainly be a legitimate concern. On the other hand, the faculty we have are generally very well-respected in their areas of research, readily accessible, and work with their advisees to publish papers. The most important thing, I think, is to ensure you are in a supportive environment where potential advisors will be committed to your future success. So just make sure you will be able to get opportunities to publish and work closely with faculty at whichever place you decide to go. PS - I sent you a PM.

If you can't readily find this information online, you should definitely ask the Graduate Coordinator at these schools. You would be well-positioned to find a job in academia at any of the places you've been admitted to so far.

It does not matter that much. Name recognition matters most for PhD students who want to go into academia (and even then, reputation of ADVISER, rec letters, and publication record will trump prestige of institution). For anyone interested in industry, name recognition may be viewed positively, but is not that essential (my Masters institution is ranked 60-100 in math and unranked in stats, but there were stats PhDs who got jobs at DeutscheBank and Google (as a PhD statistician).

I wouldn't bother retaking it. Just go ahead and apply.

I'm not sure a lot of undergrad statistics courses are really necessary for every school (unless it is explicitly stated on their Admissions FAQ page that it's strongly preferred that applicants have taken a full year of probability and statistics), since there are plenty of PhD statistics students who have Bachelor's or Masters degrees in pure mathematics. There are students in my current Statistics PhD program who never took a single stats class before enrolling. Maybe it's different for the top schools though (I attend one of the schools on the second page of the U.S. News & World Report rankings). Some schools like UNC and Duke seem to strongly prefer that incoming students have taken a Masters level Casella & Berger probability and statistical inference sequence, since it looks like their first-year courses delve into advanced inference and measure theoretic probability theory right away (at many other schools, these courses aren't taken until the second year of the PhD program). But if prior credits earned in probability/statistics is not an explicitly mentioned prerequisite for admission and if the first-year courses on the Department webpage start out with Masters-level probability and statistics (i.e. not measure theoretic probability or advanced inference), then you should be okay, IMHO.

I don't think the algebra class will negatively affect your chances. Better just not to draw attention to it, if the rest of your application is very strong and clearly demonstrates ability to do applied (rather than pure) math. The letters of recommendation should help a lot. Best of luck.

Just an fyi, this forum won't allow you to edit your posts after a little while, so to keep us posted on admissions results, put the results in your signature too.I'm sure that will help a lot of future applicants as well.

If you applied to their PhD program, I'd assume that they would give you financial support (or at least waitlist you for it) if they accepted you.

I would definitely advise taking it. For applied math programs, it seems as though the adcoms are a lot less stringent about what counts as a "good" score.

Yes, you would be a competitive candidate for pure stats programs. If your letters are all really strong, I could see you getting in some top 10 places.

Also agreed with Biostat_student_22 that previous research experience and work experience are not that crucial for stats/biostat PhD applications. It seems like the most important things are grades (esp. in upper division math/stat), coursework, letters of recommendation, and reputation of undergrad institution.

Your profile looks very strong and your grades show a definite upward trend, so I wouldn't worry too much about the P in Discrete Structures. The A's in Analysis should signify you have the chops to do proof-based math. In fact, in a lot of grad applications, you are asked to specify your GPA in your last two years separately from your overall GPA, and it is my impression that the performance your junior and senior year carry more weight If you are a domestic applicant, I think you have a very good shot at most top 40 programs... especially given that UCLA also has a great reputation. I would focus on making sure you get three very strong letters of recommendation... that should give you a very good shot at top 10-20 places, IMHO.

You should definitely mention the personal circumstances in the application (either the supplemental info section or the SoP, or ask one of your LoR writers to allude to it). Don't dwell on it too much, but the point you need to get across is that these two terms are not reflective of your true mathematical ability, and you'll want to emphasize how you improved greatly after that. You may want your LOR writers to make a case with regard to this too. Also agreed with footballman2399 that a decent score on the math GPA could help you make your case. You'll also want to get very strong letters of recommendation that can say a bit more than just "one of the best." If your professors could be a bit more specific (e.g. pointing to specific coursework, class participation, projects, etc.) and say you're one of the best students they have ever had, that your research potential is excellent, etc. that would work in your favor. Your GPA would still keep you out of the very top programs, but there are programs that tend to place stronger emphasis on letters of recommendation (case in point: there is a student in my current program who went to a liberal arts college that didn't even give grades).

I think Rice University and UNC Chapel Hill are also a good place to apply to. There's faculty working on machine learning and statistical learning stuff at these places. Your list is pretty extensive. It seems like a lot of Americans who are in stat PhD programs also opt to leave with a Masters. I think international students are less likely to quit after getting the Masters, because their job prospects in the U.S. tend to be better with PhDs (the bar is definitely higher for them -- meanwhile, a lot of American students who don't have to deal with employer sponsorship, H1-Bs, immigration, etc. have a much easier time finding work with just a Masters).

I think the work experience might help a little bit if you can put it in your SoP and have your LoR writers allude to it. It is definitely not the most important thing though (for PhD admissions, maybe it's different for MS) -- you have to communicate mathematical readiness and research potential in your application (through your SoP, LoRs, etc.) above all else. Therefore, your grades in upper division math/stat courses and your letters of recommendation are the most important part of the application (the relative weight depends on your undergrad institution too -- e.g. if you went to UChicago, MIT, or another school known for grade deflation, then slightly lower grades may be okay, or if you have stellar grades from a prestigious schools, good but not great LoRs may suffice).

That is likely to be the case at most statistics and biostatistics departments (i.e. majority international students from China). It seems like American students are less likely to pursue advanced degrees in statistics past the MS though... many more seem to be in mathematics.

If these programs do not require a one-year sequence in Casella-Berger probability/inference, regression, design of experiments, statistical computing, and linear models (at the minimum), then they would probably not be on the same level as other MS programs in Statistics, imho. I would look into MS programs that specifically have the requirements I listed... although I am not sure if a Data analysis or "statistical practice" program constitutes statistics? These might be something totally different.

I am not aware of any Statistics joint MS/PhD programs (i.e. where people who perform well in the MS can just go straight into the PhD program). Typically when people are in terminal MS programs in Statistics (as opposed to being in a PhD program and getting a Masters along the way), they have to apply again to the PhD program or to other PhD programs. However, in this case, it might be possible to obtain the PhD in four years rather than five. A lot of PhD programs will allow transfers of credits and will allow you to skip the Masters level classes if you can pass their qualifying exam. I think any MS program that requires the Casella-Berger probability/statistical inference sequence, regression, design of experiments, linear models, and statistical computing should be good preparation for a PhD program. If you get the terminal MS and then continue onto a PhD program and pass the qualifying exams before starting your first year, then that typically saves a year of classes in the PhD program. (That is what I did -- although my MS is in Applied Math, I had taken the courses that were tested on the first year exam, and I passed the exam so I can finish the coursework requirements in two years instead of three).

I think UCLA has some faculty. It may be worth applying to Sociology departments too, since they have people who do computational social network analysis. I saw that UMn Statistics Dept hired a faculty member recently who had a PhD in Sociology (undergrad in mathematics).

I think the schools you have listed are most definitely reach schools. Honestly, if you are an international student who isn't from a handful of universities and you don't have a lot of A's on your transcript, it is going to be an uphill climb at a lot of universities ranked in the top 20 or so. That is why I suggested that your best chance is to get a Masters degree in the U.S., doing well (so as to offset the earlier performance a little bit), and then applying to PhD programs. I suppose you could improve your GRE Q score, but it's not so disastrous that it raises a red flag. The grades are the bigger problem for you.

If the adcom is not as familiar with your institution, they will weigh your letters of recommendation and your mathematics subject GRE score more. So do well on the math subject GRE. But additionally, you indicate you have gotten a "D," so you may want to have one of your letter writers address this or address it somewhere in the supplemental information section (you should thoroughly explain why it does not reflect your ability and emphasize your strong performance in other classes).

University of Pittsburgh is another possibility for Stats. I think you have a lot better chance in biostats than most stats though, TBH I don't think biostats will limit you to only working in pharmaceuticals. Your marketability in industry depends not only on degree but what your skills are. If you have a degree in Biostatistics, but are familiar with Hadoop, machine learning, and data science, then you should be fine to apply to a lot of data science jobs, for example.

You should be able apply to multiple programs at the same university, but this is not typically done for PhD programs (it is more common if you're applying to professional Masters programs, e.g. applying to MSW and MPP at the same school). But as you have mentioned, this may not look good to the Math department admissions committees. These adcoms really want people who only want to do math (especially if it's pure math), so I suspect that you are right that they won't receive this warmly. What kind of math are you interested in? You could apply to programs that have applied math faculty, some of whom work in financial mathematics. The department where I got my MSc did not have a separate applied math department (applied and pure were both in the same department), and some of the faculty studied financial math.

If you want to apply to statistics programs and not just biostats, you may want to try applying to some bigger stats programs that have faculty working in biostats (NC State, Ohio State, etc.). Some schools like OSU are more forgiving of lower GPAs, and they do not separate biostat from stat. I am not familiar with the program at UC Riverwise at all. I think one of the faculty members and some other members on this forum may have indicated skepticism towards degrees offered in "Applied Statistics" (including ones offered by highly reputable institutions like University of Michigan). I think this is probably true, since I took a look at an Applied Statistics "professional" Masters program offered at my alma mater (an Ivy League school), and this program does not even require a lot of the core classes that one would have to take in a typical Statistics MS program. I think it would be better to get a degree in Statistics or Biostatistics, but to focus on applied statistics for your actual research (that way, you have sufficiently learned the theory and had to demonstrate proficiency in it on quals, even if you actually work in applied topics later). Getting a degree that is called "Applied Statistics" may make employers question if you had enough training in theory to understand the theory behind the applied stats.

Your rec letters should be from professors in statistics or in other quantitative disciplines (mathematics, etc.). And I agree that it may be more suitable for you to apply to biostatistics PhD programs than statistics (unless you get another MS in Statistics in the U.S. first but it seems as though you don't want to pursue that option). Based on your info here, you wouldn't be amongst the most competitive international candidates for stat programs (and there are some very impressive ones from ISI, Peking, and Tingshua alone). But even Biostatistics seems to be very competitive too. Maybe a biostatistics faculty member or student can speak to the OP's competitiveness for biostatistics.