Innominate

@statbiostat2017 thanks! At Hopkins there were 25 people present at the interview and roughly 20 more who would be interviewing over the phone (mostly international) for roughly 7 available slots. They also shared that they have recently had incoming classes as large as 11 for anyone reading this in not-2017. 200 PhD and 100 MS applications per year is another piece of info I'll relay. I'll be honest: I didn't know Duke biostat existed until after I had already applied. I would have been cautious to apply to such a new program as well. As you can tell from my list I applied to lots of stat programs for the MS under the idea that most are more general and theoretical than biostat, allowing easier transfer to a broad swath of PhD programs if desired. Lots of biostat masters don't have the theory you'd need for that, don't know about Duke.

Undergrad Institution: Public, Ranked 40-60 in USNWR National University rankings. See misc below also. Major(s): Mathematics Minor(s): Applied math, neuroscience GPA: 3.7 (3.77 math) Type of Student: Domestic while male, 25-35 years old. GRE General Test: Q: 170 (97%) V: 164 (94%) W: 4.5 (82%) GRE Subject Test in Mathematics: M: Did not submit. (~60th percentile) TOEFL Score: Native speaker (N/A) Grad Institution: No graduate school. Programs Applying: Statistics and Biostatistics, MS and PhD as noted below Research Experience: None, no REUs, no grant work. Awards/Honors/Recognitions: Grade-based only (dean's list etc.) Pertinent Activities or Jobs: See below, none concurrently with undergraduate education. Letters of Recommendation: Professors I got along with and took several relevant courses with: Linear algebra, real analysis and probability. Pretending that I have any clear idea how strong the letters are would be presumptuous. Coursework: Lots of linear algebra, computer science, analysis through baby Rudin and intro measure theory. Roughly 10 proof-based math courses overall, only 2 non-math grad courses. Any Miscellaneous Points that Might Help: I worked as a nurse for about a year before I completed my undergrad, which I think was noticed at biostat programs. Also I want to note two issues in my transcripts that may help future unusual applicants: 1. I did the first year calculus courses (and lots of science) at a community college. 2. I had some poor grades from a failed attempt a college earlier in life, which are not included in the GPA above. Applying to Where: I'll order by (MS/PhD, USNews Rank). I've noted the financial offers by distributing one-time bonuses over the course of the program and counting healthcare provision at +3k/year. MS applications: University of Washington - biostatistics - MS - Accepted Notification on 1/13. No financial support. University of Chicago - statistics - MS - Accepted. Notification on 2/6. 25% tuition remission which is typical. Duke University - statistics - MS - Rejected. Notification on 3/2. Iowa State University - statistics - MS - Accepted. Notification on 2/22. They actually sent me a PhD offer (after a brief email to ask if I'd be into that) though my application was for the MS only. Funding 20.5k for 9 months of non-teaching TA work (grading). Purdue University - statistics - MS - Accepted. Notification on 2/14. No financial support. UCLA - statistics - MS - Accepted. Notification on 3/28. No financial support. University of Missouri - statistics - MS - Accepted. Notification on 2/15. Offered a TAship at 19.2k for 9 months work (during a masters for those looking for a funded masters) University of Virginia - statistics - MS - Accepted. Notification on 2/22. No financial support. PhD applications: Johns Hopkins - biostatistics - PhD - Rejected. Invited to interview on 1/20. Interview weekend was 3/3. Notification of rejection on 4/8. NCSU - statistics - PhD - Accepted. Notification on 1/5. Offered a TAship initially at 21.8k/yr. Offered a fellowship on 3/30 at 23.8k/yr average, no work required during the first year. Texas A&M - statistics - PhD - Accepted. Notification on 2/23. Offered support at 16k/yr for 9 months of unspecified work. University of Minnesota - biostatistics - PhD - Accepted. Notification on 1/17, including an invitation to visit (not interview). Support at 29k / yr for 12 months of unspecified work. University of Rochester - biostatistics - PhD - Accepted. Invitation to interview on 1/16. Interview on 2/2. Admitted 2/14. Support at 26.7k for 10 months of TAship work, probably grading the first year. Boston University - biostatistics - PhD - MA offer only. Notification on 2/17. I applied for the PhD and they sent me an offer to join their MA program with no financial support. Happy to answer any questions that don't add to the plethora of identifying information above.

May help folks who want to give you advice if you post your original GRE scores. If your first score was Q: 150 and V: 160 that's very different than a score of Q: 165 and V: 145, both of which are "310". In general I'm inclined to think that improving your GRE by 10-15 points is unlikely to change your results dramatically.

Ask the department directly if they keep a placement list from the last few years. This is completely reasonable to ask for during your decision process, and for what its worth I have received this info from some departments after asking. In general I'd err on the side of being skeptical about PhD program placement, since they do not have a PhD program at the school. Also, as other people here will tell you, this forum has more statisticians than mathematicians. You might want to try sites like mathematicsgre.com which has a bigger math population.

Regarding programming: At the programs I visited I usually asked what I could do to improve before I started, and learning more R was the most common response. But this isn't a big factor for admissions, it's because you'll need to those skills to work in most RA positions. So for that reason I think more programming (especially R) is worth your time - you'll likely have to learn it now or later.

I'm a student, but having just gone through the application process and speaking to lots of admissions people I'd say this is unlikely to hurt you. Mathematical preparation, and evidence of ability, is very important to departments, but they don't seem to really care if you have a ton of stats under your belt. One department head flat out told me he doesn't care if applicants have taken even an intro class. Some would rather just teach that stuff to you themselves and others are understanding of the fact that access to advanced stats can vary even among elite undergrad institutions. I also wouldn't encourage you to take "easy" classes to "boost your GPA", even though it is a bit lower than the average at the best programs. Adcoms aren't stupid, they'll notice if you take basket weaving and pottery, but you can definitely take something like data structures or numerical linear algebra, etc. which aren't technically stats but will boost your breadth. If you really took graduate real analysis your junior year and got an A your class choices are unlikely to be limiting.

This person did you no favors with that advice, since the scant evidence that exists completely contradicts it. Stanford lists the average of admitted students at 82nd percentile. They're the only school willing to give a number, but I think there's little doubt it's the most mathematically competitive program to get into. 70th percentile will probably only hurt you at U Chicago, Berkeley and Stanford, while lots of schools outside of that range would look kindly on a score >= 50th, especially if you have a lack of mathematics courses. You can also sanity-check this by considering that there are top 30 math programs with medians around 70th percentile (see OSU for one example of school that provides their data, median 67). Obviously they're going to care more about the subject test score than stats programs.

Of course it's overkill. OP asked if it would give him a bonus, the answer is yes (I'm sure you'll agree). Whether he can spare the time is his decision.

I would not address the C+ directly as it's really not an outlier in your math grades so there's no reason to call attention to it. The advice that I'd heard is that it's usually only necessary to discuss things that are a big deviation from your profile that have an explanation that can't be inferred from your transcript. For instance if you had one quarter where you went through something personal and racked up 2 D's, but you have all A's otherwise, that merits a brief explanation. This will become more clear as you start to apply I think, for instance U of Chicago flat out says not to discuss grades in your statement. Additionally, ODEs is not one of the "core" mathematics courses that they'll give extra scrutiny to. I'm sure you know this already but that set is roughly linear algebra, calculus, probability and real analysis, with analysis being less important at the masters level. I also wanted to comment because I had a very similar profile in terms of big metrics ("top 10 public" school, same major, similar GRE, identical GPA) and was admitted for an MS at 3 out of 4 of the top 10 schools I applied for (biost/stat combined ranking). So I would encourage you to shoot for that range and be sure to include some safeties that are a good fit for your interests. You may be a candidate for a funded MS at a lower ranked school if that interests you. I'm going to offer a different take than robben did on the subject test: Taking the Math GRE and doing well will help you (of course!), and I can't see any reason that reviewing calculus is not in your best interest. You don't have to send it if it goes badly, so why not give it a shot? If you have to choose between that and getting good grades, then get good grades.

Duke's open house may help a lot - hope you get the chance to do the same at Cal

What information are you using to determine that Berk > Duke is objectively false? For instance, USNews rankings are a reputation-only measurement, with UCB sitting at 4.7 and Duke at 4.1 (out of 5), a significant difference which indicates the opposite of what you're saying. lynntoujours, this may have been pointed out to you already but you can get the placement information easily since Duke posts it (http://stat.duke.edu/people/masters-alumni). Scanning over it it seems like industry placements are by far the most common result (triangle area especially), with very few people not finding a good industry job or a PhD. However, take a look for yourself and see if those companies are somewhere you could see working! I think it's also worth pointing out that they seem to have even more folks doing PhDs in not-statistics after leaving Duke. I have no idea if similar data is obtainable from Berkeley but you can formulate a very precise estimate of your odds of landing an industry job through that info from Duke, to see if the triangle < silicon valley claim is justified.

Please let us know if you're applying for an MS or a PhD at these schools. This is an important consideration in "how bad is the C+ in real analysis". Correct me if I'm wrong but it seems like your only (completed) proof based mathematics courses are abstract algebra (B) and real analysis (C+). If you're interested in a PhD admissions committees will want to see evidence that you can handle proofs, and I think this is the biggest place for improvement in your application. You could do this by taking Real analysis II and doing well (probably the best idea) or something like complex analysis, topology, etc. You have another year left in school correct?

Regarding JHU and crime: JHU's biostatistics department is housed in the school of public health, which is by the JHU hospital geographically (southeast of the main JHU campus). This is a bad neighborhood, with violent crime around 5 times the national average (2000 incidents per 100k per year vs ~400 incidents per 100k per year as the national average). The campus itself is safe, they have invested a ton of money in security and if you visit the area you'll see security guards on literally every street corner in the area. There are also shuttles to the 'main' (Homewood) campus that seemed fine to me during the day. I think it's reasonable to say that graduate students live such a cloistered life that this is unlikely to affect them much. However, there's no question that the area is dangerous and it would be unwise to walk around alone at night near JHU biostat. My recommendation is that you look up the violent crime rate in your current zip code to see how much of an adjustment this would be for you (if you're used to big cities it may be normal for you).

I'm assuming that you mean UNC for stats not biostat, may help to clarify that for folks that want to weigh in on UNC, especially since their biostat is a bit more prestigious. I can only comment on NCSU... NCSU: I visited here and found that their placements were heavily geared toward industry. Some of the students (~15 of 130) are even funded by graduate internship training (GIT) programs while they're studying, mostly by triangle area companies. NCSU would not provide me with a complete placement list, just said what companies they had placed people into. This is a bad sign, when schools are proud of their placements they post this info publicly, refusing to provide a list to admitted students is a red flag for me. Further the NRC data (2006) is missing, but it indicates that 60%+ are going into industry. Their industry placements are very good, especially strong with the research triangle companies, but I think it's fair to say that they're not placing as many folks into academia. I'm a student so my opinion on prestige is worthless, but I think their proximity in the USNWR rankings tells you that they're too close to matter. It will come down to your adviser, work, etc.

Duke's MS has about 25 students per year (currently 51 for a two year program) and Columbia's has 160+ per year. 160 students per year just screams cash cow to me, but you can decide for yourself. I'm not aware of any other program that even approaches the size of Columbia. Duke has an extremely strong focus on Bayesian statistics, which lends itself to some machine learning work as well. I know less about Columbia but they do tons of social science work and I think they tend to place people into financial positions generally. Also duke has unusually clear information about MS grads so scroll through them and see what you think of where they're placing (same link as what machine scholar posted) http://stat.duke.edu/people/masters-alumni-pages

What does this even mean? Are you basing this on any data?

Schools in this tier that I've spoken to have acceptance rates that are still around 5-10%, but some of them have higher international application percentages (70%+), and others just get a ton of applications that aren't anywhere near yours, so I wouldn't stress about the percentage. You've left off two pieces of information which would help folks give you a subjective probability: Your school and major/coursework. State school isn't helpful, there's obviously a huge difference between UC Berkeley and New Mexico State, it does matter for your application odds (for better or worse). You can remain anonymous by saying "unranked" or "top 50" or whatever. Your coursework/major matters too, especially if you've had a ton of analysis or happen to have zero linear algebra, for instance.

I'm a student, not a professor, but for what its worth this doesn't seem to be a big deal. I had an interview at a school where the admissions committee member flat out told me that they don't care how much stats students have seen since mathematical preparation is a vastly better predictor for program success. Also, if you search there's a recent thread about SIBS where some more qualified folks say that it's unlikely to differentiate your application or make you stand out. Too many people do it and it's not proof of ability to research of course. The one thing that a lot of schools will want to see is at least one course in distribution / probability theory (calculus based but not rigorous). I didn't have access to any mathematical statistics courses, no research (even in math) and got into both the top 10 schools I applied to.

I'll let more experienced professors weigh in on the unknown-but-good programs, but its worth some skepticism if they're not ranked by either USNews or the NRC. These rankings are primarily (completely in USNews case) based on professor surveys, so if they didn't get enough responses about a school how likely is it that their biostat program is producing good work? Also, some schools above are based in medical schools so I don't think it's true as a rule that the USNews list excludes medical school based programs. For instance University of texas health science center (MD Anderson now), Medical College of Wisconsin and University of Rochester are are medical school based programs. Was there a specific program you were curious about, waitforever?

USNews ranks all the statistics and biostatistics programs together, to separate out the biostat departments just click on the schools and see if the department is "biostatistics" or "statistics". Some schools may be a bit more complicated, for instance stanford has a biostat concentration in their stat department, but you can figure these things out by reading the school's website. Here's the biostatistics ordering though, with the combined biostat-and-stat rank in parens: 1. Harvard (3) 1. University of Washington (3) 3. Johns Hopkins (5) 4. University of Michigan (12) 4. UNC (12) 6. University of California Berkeley (15) 7. University of Minnesota (24) 8. University of Pennsylvania (27) 9. Columbia (30) 9. UCLA (30) 11. Yale (34) 12. Emory (38) 13. University of Rochester (42) 13. Brown (42) 13. Univ Iowa (42) 15. Univ Pittsburgh (47) 16. Boston University (49) 17. Medical College of Wisconsin (53) 17. UIC (53) 17. Univ Texas Health Sci Ctr (53) .... etc I don't believe any of the 4 schools you've listed have ranked biostat departments, why have you chosen those to apply to exactly?

Funded masters in pure math will be a tough one. A good place to start might be with the USNews specialty rankings for geometry: http://grad-schools.usnews.rankingsandreviews.com/best-graduate-schools/top-science-schools/geometry-rankings

Thanks for the response, that's the direction I'm leaning. Both are great schools, couldn't agree more, but it seems that people make a big distinction between Minnesota (as an example) and UW/Hopkins/Harvard, hence my hesitation.

So far I have the choice between an MS at a tier 1 school and a PhD at a tier 2 school. An academic career is a possibility, with a certainty no higher than what a typical undergraduate has (I'm really not sure yet). Since this decision is likely to be made by many other students I wanted to start a thread and collect opinions from the stat/math domain. My question is: Is the MS a potentially valid choice, or am I kidding myself thinking that MS -> top PhD is a realistic option? My GRE scores are already excellent, so my most viable option to improve my profile would be future grades and publications, but counting on those seems risky. In order to give the most concrete possible context I'll spell out my choices. I have applied to both biostat and stat programs so here's an example of each: Biostat: MS at Univ. Washington (unfunded) vs. PhD at Minnesota (full funds + stipend) Stat: MS at Univ. Chicago (unfunded, have not heard yet) vs. PhD at NCSU or TAMU (full funds + stipend) One other relevant note is that I have enough saved to pay for a masters if need be, so the funding is less of an issue. I'm more concerned that it would be for naught and I'd end up with the same or worse PhD options later. I would be especially interested in the opinions of those who are already into an academic career (e.g. cyberwulf, biostatprof) as they have the most perspective on what their peers did or regretted, what's realistic and isn't, etc. I would be grateful also if discussions about school tiers, admissions notifications, etc. could be sent to me in a PM or put in another thread. Thanks in advance, this forum has been a great resource while applying!