wine in coffee cups

I agree with bayessays, who has made a useful contribution to this discussion by listing examples of statistics programs ranked in the 60s and why you might choose them. Generic questions like "would it be in my best interest to accept the PhD offer I currently have or go for a funded MS program" have one right answer: "it depends". You need to give more info! The specifics of which PhD/funded MS programs you are considering and what research you are interested in are important on the off-chance that someone here is knowledgable about those departments and can offer more tailored advice.

I am not knowledgable about these specific departments, but I'd like to say things about some issues you raised. Funding: sounds like the differences in funding packages are not a concern here. If you have a livable stipend and the work it requires is not unduly burdensome according to current students, consider that good enough. Professors who have no students: ask current students what the deal is. If it's just the case that what they work on happens not to mesh with the interests of current students, fine. But every department has its a-holes, and it's good to find out who they are before you decide. If this person is an intolerable a-hole, you should look for other options. Professors with many students: talk to several of their current students and ask how independent the students' working styles are, how often they have one-on-one meetings, email responsiveness, group meetings and how useful those are, can they get attention when they face roadblocks or deadlines, how progress is monitored, if postdocs or advanced students of the same advisor competently guide newer students, etc. A huge number of advisees might not be a problem if the professor has stellar supervisory and organizational skills. I would worry if advisees suggest they often feel out to sea, abused and overworked, pressured into specific research directions they aren't interested in, or if favorites are treated well while everyone else is neglected. Paraphrasing advice a friend (surely lurking, hi!) shared with prospectives recently: at some point academic differences between programs are too minor to matter in the grand scheme of things. You are considering departments that have similar reputations, adequate funding, (hopefully) multiple viable potential advisors, diverse research areas when you change your mind, and outcomes you are content with. I think quality-of-life considerations are the deciding factor. Anywhere you go, you will face research setbacks, career anxiety, existential crises, deadline pileups, advisor disagreements, and personal problems at inconvenient times. When those things inevitably happen, you want to have a supportive community of students to commiserate with. (Could happen in a tight-knit program, or could happen in a big program with enough like-minded people.) You should feel able to get mentoring from faculty besides your advisor and not be in a fractious political environment preventing this. You need a place where you can have some degree of work-life balance to preserve your physical and mental health. Good luck with your eventual choice!

^ I would be skeptical of the associate professor placement at Purdue I was just reading a paper by the 2014 A&M graduate who started as an assistant professor at Purdue. The affiliations indicate his advisor (F. Liang) has recently moved to the biostat department at Florida, FYI.

Unless they've changed things in the past couple years, they suggested that admitted students schedule an individual visit over a couple of weeks in March after their spring break. There was not one big recommended admitted student day the way some places like Berkeley, UW, Michigan do it. I ended up visiting Columbia the same time as a few others and they had us all follow the same itinerary of meeting with professors together, lunch with a few students, etc., but it was a Friday and unfortunately not many people were around that day.

Many of the faculty in that department specialize in very theoretical work related to misspecified models, post-selection inference, high-dimensional theory, and statistical learning theory. Collectively they have a lot of recent Annals of Statistics papers (top theory journal). The more applied people in the department work on Bayesian variable selection, hierarchical modeling, causal inference, experimental design, and genetics. I think of mathematical finance as being more in the probability, time series, stochastic process, extreme value theory, and copula literature, none of which I know much about. There's a couple faculty at Wharton whose work is adjacent to or in these areas but nothing like you might expect based on the knowledge that the department is housed in this fancy business school. (Someone should pipe up if you think I'm mischaracterizing the program.) I perceive Columbia stat as the most finance-oriented of these three by a good margin, followed by Chicago stat and then Wharton stat.

That rate is before taxes, but you will not have an effective tax rate of 20%. It'll be closer to 10%, maybe less (mine is 6% in a state without income tax). Grad students pay federal and state income taxes on the stepped schedule like everyone else, but full-time students working as academic student employees are exempt from FICA and Medicare withholdings. Using an online calculator for an estimate and ignoring the FICA/Medicare lines, a student on a $27K stipend for 12 months in California would owe about $2074 in federal income tax and $484 in California income tax. University-owned grad student housing is often more expensive than comparable non-university housing even after accounting for bundled utilities, so I wouldn't get stuck on the $1300/mo rate until you talk to current students. That *might* be about right if you insist on living alone in a nice-ish 1BR, but sounds like an overestimate if you are willing to share a house/multi-bedroom apartment or live in a crappy studio.

I just looked and there are 9 entries for fall 2015 Columbia biostat PhD decisions, with specific comments about fellowships and visits. Those are all fake?!

Make sure you look at the period of the funding. $20K for 12 months or $20K for the 9 month school year? If the latter, the possibility of working on campus during the summer as an instructor, TA, or RA for an additional $5-7K sounds much more doable. Are you sure they expect you to live in on-campus housing? Graduate programs aren't paternalistic about campus residency. Ask current students when you visit. Not everyone will talk financial specifics with you, but many will share rough information about their monthly housing, utilities, and food expenses. I wonder if you might be mistaken about the stipend ($1600/mo is low for the Bay Area, and at least for Berkeley stat PhD, it was $2250/mo in 2012), or if indeed students commonly take out supplemental loans or split costs with a partner to get by.

Booth has quite a few Bayesian statisticians doing work I think is pretty interesting that is pretty far removed from classic econometrics (Gramacy, Lopes, McCulloch, Polson, Taddy). Most of the co-authorships on the papers I like appear to be with each other and collaborators at other universities, though, not so often with current Booth students. Seems like kind of a shame to me not to have more students involved. The student thesis proposals/defenses for the econometrics/statistics students don't sound that reflective of the methods work those faculty are strong in. Is there a mismatch between the research interests of the students and faculty there, what's going on?

Awesome! My main suggestion is to have the data frame returned by brewdata() contain the original program name as a column. Setting map=TRUE lets you get the school name, but I think it makes sense to also return the program name. That way users can remove false positives, e.g. exclude programs like "Educational Psychology - Learning Sciences (Research, Measurement, And Statistics)" from statistics-related results. This seems really important for disciplines like math, where searching for "math*" gets you both pure and applied programs, which are impossible to disentangle without the program name. I also suggest changing the default query to "(stat|stats|statis*)". You actually miss out on a decent number of Duke results because their program is formally called "Statistical Science", for example.

I'm in a program that is much more heavily American than most, so my department is pretty social (we plan frequent happy hours and parties, good intramural sports participation, etc.) The other students here are great, though it took a little while to get to know people who I didn't have classes with besides my office mates.

Agreed, I was just about to post the same thing! The emphasis in the first year or two of an American statistics PhD program is on coursework and learning the material well enough to pass qualifying exams, which functionally means spending a ton of time studying. You can work on research, but it's more on the back burner until your second or third year when you're done with the hard requirements.

It's common for undergrad math majors interested in statistics to take a semester of probability and a semester of math stats (like bsharpe269 did), which are usually slower versions of that master's-level C&B course sequence you're in. Even though the undergrad sequence is not a formal requirement, the class you're taking is probably designed with the expectation that most people have seen the less-advanced version before. But there are probably others in the same boat as you. It's hard to learn completely new things in a fast-moving course, but you'll start catching up as the semester goes on. There are a lot of little tricks you use over and over again. These are surprising the first few times you see them but become old hat as you get more practice. (For example: recognizing a term as proportional to the integral of a known probability density over its support, which has to integrate to one, so you just have to mess with the normalizing constant.) I hope that even if you struggle to come to solutions yourself now, you at least understand why they're right. After you see enough of them, you will be get better at generating them yourself. I would make a regular study group with some other people in the class. It'll be good for you all to practice explaining the intuition behind how you approach various problems. Be sure to take advantage of office hours too.

Specifics depend on what courses you start with, but I think you're better off reviewing calculus and linear algebra than anything proof-driven. Having stuff like the proof of Heine-Borel fresh in your mind will not speed up work, but having a good command of calculus will. I had made a post summarizing the specific math I wish I had reviewed before starting my program Get comfortable with LaTeX before you start. Even better, learn how to integrate it with R using knitr. Some of my classes required homework to be typeset, but nobody is going to teach you how to do this if you don't already know. The people who had not used LaTeX seemed to have a hard time learning as they went along, and Microsoft Equation Editor output looks like garbage so you'll want to not use that as a crutch.

If you are offered funding, take it. That either means they think you are one of the strongest incoming students, or it's a small program and you're going to have the opportunity to work closely with faculty (e.g. as a research assistant or on a master's thesis). In either case you are being set up to get good advising and have nice references for employment or for a PhD down the road. Remember that rankings of departments are based on their faculty and PhD programs, which won't necessarily mean you get a good master's experience. As a rule of thumb, the larger the program, the more you're competing with other students for attention. Being one of the top students in your classes is the main way you distinguish yourself in a non-thesis MS program, and that's easier said than done in a big program with many other smart and hard-working people. For PhD applications, it matters that you get faculty enthusiastically on your side. To do that, I would guess that you're better off being one of the top couple of students in a smaller less-known master's program than being outside of the top few students in a large master's program that just happens to be linked to a highly ranked PhD. (The exception would be if that program happens to recruit a lot of its PhD students out of its master's cohorts.) For jobs, this doesn't really matter, because employers mostly care about whether you have the experience they are looking for and come off as quick-learning and competent. I'd pick the program primarily based on the geographic area you want to be in and what kinds of jobs its graduates get, maybe with debt as a tie-breaker. A fancy name will open more doors nationally, but a regional job search is logistically much easier. For local job hunting, the added value of the fancy name is smaller since there is more of an awareness of the non-fancy programs nearby.

No clue.

I don't think part-time master's students are a large share of that applicant pool. They are basically all already living in Seattle and employed off-campus part-time or PhD students in other departments at UW.

Sit tight guys, I don't think they sent everything out at once.

There is a current 3rd year biostat PhD student who was admitted off the master's pathway at the end of her first year. I'm pretty sure there is another who was also admitted to the PhD off the MS and would be a third year, but she took a job in Seattle a few months ago and I think is returning at a later point. One thing you have to do to get accepted off the pathway is pass the master's exam at a designated PhD level. I don't know about the most recent exam, but in 2012 and 2013 only about 60% of statistics and biostatistics graduate students taking the exam passed at the PhD level.

Would not be at all surprised if someone forgot to update a line in the initial wave of offers that went out yesterday. The official numbers from last year was 284 applicants, 28 offers. For confusing reasons these include part-time master's students (who were 3 of the enrolled students, unsure how many were offered or applied), but not full-time.

Even if you hadn't missed the financial aid deadline, it sounds like the funding situation at Rutgers is dicey, as a few applicants from recent years have mentioned in the results database. Obviously these are not official statements, but I think comments like "Rutgers is in financial disarray right now and I believe they have almost no funding" from the previous application season are a little scary. Also, if you look at their current PhD students, you see that only one listed as starting Fall 2014 (appointed as a TA) and two in Fall 2013 (both TA). I see they have start dates missing for a few students, but to me this suggests they haven't been able fund more than a couple of new PhD students in recent years. Good luck dude, something will work out!

For UW biostat, looking at what current students are working on might help you get a better sense of the range of research. The department has weekly student seminars and all the talk abstracts are conveniently posted here. As for job prospects, biostatistics faculty positions or postdocs are the most common. Graduates often want to stay in Seattle and take a position at a local research organization (the Fred Hutchinson Cancer Research Center or one of the many hospitals or health care systems in the area). Few alumni end up in private industry (~5% of PhDs). Right now, alumni tend to stay within biostatistics, and the department is not a huge feeder for non-biostat jobs at tech companies, general "data science" positions, or anything like that. I don't know how this compares with other biostatistics departments.

I think work experience can help you come off as an interesting, mature person with more focused research directions and is helpful in that sense. However, I'm not sure it'll make you much more competitive than you were when you applied two years ago. The sense I get is that academic statisticians don't give work experience much weight in admissions relative to things like letters of recommendation, coursework, and grades. This is probably because your job experience has little to do with whether you're going to eventually pass PhD courses and qualifying exams. Also, most people involved in admissions have never held non-academic positions and just aren't going to be impressed by anything you did that wasn't scholarly research, though they might have a vague appreciation for positive personal qualities and technical skills you developed from working.

The UW program is a normal MS in statistics program with a funny name. It takes a little under two years (5 quarters, no classes in the summer) and covers Casella and Berger theory, experimental design, regression analysis, statistical computing, GLMs/GEEs/mixed effect models, categorical data analysis, plus whatever electives you take (e.g. nonparametric regression, stochastic processes, statistical learning, causal inference, hierarchical models, social network analysis, time series). Both master's and PhD students take these courses along with biostatistics PhD students (and biostat master's students in the theory course). About 20 students have entered the stat MS program each year for the past two years.

Current student at one of the schools you mentioned. I have a friend in the department whose background is similar to yours, so maybe this will be helpful in setting your expectations. My friend did not get in the program initially but were admitted off the waitlist after the April 15 deadline. (They were originally going to go to UCLA.) They did very well as a statistics undergrad and thesis-based master's student at a big university better known for sports than for its statistics program. They took many statistics courses, had research experience as a master's student, established good relationships with the faculty, I'm sure had excellent letters of recommendation. They were basically the star of their old program, but like you, they did not have advanced math coursework past linear algebra. No clue about GRE scores. However, everyone else in the past couple of PhD cohorts came in with more math and almost all came from academically well-known schools, so this person's preparation is not typical of students in our program. As for you, I would guess that NCSU is reasonably likely since you're getting a letter from someone in that department, and that's a really big program that admits students from a variety of backgrounds. The others, can't hurt to apply, but don't get too attached to anywhere since your math preparation might cut you out of the running. With your excellent grades and good references, though, I bet you will be competitive at a lot of departments outside the top dozen or so that are more willing to take an otherwise strong student without real analysis. I agree with cyberwulf that you should give biostatistics departments more thought, too. You might be able to get into UW slightly more easily through the biostat department instead of the stat department, for instance. Good luck!