Statistics PhD Admission Advice Thread

[Bayesian1701]

I am stuck in an airport with nothing else to do and not enough WiFi for Netflix so I figured I would start an advice thread on all things Statistics PhD admissions that I have learned from the process.  I wanted to share all the things I wish I knew when I was applying.  Anyone else feel free to add to it.  I wrote this from a general statistics perspective but some of this probably will be similar in biostats and perhaps even math.   

Admissions - Where to apply

The best indicator of how you will do is looking at the past profiles that are most similar to you.  You will unlikely find a perfect match, but try to look at similar applicants first and then see what kind of programs they got into.  Expect a lot of randomness and try to have a more balanced list than I did.  Also know that safety programs really don’t exist and lower ranked programs (particularly smaller ones) can be just as hard to get into as large high ranked programs.  There are programs that you have a higher probability of getting admitted at then others.  Also research fit is very important so consider that if you have known interests.

Admissions - What to Expect When you are Expecting to go to Graduate School

TL;DR Expect a lot of waiting.  

I am going to be totally and completely honest with you about my experience applying for graduate school.  For undergrad I was admitted two weeks after I applied and had an departmental scholarship two months after that.  I knew I was going there because it was my first choice and I had automatic acceptance.   So I have never not know what I was doing next.  This whole process of not knowing where I am going to be in Fall 2018 for over a year now has been hard.  I still don’t know where I will go.   I have watched results tell me that I was waitlisted and rejected before I knew officially and checked admissions portals multiple times a day.  To protect my sanity and prevent me for spending my entire life on the gradcafe I blocked it from 9-5 every day on my laptop.  I highly recommend it.   I have cried more than I ever have in my life and have questioned myself almost daily.   My impostor syndrome has been awful.  I have doubted whether or not I am ready to jump into a PhD program at 21.   I have questioned why programs admitted me and felt intimidated by the people at visits who already have masters degrees.  I have read all the profiles dozens of times and knew that most domestic students with my level of research experience typically do well but I didn’t think I would do as well as I did.  I have questioned if I really do know what I want to do and whether or not being able to pursue my research is a deal breaker.  I have struggled over how to decide and what qualifies as a deal breaker.   If you had/have any of these feeling you are not alone.

Interviews

I interviewed two times (three if you count Baylor) so thought I would share my experience.  

For Duke the interview was very informal and was more about me as a person than me as a prospective admit.   I was asked what my hobbies were (which totally threw me off) and we talked about life at Duke and in the research triangle. It was about 20 minutes long and with a single assistant professor on the AdComm who was the original reviewer for my application. I was asked for more technical details on my research but that was the only thing application related we talked about.  

For Virginia Tech, they just asked if I had any questions and I asked about funding and their completion rate.  It was implied that funding was competitive and I was told the completion rate was “around 30%” because a majority of people fail quals.  They were definitely trying to court me calling me their “top applicant” and “favorite application” and I really though I had gotten the fellowship.   I had an single semester of support with the possible option of renewing but it didn’t have the language that my other offers did which offered more security.  I mainly applied for the fellowship so when I didn’t get it I wasn’t very interested.

Baylor was a causal visit and I meet with everyone at the department that was there. It wasn’t really an interview.

Visits

I have completed three visits so far and I wanted to help answer the what do I wear question that I had.  On both my prospective student weekends there was a range from button down and tie for men and nice blouse and heels for women to tee shirts and jeans.  The average was around business causal and that’s what I would recommend you do.  Don’t worry about knowing anyone research interests and memorizing CVs if they aren’t your POIs.   All my meetings with professors started with a basic explaination of their research and since you will likely be supported by a TA ship your first year you don’t need to find an advisor right off the bat.  I didn’t really take notes and made a list of people I liked.  You will have information overload and it takes time to fully realize the pros and cons after visiting.

Program Impressions

Next I wanted to share my impressions of programs based on my interactions with them.  

Baylor

Baylor is my favorite program so I am biased.  Demographically it is a majority domestic program (~75%) with about 40-50% women.  Waco is very much a college town but it still has a lot of things to do.  They told me 82% of their students graduate in four years with almost all of them finishing by year 5.  Their placements are mainly in industry with a lot of people at Eli Lilly which is huge supporter of the program and has funded RA positions and grants for the faculty.  It has 3 Bayesian statisticians and 5 other faculty members doing different things.  The main fields of application are biostats and environmental statistics.  They have a nice bank of computers for faculty and students to use to run simulations on.  I think it is severely underrated.  On the Academic side of placements they occur mainly at state universities and liberal arts colleges.  They offered to buy me a plane ticket but instead they reimbursed me for mileage which was over $300 plus a nice hotel room.   It’s a great department that is totally underrated.  

UT Austin

It may be a young department but it is growing fast.  They have a ton of people doing Nonparametric Bayesian statistics with some machine learning and hierarchical modeling mixed in.  It is entirely Bayesian to the point of the visitors being told that if you want to be a frequentist don’t come here.  A lot of the faculty is double appointed in both stats and the business school but they are committed to the Stats PhD program.  It seemed to be about 40% female and a slight international  majority but the domestic and international students were really integrated with each other which doesn’t always happen. Austin is a very expensive city (at least for Texas) and most of the graduate students I talked to spend around $700-1000 on housing a month, my stipend offer was $2000 a month so that seemed expensive. They also have a lot of computing resources for research. They are very proud that their first graduate got a job (post doc I think?)  at Berkeley straight out of a PhD (who wouldn’t be)  My visit was completely funded.

University of Missouri - Columbia

Also know as Mizzou,  the University of Missouri is state university in a college town.  Columbia’s airport is awful (I have been stuck here for 5 hours since there are only like 5 flights a day), but Columbia is a nice city that is bikeable and the bus system is apparently good.  The department has a wide variety of research  but it probably more Bayesian than classical and more focused on environmental statistics than most programs.  Demographically it appeared to be about half and half domestic and international but not a single international student did anything at the visit until the free dinner where they didn’t sit with the prospective students.  A grad student made a comment that they never want anything to do with us (referring to the domestic students).  This is a lot different than UT and Baylor where  all the graduate students knew each other pretty well.  The faculty do seem nice and hierarchical modeling was a common theme from the research presentations.  Their placement was a mix of academic and industry and generally good for a mid tier program.  My visit was completely funded.

Virginia Tech

I never actually visited but it I didn’t like the completion rate and the potentially unstable funding package.  The professors seemed nice but I am kinda bitter about how they called me their top applicant and then didn’t even give me two semesters of funding.  I am sure it is not a bad place but I was no longer interested when I began to get my other packages.  Their academic placements aren’t great as well.  I personally wasn’t interested in visiting given my concerns but they offered to fly me out on a paid visit.  

I talk about TAMU when after my visit.  

[speowi]

On 3/6/2018 at 3:21 PM, Bayesian1701 said:

The best indicator of how you will do is looking at the past profiles that are most similar to you.  You will unlikely find a perfect match, but try to look at similar applicants first and then see what kind of programs they got into.  Expect a lot of randomness and try to have a more balanced list than I did.  Also know that safety programs really don’t exist and lower ranked programs (particularly smaller ones) can be just as hard to get into as large high ranked programs.  There are programs that you have a higher probability of getting admitted at then others.  Also research fit is very important so consider that if you have known interests.

This is a good post with some helpful insights, but I would take issue with some of the things you said in the quoted section above.

First, I don't think that the best indicator of how you will do is to look at past profiles that are most similar to you. While this might be true in theory, in practice people often provide insufficient detail for meaningful comparisons to be made. For example, someone might say that they have published several articles or presented at several conferences, but without knowing the kind of journal or conference they're talking about, it's hard to really gauge how much weight admissions committees gave to this aspect of their application. There is a difference between publishing in JASA or Annals of Statistics vs. an undergraduate journal, between presenting at the top conference in a particular field of statistics vs. presenting at a poster presentation at one's university/college or summer research institution, and between being first author vs. some other author. Similar things can be said about GPA, coursework, and academic awards with respect to rigor. In addition, letters of recommendation can play a big role in the process, but people often describe their letters simply as "very strong," "probably strong," "decent," etc., which isn't very descriptive at all.

I think that a better way to get a sense of how you will fare in the admissions process is to talk to people in your social/academic circles, particularly your professors and any friends who have gone through the process already. Since they are presumably more familiar with your work and with the admissions results of students coming out of your university/college, they are probably in a better position to gauge how you might do.

Second, while I agree that there is some variation in the process, I don't agree that "safety programs really don't exist and lower ranked programs (particularly smaller ones) can be just as hard to get into as large high ranked programs." Of course, I would encourage applicants not to take admission to any particular program as a given. But I think it is fair to say that someone who is competitive for Berkeley or Stanford--assuming that we can know such a thing--probably does not need to apply to as many lower-ranked programs as someone for whom said lower-ranked programs are a match. I think that within broad tiers of schools--say, top 1-30, top 30-50, top 50-100, etc. (I'm just making up the numbers here)--a single applicant can expect some variation in results due to factors like research fit and sheer randomness, but it seems to me that there are clear trends across/between these vaguely defined tiers. In other words, someone who is a match for top 1-30 schools would probably fare better among those schools than someone who is a match for top 50-100 schools. (I seem to recall that some people, historically, would disagree with me about the variation within tiers; they think that if someone is admitted to one top 5-10 school, they will probably be admitted to all or most top 5-10 schools. But I think that these people would agree even more with me, then, about the presence of trends across/between tiers.) Of course, people do sometimes get into their reach schools and rejected from their safety schools; my view is just that it is still meaningful to categorize one's list of schools into reach, match/target, and safety schools, contrary to what was implied in the OP, if the categorization is done well and doesn't just split hairs between, say, the top 5-10 programs. Also, this is only tangentially related, but one should note that acceptance rates are not necessarily a good indicator of admissions competitiveness, since there's a good deal of self-selection among applicants to these schools.

Edited March 8, 2018 by speowi

[Bayesian1701]

 

2 hours ago, speowi said:

This is a good post with some helpful insights, but I would take issue with some of the things you said in the quoted section above.

First, I don't think that the best indicator of how you will do is to look at past profiles that are most similar to you. While this might be true in theory, in practice people often provide insufficient detail for meaningful comparisons to be made. For example, someone might say that they have published several articles or presented at several conferences, but without knowing the kind of journal or conference they're talking about, it's hard to really gauge how much weight admissions committees gave to this aspect of their application. There is a difference between publishing in JASA or Annals of Statistics vs. an undergraduate journal, between presenting at the top conference in a particular field of statistics vs. presenting at a poster presentation at one's university/college or summer research institution, and between being first author vs. some other author. Similar things can be said about GPA, coursework, and academic awards with respect to rigor. In addition, letters of recommendation can play a big role in the process, but people often describe their letters simply as "very strong," "probably strong," "decent," etc., which isn't very descriptive at all.

I think that a better way to get a sense of how you will fare in the admissions process is to talk to people in your social/academic circles, particularly your professors and any friends who have gone through the process already. Since they are presumably more familiar with your work and with the admissions results of students coming out of your university/college, they are probably in a better position to gauge how you might do.

Second, while I agree that there is some variation in the process, I don't agree that "safety programs really don't exist and lower ranked programs (particularly smaller ones) can be just as hard to get into as large high ranked programs." Of course, I would encourage applicants not to take admission to any particular program as a given. But I think it is fair to say that someone who is competitive for Berkeley or Stanford--assuming that we can know such a thing--probably does not need to apply to as many lower-ranked programs as someone for whom said lower-ranked programs are a match. I think that within broad tiers of schools--say, top 1-30, top 30-50, top 50-100, etc. (I'm just making up the numbers here)--a single applicant can expect some variation in results due to factors like research fit and sheer randomness, but it seems to me that there are clear trends across/between these vaguely defined tiers. In other words, someone who is a match for top 1-30 schools would probably fare better among those schools than someone who is a match for top 50-100 schools. (I seem to recall that some people, historically, would disagree with me about the variation within tiers; they think that if someone is admitted to one top 5-10 school, they will probably be admitted to all or most top 5-10 schools. But I think that these people would agree even more with me, then, about the presence of trends across/between tiers.) Of course, people do sometimes get into their reach schools and rejected from their safety schools; my view is just that it is still meaningful to categorize one's list of schools into reach, match/target, and safety schools, contrary to what was implied in the OP, if the categorization is done well and doesn't just split hairs between, say, the top 5-10 programs. Also, this is only tangentially related, but one should note that acceptance rates are not necessarily a good indicator of admissions competitiveness, since there's a good deal of self-selection among applicants to these schools.

2

For the first point if you have access to professors who are knowledgeable then yes listen and seek their opinions but a lot of people I think don't have that.  The profiles are generally vague on research experience which I think made it really hard to tell for myself personally since I didn't have a lot of knowledgeable professors who had sent their students to stats PhD program and it's hard to judge how good research experience is.  On the GPA/GRE and coursework side it's a lot easier to judge.   And comments on rec letters are basically pointless and nobody really nows what a "good sop" is and what it means.

What I meant about there are no safety schools is that from a US undergraduate prospective safety schools are places where it is highly likely,  say 75%, you will get into them.  But graduate safety schools don't have that large degree of security imo.  I totally categorized my programs into reach, match, and safety but I also knew that I could not get into all my safety schools (even though I actually did but other's don't).   That's what I was trying to say, it that is almost nothing like US undergraduate admissions,  there are no places with automatic acceptance and safety school rejections happen.  If they are a Stanford/Berkeley level candidate a "safety school"  for them might be like a top 20 vs a 50-100 for most people.  For another applicant at the 50-100 level Missouri might be a reach,  where Missouri was a safety school for me.  It's all relative. It's also true that self-selection does occur and acceptance rates are a bad judge because you don't know what they are in most cases.  Even if they tell you how many applications and how many spots,  without a yield number you don't know how many offers are made.

[Stat Assistant Professor]

I would say to all prospective applicants to ultimately choose the program that you feel is the best fit for you personally. This includes factors like program size (some people thrive better in a smaller dept), location, and "culture" of the department. These are highly personal and it's perfectly fine to prioritize one thing over another (like location, for example).   

I would also tell prospective applicants to not be overly concerned with rankings. It would be wrong to say that some people won't make a snap judgment based on your pedigree. But if you want to go into industry, it probably isn't that important. If you want to go into academia, on the other hand, the PhD advisor matters the most, and the most recent (postdoc/VAP) appointment and publication record are what receive the most consideration in hiring. Nobody is going to hire someone who got their PhD from a top school, but who has no publications. 

For reference, my department (a reputable stats dept) has hired two new faculty this year who got their PhDs from University of California-Santa Cruz (UCSC) and University of Cincinnati. These are excellent schools, but their statistics PhD programs are not considered "top tier" (I'm not sure if they are even ranked by USNWR). However, these new incoming Assistant Professors DID do prestigious postdocs and were quite productive during their time there (publishing their postdoc work in top journals), which gave them an edge in the hiring process. In the current job market, it's very difficult to get a tenure-track job without doing a postdoc (it does happen but the vast majority of aspiring academics in stat/biostat need to do a postdoc now). 

[BayesianLove]

2 hours ago, Applied Math to Stat said:

I would say to all prospective applicants to ultimately choose the program that you feel is the best fit for you personally. This includes factors like program size (some people thrive better in a smaller dept), location, and "culture" of the department. These are highly personal and it's perfectly fine to prioritize one thing over another (like location, for example).   

This sounds like very solid advice, but it is very hard to know about potential advisors and the culture of the department. Although it is reasonably easy to gather some hard measures on the performance of the Faculty (like checking their recent work and see if they're active and publishing on top journals), I'd say that for most incoming students it is very difficult to know whether faculty with similar interests are taking students, whether they're cool people and supportive with their students, and overall whether the department has a healthy culture.

[Stat Assistant Professor]

30 minutes ago, BayesianLove said:

This sounds like very solid advice, but it is very hard to know about potential advisors and the culture of the department. Although it is reasonably easy to gather some hard measures on the performance of the Faculty (like checking their recent work and see if they're active and publishing on top journals), I'd say that for most incoming students it is very difficult to know whether faculty with similar interests are taking students, whether they're cool people and supportive with their students, and overall whether the department has a healthy culture.

That's true. If feasible, I would recommend visiting the programs and learning more about them firsthand. Or barring that, reaching out to current PhD students and asking about their experiences. I think completion rate and job placements are very pertinent information to consider as well (usually job placements data will also list the PhD advisor). Institutional reputation and rankings ARE also important, but not as important as doing good research and getting a good postdoc (if you want to go into academia). So you need to make sure you are at a place that is conducive to getting research done. 

[BayesianLove]

Are quals part of your decision? I've found that coursework and quals vary widely among schools:

• At one school, quals include only two courses on probability theory and statistics (i.e. Casella & Berger).
• NCSU includes two courses on probability theory and statistics (i.e. Casella & Berger) plus two courses on statistical methods (mostly linear models, ANOVA, etc). (ST 701, 702, 703, and 705).
• Then there's crazy Iowa State University with an exam covering eight courses, including stuff more advanced than Casella & Berger as well as measure theory. (https://stat.iastate.edu/phd-requirements).

I would assume that less coursework and a reasonable qualifying exam means that you can focus on research sooner. Isn't that a plus?

[statscan9]

On 2018-03-12 at 9:36 AM, BayesianLove said:

Are quals part of your decision? I've found that coursework and quals vary widely among schools:

• At one school, quals include only two courses on probability theory and statistics (i.e. Casella & Berger).
• NCSU includes two courses on probability theory and statistics (i.e. Casella & Berger) plus two courses on statistical methods (mostly linear models, ANOVA, etc). (ST 701, 702, 703, and 705).
• Then there's crazy Iowa State University with an exam covering eight courses, including stuff more advanced than Casella & Berger as well as measure theory. (https://stat.iastate.edu/phd-requirements).

I would assume that less coursework and a reasonable qualifying exam means that you can focus on research sooner. Isn't that a plus?

Quals actually influenced my decision in the opposite way of what you said. I picked Toronto, which has 3 quals with 1 dedicated to measure theoretic probability, over CMU, which just has a data analysis project. I felt that this showed me Toronto was a better fit for my rigorous theoretical interests, so I think it's really a personal preference on what you want to study.

[Bayesian1701]

If quals are at the end of the first year then I don't think it would matter that much for spending more time on research.   Harder quals (at the end of the first year) might indicate that hard material is covered quickly giving you possibly a better background for research than less difficult coursework.  It's a factor for me but hard quals aren't a deal breaker unless I think I can't pass and run the risk of being kicked out.  It definitely does vary but in most cases, I don't think its common to be kicked out because of quals.   Duke's quals (first-year exam) looked tough but during my interview, my interviewer said no one actually fails them,  but at Virginia Tech most people fail.     At University of Texas, they said their quals were difficult but passing was more determined by whether or not you were ready to do independent research and not by getting certain grade.   Baylor and Mizzou seemed to have a similar philosophy to UT, as well as my undergraduate institution's statistics quals for their program. 

 

[Stat1996]

@Bayesian1701 Did Virginia Tech offer any reasoning as to why most people fail, or give you any more information about quals? I'm considering their program but I don't want to go if almost everyone fails.

[Bayesian1701]

7 minutes ago, Stat1996 said:

@Bayesian1701 Did Virginia Tech offer any reasoning as to why most people fail, or give you any more information about quals? I'm considering their program but I don't want to go if almost everyone fails.

Their response was that they can't find enough qualified students so they have to admit subpar ones they know will not complete the degree.   I explicitly asked about this after they told me the completion rate was approximately 30% which shocked me because everywhere else I know has one easily double that so I was curious about why  so many people didn't complete.  They said I shouldn't worry about it but I had better offers at better programs so I wasn't willing to risk it personally.   Looking at your profile I would guess that they aren't worried about you completing and if they gave you the fellowship I really wouldn't worry about it.

[DJ3Sigma]

On 3/13/2018 at 4:19 PM, Bayesian1701 said:

Their response was that they can't find enough qualified students so they have to admit subpar ones they know will not complete the degree.   I explicitly asked about this after they told me the completion rate was approximately 30% which shocked me because everywhere else I know has one easily double that so I was curious about why  so many people didn't complete.  They said I shouldn't worry about it but I had better offers at better programs so I wasn't willing to risk it personally.   Looking at your profile I would guess that they aren't worried about you completing and if they gave you the fellowship I really wouldn't worry about it.

Wow haha I can't believe they are telling people that their students are sub par.

On 3/6/2018 at 3:21 PM, Bayesian1701 said:

 

Admissions - What to Expect When you are Expecting to go to Graduate School

TL;DR Expect a lot of waiting.  

I am going to be totally and completely honest with you about my experience applying for graduate school.  For undergrad I was admitted two weeks after I applied and had an departmental scholarship two months after that.  I knew I was going there because it was my first choice and I had automatic acceptance.   So I have never not know what I was doing next.  This whole process of not knowing where I am going to be in Fall 2018 for over a year now has been hard.  I still don’t know where I will go.   I have watched results tell me that I was waitlisted and rejected before I knew officially and checked admissions portals multiple times a day.  To protect my sanity and prevent me for spending my entire life on the gradcafe I blocked it from 9-5 every day on my laptop.  I highly recommend it.   I have cried more than I ever have in my life and have questioned myself almost daily.   My impostor syndrome has been awful.  I have doubted whether or not I am ready to jump into a PhD program at 21.   I have questioned why programs admitted me and felt intimidated by the people at visits who already have masters degrees.  I have read all the profiles dozens of times and knew that most domestic students with my level of research experience typically do well but I didn’t think I would do as well as I did.  I have questioned if I really do know what I want to do and whether or not being able to pursue my research is a deal breaker.  I have struggled over how to decide and what qualifies as a deal breaker.   If you had/have any of these feeling you are not alone.

 I do not understand why they cannot be more or as efficient as the undergraduate applications. For undergraduate everything was known for most schools by February and yet it is late in March and still people are waiting on decisions from programs. If the delay is in making sure that they include the applicant's name in the rejection letter then I would prefer they send out rejections like spam mail and only include "You are the weakest link!" in the subject line.

[GoPackGo89]

27 minutes ago, DJ3Sigma said:

Wow haha I can't believe they are telling people that their students are sub par.

 I do not understand why they cannot be more or as efficient as the undergraduate applications. For undergraduate everything was known for most schools by February and yet it is late in March and still people are waiting on decisions from programs. If the delay is in making sure that they include the applicant's name in the rejection letter then I would prefer they send out rejections like spam mail and only include "You are the weakest link!" in the subject line.

I did not go through the undergraduate admissions process (or at least I knew I was guaranteed admission transferring from a CC) but I suspect with PhD programs there is a lot of moving parts to ensure students are fully funded. Some schools juggling tons of students moving them on and off training grants are just better at it than others. I could be wrong but I also suspect most if not all respectable PhD programs have at least sent out initial offers and for the rest of their pool it comes down to how quickly offers are accepted/declined. 

Edited March 21, 2018 by GoPackGo89

[DJ3Sigma]

7 minutes ago, GoPackGo89 said:

I did not go through the undergraduate admissions process (or at least I knew I was guaranteed admission transferring from a CC) but I suspect with PhD programs there is a lot of moving parts to ensure students are fully funded. Some schools juggling tons of students moving them on and off training grants are just better at it than others. I could be wrong but I also suspect most if not all respectable PhD programs have at least sent out initial offers and for the rest of their pool it comes down to how quickly offers are accepted/declined. 

@GoPackGo89 I understand that it is more complicated. When it comes to funding should not all of these programs have an estimate of the kind of funding they can offer for the following year? They should at least have some prediction interval for the amount of funding available. I would only see funding being complicated if the program was restructured somehow or if this is the first year the program is being offered. I would also say that by the time the end of January comes around basically every school has sent out admissions letters to their top applicants. However, the system in place is inefficient because as you mentioned they must wait for applicants to make their decisions. But how can an applicant make a decision if not all schools have responded to said applicant? This could be solved if schools released information at the same time and were more forth coming with information. If the schools agree on releasing all the information at the same time then they could allow the applicants at most a month to decide which school they would like to attend. By the time that the first applicant has been notified of acceptance the schools have information on everyone that applied to their program and so at the same time rejections should be going out to the majority of applicants or at least if you are going to be placed on a waiting list you should be notified of the number of spots that they have yet to fill and the amount of people in front of you on the waiting list. You are most likely right in that I am oversimplifying this issue because this process is not new and I would hope that much thought has been placed into making sure it is as efficient as possible.

[Bayesian1701]

As promised:  Texas A&M (TAMU)

I paid for a ticket but I will be reimbursed.   This was the shortest visit I had (only about 6 hours) but I did an apartment hunting  and city tour trip earlier during spring break.   Although it is the largest department I visited (40 PhD students and over 500 online MS students) it didn’t feel that way since you they don’t have a lot of in town students.  A plus of their huge online program is that you have the option to be a Technology Teaching Assistant and record the lectures for the online classes and help manage the remote Q&A sessions for a higher stipend.   While they are probably more known for their Bayesian group they still have plenty of classical statisticians.  They revamped the quals process to where if you get at least 3 As and 2 Bs you don’t have to take the qualifying exam.  They do have a rough first year which includes Measure Theoretic Probability and a Bayesian class on quals.  Delaying quals by a year might be an option if you don’t have a Casella & Berger sequence (they offered me the option) Most of there students are international and/or come in with a masters but they obviously still accept domestic students straight out of undergrad (me).  They have demographical and other information here.