Fall 2020 Statistics Applicant Thread

[ENE1]

I'm trying to decide whether to attend Duke's open day. I'm coming from Australian and I didn't realise until now how far away it is and how expensive the (internal US) flights are to get there. 

I've talked to faculty but I wanted to get people's advice on how rigorous/tough the courses are there? I talked to a Duke student who said they assume no stats knowledge and he didn't take the classes because he didn't think they were worth his time. 

I'm actually quite interested in building up my stats knowledge so I have a solid background across the field, so good, rigorous courses are important to me.

Any comparison to U of Washington/Harvard's coursework would be helpful since they are the other programs I'm considering.

Also, any other feedback/advice on Duke would be much appreciated. On the off chance that anyone here is a rock climber, I'd be interested in hearing what's available around Durham in terms of rock and plastic.

[casummit]

2 hours ago, ENE1 said:

I'm trying to decide whether to attend Duke's open day. I'm coming from Australian and I didn't realise until now how far away it is and how expensive the (internal US) flights are to get there. 

I've talked to faculty but I wanted to get people's advice on how rigorous/tough the courses are there? I talked to a Duke student who said they assume no stats knowledge and he didn't take the classes because he didn't think they were worth his time. 

I'm actually quite interested in building up my stats knowledge so I have a solid background across the field, so good, rigorous courses are important to me.

Any comparison to U of Washington/Harvard's coursework would be helpful since they are the other programs I'm considering.

Also, any other feedback/advice on Duke would be much appreciated. On the off chance that anyone here is a rock climber, I'd be interested in hearing what's available around Durham in terms of rock and plastic.

I'm familiar with Duke biostats but not so much Duke stats so I'll let other people help answer those questions for you. 

I will say that rock climbing in the Triangle is pretty popular - we have quite a few gyms within 20 minutes of Duke and some great mountain climbs ranging from 1 to 3 hours away.  I would say that there's definitely a climbing culture here! (Feel free to message me if you want to know any other specifics about the Triangle area)

[rfan]

2 hours ago, ENE1 said:

I'm trying to decide whether to attend Duke's open day. I'm coming from Australian and I didn't realise until now how far away it is and how expensive the (internal US) flights are to get there. 

I've talked to faculty but I wanted to get people's advice on how rigorous/tough the courses are there? I talked to a Duke student who said they assume no stats knowledge and he didn't take the classes because he didn't think they were worth his time. 

I'm actually quite interested in building up my stats knowledge so I have a solid background across the field, so good, rigorous courses are important to me.

Any comparison to U of Washington/Harvard's coursework would be helpful since they are the other programs I'm considering.

Also, any other feedback/advice on Duke would be much appreciated. On the off chance that anyone here is a rock climber, I'd be interested in hearing what's available around Durham in terms of rock and plastic.

I would say that the courses are rigorous, but that there are very few requirements. Until the last year or so, technically there were no class requirements (although most students still took a majority of them). The qualifying exam after the first year is based on STA 711, STA 721, STA 611, STA 831, and STA 732.

 

The courses tend to be decently rigorous. STA 711, STA 601, and STA 721 are taken in the Fall. STA 732, STA 831, and a case studies course are taken in the spring.

STA 711 is a normal 1 semester measure-theoretic probability class (using A Probability Path). Here is the most recent course website: https://www2.stat.duke.edu/courses/Fall19/sta711/

STA 601 is an introduction to Bayesian Statistics, though it is better if some Bayesian Statistics is already known. It follow's Peter Hoff's book and is taught by Alexander Volfovsky. 

STA 721 is a theory based linear models class (mostly based off of Plane Answers to Complex Questions) that has a strong Bayesian emphasis. Here is the most recent class website: https://www2.stat.duke.edu/courses/Fall19/sta721/

STA 732 is the only "frequentist" theory course. Classical statistical theory is definitely minimally developed in the curriculum.

STA 831 is a follow-up course to STA 601 and is purely Bayesian. There is no textbook the class follows.

So the courses are difficult enough, but many students don't learn much non-Bayesian statistics, and some basic topics such as study design or generalized linear models are almost completely ignored (and neglected often in the elective offerings). Students going to Harvard or Washington (I know someone going to the latter) are taught far more classical statistics topics but don't have as much formal Bayesian coursework.

 

[IdoMath]

19 hours ago, captivatingCA said:

Are they requiring you to respond by Feb 28 for funding? Because that would be violating the April 15 Resolution which Indiana University is a part of. I'm not familiar with the department, but talking to students and alumni should give you a better idea of what it's like there.

Yes, they need a decision before end of Feb. This is my first time applying to US universities, so I was not aware of this resolution. I will contact some current students. Thank you.

[captivatingCA]

@ENE1 I climb too! I've been to a bunch of different gyms on the east coast, so those are my main frame of reference. I visited Triangle Rock (I can't remember which branch) and it was definitely one of the better gyms I've been to. I can't speak much to the rope climbing, but the bouldering was great! I'm pretty sure the gym was in a former big box store, so there was a LOT to climb and a lot of variety. They grade pretty hard too, much harder than most gyms I've been to. I only spent a few hours there, so I can't speak to how often they reset or the vibe of the community. The few people I did meet were really nice though. 

[bayessays]

@ENE1 I suppose it depends on your goals, but I wouldn't be too concerned about Duke's courses not giving you a solid foundation (although obviously you should probably not go to Duke if you aren't ok doing bayesian stuff).

UW has some pretty heavy course requirements and I assume you're comparing it to that, but looking at the syllabi I don't think you'll come out with a much different knowledge base.  At both schools you'll take measure theoretic probability and have the option to take stochastic processes.  Both schools will obviously have you master Casella/Berger level theory material and take basic linear models classes.  Any stat theory class beyond the Casella/Berger level is essentially a special topics course.  A lot of programs will teach you basically more theoretical stuff about sufficency, exponential families, etc, while some will have their higher-level theory classes be about bootstrap stuff, while Duke will focus mostly on Bayesian theory. Duke seems to not have an entire class devoted to large sample theory, likely because Bayesian statistics relies on computation more. Insofar as there is a "core knowledge base", all the programs you've been accepted to will provide it.

I've had the same line of thinking in the past regarding wanting rigorous coursework, but I don't think that should be a major contributor to your decision.  You're going to forget 90% of your coursework in a year if you're not using it in your research.  For instance, the poster above mentions that Duke doesn't teach much about GLMs.  As someone who went to a program that had multiple semester-long courses on them, that's crazy to me.  But 5 years later,  I don't remember much about the technical details of GLMs beyond what I could read in 5 minutes of googling.

Since all the programs will give you a solid base,  I think flexibility to take courses you want and to do research you enjoy are way more important.  The syllabi for Washington's classes are all up online, so I don't think going there based on coursework you can access for free is a good decision.

[Geococcyx]

Would we consider TPE2 more advanced than Casella and Berger?  A friend of mine at Duke says that their inference class is straight from TPE2 rather than Casella and Berger, and I'm not really familiar enough with both books to judge difficulty.

[bayessays]

11 minutes ago, Geococcyx said:

Would we consider TPE2 more advanced than Casella and Berger?  A friend of mine at Duke says that their inference class is straight from TPE2 rather than Casella and Berger, and I'm not really familiar enough with both books to judge difficulty.

Yeah it's more advanced than Casella and Berger but a more in-depth coverage of a lot of the same material (UMVUE, sufficiency, etc).  It's hard to tell without looking at the homework, but stat 732 is pretty similar to what a lot of people would probably take as their their first post-CB theory course, followed by a course in large sample theory.  Some top programs just skip having a CB course and learn the stuff at a higher level to begin with.  I think a lot of programs are changing their curriculum and view spending multiple semesters on classical theory as irrelevant to most statisticians, and are changing the topics covered in advanced theory courses more to things like bootstrap, empirical processes, random matrix theory, etc.

[Stat Assistant Professor]

2 hours ago, bayessays said:

@ENE1 I suppose it depends on your goals, but I wouldn't be too concerned about Duke's courses not giving you a solid foundation (although obviously you should probably not go to Duke if you aren't ok doing bayesian stuff).

UW has some pretty heavy course requirements and I assume you're comparing it to that, but looking at the syllabi I don't think you'll come out with a much different knowledge base.  At both schools you'll take measure theoretic probability and have the option to take stochastic processes.  Both schools will obviously have you master Casella/Berger level theory material and take basic linear models classes.  Any stat theory class beyond the Casella/Berger level is essentially a special topics course.  A lot of programs will teach you basically more theoretical stuff about sufficency, exponential families, etc, while some will have their higher-level theory classes be about bootstrap stuff, while Duke will focus mostly on Bayesian theory. Duke seems to not have an entire class devoted to large sample theory, likely because Bayesian statistics relies on computation more. Insofar as there is a "core knowledge base", all the programs you've been accepted to will provide it.

I've had the same line of thinking in the past regarding wanting rigorous coursework, but I don't think that should be a major contributor to your decision.  You're going to forget 90% of your coursework in a year if you're not using it in your research.  For instance, the poster above mentions that Duke doesn't teach much about GLMs.  As someone who went to a program that had multiple semester-long courses on them, that's crazy to me.  But 5 years later,  I don't remember much about the technical details of GLMs beyond what I could read in 5 minutes of googling.

Since all the programs will give you a solid base,  I think flexibility to take courses you want and to do research you enjoy are way more important.  The syllabi for Washington's classes are all up online, so I don't think going there based on coursework you can access for free is a good decision.

+1 to this. Also, if you look at UC Berkeley's Statistics PhD requirements (a very top program, if I do say so myself), it seems like students there get a choice of taking two out of three sequences: theoretical statistics, applied statistics, and probability theory. So it seems as though one could actually go through UCB's Statistics PhD program without having learned even measure-theoretic probability theory (which, while good to know, probably isn't super-relevant to everyone's research --if you're not a probabilitist, you can probably learn enough of it to get by without having taken a whole course on it). And UCB also has no written qualifying exams. 

Doing original research truly is the primary focus of the PhD. For your PhD research, you are mainly teaching yourself the things needed for your research. One cannot reasonably expect to learn *everything* there is to know through classes anyway. I kept a lot of my notes from my classes but I barely look at them, because I can just Google what I need to know if it comes up (e.g. what the expectation of a quadratic form is, various norm inequalities, etc.). Plus, I will say that although my research focuses on high-dimensional Bayesian statistics, it was easy enough for me to pick up on the analogous frequentist methods/theory (like LASSO, elastic net, etc.) once I had enough experience reading and understanding academic papers. I'm sure that will be similarly the case for you.

[honguyen94]

Hi, anyone apply to Duke M.S in Statistical Science program? 

[dberlind]

Is anyone else surprised we have not heard from Penn yet?

[stochasticworld]

6 hours ago, dberlind said:

Is anyone else surprised we have not heard from Penn yet?

I emailed Upenn's admission office last week and they say they are about to finalize decisions this week.

[AM61]

15 hours ago, honguyen94 said:

Hi, anyone apply to Duke M.S in Statistical Science program? 

Yeah, I actually just got my decision notification today.

[Cofcof]

I guess i will get a rejection from all schools I applied. I think this is due to my institution is unknown and I am a international. I think getting an MS in math will make my profile a little bit stronger even though I have no idea which school to apply because most of deadline is over. Can you give me any advice for me? 

[statapp]

Can people receiving offers share the funding details...it will be useful for people who are in the waiting list to decide whether to consider the place as well as helpful in getting a better idea of the funding scene elsewhere.

[JD009]

4 hours ago, statapp said:

Can people receiving offers share the funding details...it will be useful for people who are in the waiting list to decide whether to consider the place as well as helpful in getting a better idea of the funding scene elsewhere.

My offer letter from UIUC says: We also are offering a teaching assistantship during the academic year. The assistantship includes a waiver of tuition and the majority of charged fees. The stipend for the current (2019-2020) academic year is $20,589.48 ($2,287.72 per month for nine months). Summer support is available on a limited basis with priority going to first year PhD students. In addition to this support, we also offer travel support for our PhD students for research-related purposes, e.g., to present at conferences.

[IdoMath]

Has anyone conducted the interview with University of Florida (professor Xiang-Yang Lou)?

[Ginger?]

If I haven't heard from Duke Statistics or Colorado State Stats does that mean I'm rejected? It seems like they've sent out their decisions already

[DanielWarlock]

2 hours ago, Ginger? said:

If I haven't heard from Duke Statistics or Colorado State Stats does that mean I'm rejected? It seems like they've sent out their decisions already

Technically, there is still "hope" but my feeling is that once you do not receive offer and others do, you are waitlisted. You see, Duke sent nearly 30 offers (I saw it from their mailing list). For example, I'm just officially confirmed to be waitlisted by Michigan but I knew that a few days ago-- some people here even argued with me about this back then. For me, same thing is now going on with UFlorida, Minnesota and Columbia where others already got offers/interviews while I didn't. I'm hopeful for Columbia because there is just one guy who got interview and it is possible that candidates will be admitted without an interview (when they deem interview to be unnecessary for such candidates for example). But hope is not very, very high at this point.

[zscki]

Yet to hear from 5/6 of my programs...a bit unnerving. Anyone have any intuitions about UCLA, UW, or UT Austin? The other programs are NYU and USC data science, which I assume not very many people here applied for.

[bayessays]

31 minutes ago, DanielWarlock said:

Technically, there is still "hope" but my feeling is that once you do not receive offer and others do, you are waitlisted. 

This is incorrect and needs to stop being repeated.  Some programs release decisions over a period of weeks for a variety of reasons.  This is demoralizing to people whose applications haven't even been looked at yet and also gives false hope to people who are going to be rejected. Michigan probably has 100s of people they haven't sent rejections to yet and those people are not all "on a waitlist."

I'm not sure if this is a language barrier, but a waitlist is a short list of people who may get acceptances depending on availability. The only wait list many applicants are on now is the waiting list to be sent a rejection email.

31 minutes ago, DanielWarlock said:

For example, I'm just officially confirmed to be waitlisted by Michigan but I knew that a few days ago-- some people here even argued with me about this back then

We argued with you because you are wrong and claiming you know a process that you don't understand.  You happened to get a waitlist from Michigan - lucky guess.  Again, 100s of people did not get such an email, are waiting for their rejection letters and are not on a waitlist.

[ENE1]

35 minutes ago, bayessays said:

Some programs release decisions over a period of weeks for a variety of reasons

To back this up, I got an email from Penn State yesterday with an offer despite a bunch of offers being sent out more than a week ago.

 

 

@dberlind I’m not sure if you are referring to Upenn or Penn State but see above if the latter.

Edited February 19, 2020 by ENE1

[ENE1]

 

@bayessays @rfan @Stat Phd Now Postdoc

@bayessays @rfan @Stat PhD Now Postdoc thanks for your replies. You raise some good points - a PhD is mostly about research and pretty much all the top programs will have good syllabi.

@casummit thanks for the info - that sounds great. @captivatingCADo you know what the gyms are like in Cambridge/Boston by chance?

I think I will be attending visit day now. It's going to be a really hard decision.

[rfan]

1 hour ago, DanielWarlock said:

Technically, there is still "hope" but my feeling is that once you do not receive offer and others do, you are waitlisted. You see, Duke sent nearly 30 offers (I saw it from their mailing list). For example, I'm just officially confirmed to be waitlisted by Michigan but I knew that a few days ago-- some people here even argued with me about this back then. For me, same thing is now going on with UFlorida, Minnesota and Columbia where others already got offers/interviews while I didn't. I'm hopeful for Columbia because there is just one guy who got interview and it is possible that candidates will be admitted without an interview (when they deem interview to be unnecessary for such candidates for example). But hope is not very, very high at this point.

Not to be confrontational, but if you're nearly as argumentative in person as with this internet persona, it might be hard developing the strong connections with your fellow PhD students that are often really beneficial to having a positive PhD experience. You have a terrific sounding background and got in to some really great schools so major props to you on that, but as a current PhD student in one of those schools I'll say that the confrontational style is probably best toned down a tad! We're all going through this wild, challenging experience, and I think helping each other out and building each other up is a good perspective to have. ?

To the poster asking about Duke's admission chances, I do know that they *plan* on having another wave of acceptances. This past class was very large compared to their expectations, so I wonder if there is concern about having too large of a class this go around. Anywho good luck to all y'all that are waiting to hear back!

[cyberwulf]

Here's the thing: It's much easier for programs to make "no decision" on applicants who aren't admitted in the first round and aren't obvious rejections than to come up with an official waitlist of people who are first in line to be admitted if first round offers decline. So, many schools keep a bunch of applicants hanging, even though most of these aren't really in the running for admission (editorial comment: I think this is unfair to students, but it's the reality). If you haven't heard from a school, offers don't seem to be trickling out gradually, and you haven't been notified that you're on an official waitlist, then you should probably be prepping yourself for bad news.