Incoming junior looking to do a Statistics PhD at some point. How are my credentials so far?

[notafrequentist]

Undergrad:  Relatively large public university in Pennsylvania, double major in Statistics and Mathematics.

 

GPA:  Currently a 3.377 (due to a mediocre freshman year), with a 3.625 for last semester.  Statistics GPA is 3.52,  and 3.60 for both Statistics and Math.  As for individual courses:

 

Calc 1:  B-

Calc 2:  Transfer credit from a community college for money reasons, but an A there

Calc 3:  B+
Intro to Theoretical Mathematics (covers set theory, metric spaces, Cauchy sequences, etc.):  A-

Linear Algebra:  A

Numerical Linear Algebra:  TBA.  Taking it now.  It's a scientific computing-based approach to linear algebra and we use MATLAB for everything.

Ordinary Differential Equations:  Not sure yet, but I'll be taking it this summer in a few weeks.

Java:  Withdrew due to a terrible professor in a previous semester, but taking it now and doing well so far.  It's required for the math major.

 

Intro to Statistics:  A+

Regression:  A
Intro to Probability:  B

Intro to Mathematical Statistics:  A-

 

 

In the fall I'm taking Combinatorics, Data Mining (graduate business course), Applied Nonparametric Statistics, Applied Time Series, and Data Structures, although I'm considering dropping Nonparametrics for a second analysis class on Calculus so I don't have to put that off until the Spring.  The issue would be there aren't that many applied courses offered for a specific requirement and Nonparametrics is only offered once every two years in the fall.  For the future, the other courses I'd be taking are:

 

Intro to One-Variable Calculus

Physics 1 (required for math major)

Abstract Algebra

Advanced Calculus 1 (uses Rudin's book)

Advanced Calculus 2

(Possibly) Numerical Analysis

Intermediate Probability

Intermediate Mathematical Statistics

Stochastic Processes

(Possibly) Applied Probability Theory

Algorithms (for my own benefit)

 

 

Recommendation letters:  I know both the undergrad department heads for Statistics and Mathematics and have taken classes with both.  I also have a Math professor who I've taken two classes with and am taking a third in the fall.

 

 

Work:  Currently tutoring someone for an Intro to Statistics class.  I will have work unrelated to either major in the fall as someone hired by the university to fix peoples' computer problems.  This is for financial reasons.

 

 

GRE:  Not sure yet

 

 

Research:  Did a semester-long research project in a Bio course I took freshman year (when I was a Bio major) and did a Bioinformatics research project that same semester.  I did no research sophomore year, but did a week-long statistical modeling workshop through SAMSI at NCSU a few weeks ago where I modeled climate data.

 

 

What should I be looking to do now if I want to get into a really good program?  I've been drawn a lot to Bayesian statistics, machine learning, and data mining lately, but my university's Statistics department has nothing with regard to any of these and for data mining I've had to go out of department.  I'm liking some of the stuff at Stanford even though I doubt I'll get in with my current credentials, and there's a major university across the road where I've been trying to find research experience but failing.  I am also highly interested in said university, along with Duke and UW Seattle.

[wine in coffee cups]

What should I be looking to do now if I want to get into a really good program?

This will come off as obvious and patronizing, but seriously, get way more A's. The three programs you specifically identified have very, very high average overall GPAs for incoming students.

[notafrequentist]

Still useful though, so thanks!  For some reason I tend to do mediocrely in intro classes and really well in more advanced classes because the depth forces me to understand the material better.  Like I got a B in Microeconomics last semester because I couldn't figure out where to apply Calculus, which would have made a lot of the curve areas much easier.

 

But so far I have As in my summer classes and plan to keep that up into the fall.

 

Also would I be better off getting a Master's somewhere else first?

Edited June 1, 2013 by notafrequentist

[Shostakovich]

Still useful though, so thanks!  For some reason I tend to do mediocrely in intro classes and really well in more advanced classes because the depth forces me to understand the material better.  Like I got a B in Microeconomics last semester because I couldn't figure out where to apply Calculus, which would have made a lot of the curve areas much easier.

 

But so far I have As in my summer classes and plan to keep that up into the fall.

 

Also would I be better off getting a Master's somewhere else first?

 

Like wine in coffee cups said, the problem for you is that many students getting into the schools you mentioned perform really well in just about every class they take. There would really be no reason for the admissions committee to go with a 3.38 when they have a bunch of 3.8's and 3.9's, all other things equal. They do look at your application as a whole so if you could basically put up straight A's for the next 2 years or so (assuming that you're going into junior year) you might have a shot at Duke or UW. There is an applicant profile thread on this site as well as mathematicsgre.com where you could get an idea of what successful applicants are doing, and how your application might stack up against the competition at different programs.

 

Starting with a masters somewhere is a good option, where a good performance based on grades and research experience could qualify you for decent PhD programs.

[notafrequentist]

Thanks! My university has a five-year program where I'd stay one extra year and come out with my Master's. I've already applied for it and as far as I know, I think I'm the only person my year who's actually applied for it.

How much will an REU help me? Also will the fact that I'm coming from a non-Ivy or a school without an amazing math department hurt me? Finally, what about not being able to take Measure Theory because my university requires us to take 13 gen eds and I can't fit it in, not just because of that but because we're required to take two real analysis classes before touching Advanced Calculus and Baby Rudin?

Edited June 2, 2013 by notafrequentist

[Shostakovich]

Thanks! My university has a five-year program where I'd stay one extra year and come out with my Master's. I've already applied for it and as far as I know, I think I'm the only person my year who's actually applied for it.

How much will an REU help me? Also will the fact that I'm coming from a non-Ivy or a school without an amazing math department hurt me? Finally, what about not being able to take Measure Theory because my university requires us to take 13 gen eds and I can't fit it in, not just because of that but because we're required to take two real analysis classes before touching Advanced Calculus and Baby Rudin?

 

1. If you are attending a school without a good Math/Stats department and you want to do your PhD at a top program, it might be to your advantage to go do your masters at a more prestigious institution (MS programs at lower ranked institutions don't seem to be THAT respected in the eyes of PhD program admissions committees). cyberwulf explained it somewhat here as well as on other threads: 

 

2. From what I've heard, REU probably won't help too much unless you can get some papers published in a reputable journal (preferably first-authored). A lot of people get accepted with close to zero research experience, and a lot of people get rejected with a lot of research experience. But obviously still better to have experience than to not have experience

 

3. Reputation of your undergrad institution matters, but you don't have to be from an ivy or a top math department to be considered for admissions. It might be easier to get accepted coming from a top undergrad school, but I've also seen students coming from schools ranked a bit lower who put up good numbers and softs.

 

4. Not having a course on measure theory probably won't sink your application, but I heard somewhere that two courses in real analysis is somewhat standard for a PhD applicant. If you don't feel super strong on your Math background, Biostat programs seem to be somewhat more lenient on the Math requirements (and in general less competitive). It's possible to work in Bayesian/ML/data mining type of research at Biostat programs also, so it could be a good idea to take a look.

 

Good luck!

Edited June 2, 2013 by Shostakovich

[wine in coffee cups]

Thanks! My university has a five-year program where I'd stay one extra year and come out with my Master's. I've already applied for it and as far as I know, I think I'm the only person my year who's actually applied for it.

How much will an REU help me? Also will the fact that I'm coming from a non-Ivy or a school without an amazing math department hurt me? Finally, what about not being able to take Measure Theory because my university requires us to take 13 gen eds and I can't fit it in, not just because of that but because we're required to take two real analysis classes before touching Advanced Calculus and Baby Rudin?

The five-year program could be a good idea for a couple of reasons. I think one big benefit is you will be delaying applying for PhDs for another year, and so you'll have an extra year's worth of (hopefully very good) grades in advanced classes at the time of applying.

 

The other upside is that you'll have the extra year to impress potential recommenders. You'll absolutely need to have strong recommendations to be competitive. I would direct you to the chain of posts from cyberwulf and biostat_prof starting to get some ideas about how much where you go matters vs. your grades vs. your letters. I think an REU could help strengthen your references so that it's not just three letters saying "he was good at classes".

 

You might investigate where previous math/stat majors (and particularly stat master's students) from your school have ended up for stats PhDs and how you stack up to them to get a rough sense of where you might be competitive, though it's premature for serious comparisons.

[notafrequentist]

Thanks for the advice!  The thing is, very few people at my school major in Statistics.  We have Statistics and a joint major between Statistics and Economics, and we only have about ten pure Statistics majors.  Neither major requires analysis.  Also very few of us even apply to PhD programs.  My Linear Algebra professor (who's also the head of undergrad math advising) said it's basically unheard of at my university for a Statistics major to take a theory-oriented linear algebra class instead of the standard applied one like I did, and I figure it's because Statistics here is a very applied major.  The only theory required is intro and intermediate versions of Probability and Mathematical Statistics.

 

Also when do REU applications generally open up?

Edited June 2, 2013 by notafrequentist

[notafrequentist]

I e-mailed Kadane and Verdinelli because they seemed to have research interests I was interested in the most.  A major issue with e-mailing professors is I have no idea which ones actually take students for research.  Any recommendations for who to e-mail with regard to machine learning?

 

Also I'm thinking about dropping Nonparametrics for an Advanced Calculus course that covers Baby Rudin so I can take graduate analysis (aka Measure Theory) as a senior.  The only thing is I'll probably be overloading myself with all my other classes.  Is it highly necessary to take Measure Theory to get into a good Statistics grad program?

Edited June 14, 2013 by notafrequentist

[coffeehouse]

I e-mailed Kadane and Verdinelli because they seemed to have research interests I was interested in the most.  A major issue with e-mailing professors is I have no idea which ones actually take students for research.  Any recommendations for who to e-mail with regard to machine learning?

 

Also I'm thinking about dropping Nonparametrics for an Advanced Calculus course that covers Baby Rudin so I can take graduate analysis (aka Measure Theory) as a senior.  The only thing is I'll probably be overloading myself with all my other classes.  Is it highly necessary to take Measure Theory to get into a good Statistics grad program?

 

A professor Emeritus is definitely not likely to take on students for research.

[notafrequentist]

Yeah I realized that once I got deeper into the academic process and someone finally pointed out to me that an Emeritus is retired.  He did recommend me a few books to go through though, so I've been teaching myself Bayesian methodology.

[notafrequentist]

Thanks!  Any specific recommendations?  Also I just got an A- in my Numerical Linear Algebra class after a particularly difficult final.  I'm worried because it seems like it'll put me in the "good, but not good enough" category when it comes time to apply.  How much do they care about -s, considering a lot of schools don't even bother with them?

Edited June 22, 2013 by notafrequentist