Statistics PhD Program Evaluation - 2021

[Poisson_Derulo]

Hi all,

I'm planning on applying to statistics PhD programs for next cycle (entry in Fall 2021) and would love an honest assessment of my profile, with strengths and weaknesses, and recommendations on programs to apply. My interests broadly are statistical machine learning with applications to healthcare and social sciences!

Undergrad Institution: Top LAC
Majors: B.S. Mathematics and Economics
GPA: 3.97/4.0

Type of Student: Domestic White Male

GRE General Test:
Q: 170
V: 170
W: 6.0
GRE Subject Test in Mathematics: 760 (71%) - April 2019

 

Programs Applying: Statistics PhD

 

Research Experience:  

In my job since graduating college I have been working on health economics and epidemiology studies, and have a number of co-authorships on publications (although not directly related to statistics). Hoping to move away from biostatistics even though it's more relevant to my current work experience (I am aware biostats and stats departments are very similar!). I also presented an economics research paper at a well known economics conference with my thesis advisor from school since we wrote a continuation of my thesis after I graduated.

Awards/Honors/Recognitions: 

Summa cum laude, award for top economics student in graduating class.
Pertinent Activities or Jobs: 

1.5 years of research experience in health economics and epidemiology in my job.

Letters of Recommendation:

I have two math professors who know me well and said they would be happy to write me strong letters (both of them offered when I mentioned I was thinking of applying to PhD programs), one of them very well known in their field. The other is my thesis advisor from the economics department. None of them strictly from statistics department, but hopefully they are relevant!

Coding Skills: 

R, Python

Relevant Classes, Grades: All undergraduate level ?

 

Mathematics: Statistics (A); Multivariable Calculus (A); Linear Algebra (A); Discrete Mathematics (A); Mathematical Modeling and Computation (A+); Data Mining and Machine Learning (A); Financial Calculus and Probability (A); Real Analysis I (A+); Real Analysis II (A); Abstract Algebra I (B+)

 

Other Relevant: Introduction to Programming with C++ (A). Advanced Econometrics (A).

 

Comments:

1. I think that my biggest weaknesses are my lack of graduate classes (I added a math major very late in my career at the recommendation of one of my letter writers from the math department) and while I have research experience in publications in a related field, none of it is directly statistics. Also, I took the math GRE after a year out from college and was definitely rusty - I feel like it's a respectable score but am unsure if it will help or hurt my application for those that don't explicitly require it (i.e. not Stanford!) Could you guys help me get some clarity on that? I know it's a hotly debated subject (pun intended) on this forum!

 

2. One recommender said I would have a solid shot at some top 10 programs but I am not feeling as confident - what do you guys think? I will definitely have a few "shoot for the stars" schools on my list that really align with my research interests of machine learning and social science applications (Berkeley, Harvard, Michigan) but I need to add some more realistic schools I think, unless you guys feel otherwise!

 

 

Thanks you all so much for reading - I've learned a lot from going through this forum and would love your guys' input!

Edited January 22, 2020 by Poisson_Derulo

[bayessays]

I think you'd probably get into a few top 10s and you don't need to apply outside the top 20.  I don't think Michigan is a shoot for the stars school at all, that should probably be your main target range.

[Poisson_Derulo]

@bayessays - Thank you so much for your quick reply! That's really great to hear (I guess I am getting a bit too freaked out from all the qualified people I've seen on here not get into these programs). With regards to math GRE - I know there is no simple answer to this based on the amount of discussion on this thread, but in my case, do you think it's worth sending in to schools that "strongly recommend" or do not recommend at all the subject test? I think it would help me in the event my math experience was weak. I don't think it is by any means, but it's also not stellar like some applicants I see on here based on the number of advanced classes I was able to take. Do you think it would help or hurt my application to submit it to those schools?

Thank you so much, yet again!

[bayessays]

We can only speculate based on what we've seen in past results. Some people say only send it if it's 80%tile+. I think a 70 is pretty good though and I don't think it would hurt you, but it's probably in that neutral zone where it won't help a ton either. I know people who have gotten into programs like Chicago without sending their score so I don't think it's usually a big deal if domestic students don't have a subject score.

[Poisson_Derulo]

That makes a lot of sense to me, and thanks for the information. What I will probably do is try and re-take once again in April if I'm feeling up for it and see if I can break the 80 mark - if I will, I'll send it in, and otherwise likely not submit to schools that don't require it. If anybody else thinks otherwise please let me know, but I really appreciate your help on this, @bayessays!

[bayessays]

Also, I think there are definitely some reasonable reasons to avoid biostat departments (for instance, some biostatistics programs have zero opportunities to teach, or if you are only interested in theoretical statistics and probability and never want to look at real data), but if you're interested in "statistical machine learning with applications to healthcare and social sciences" then I'm not sure what you mean that you want to move away from biostatistics.  You'd be a strong candidate for Hopkins/Harvard/Washington biostat which would be some of the best places to work on these problems.

[Poisson_Derulo]

That is a good point - I definitely want teaching experience and I guess I could have been more specific about my aims. My research goals (at least in terms of application areas relating to healthcare) are more policy-focused and social science oriented as opposed to biological phenomena/systems/etc. and from my research on department faculty it seems like statistics programs as opposed to biostatistics programs have more people working on these areas. But your point is well taken and maybe I should adjust my view and take a closer look at biostatistics programs again. Thanks for the advice!

[captivatingCA]

Just to reiterate what bayessays said, I think biostats departments do a good amount of research related to your interests. From the glimpse I got at Harvard biostats, it seems like more people are working on clinical and public health problems than on biological phenomena. I heard from a few faculty who work on policy-related problems, such as environmental health and health disparities. There's also a good amount of work in understanding these problems but also in solving them iirc, which is why I imagine you're drawn to policy. In terms of methodologies related to policy and the social sciences, a good amount of faculty work on causal inference and network analysis. I don't recall hearing anyone speak specifically on healthcare though.

I think it's worthwhile to apply to a few biostats departments and talk to the faculty there if you get in. Of course you should also apply to departments that specialize in machine learning and applications to the social sciences. CMU, UW, and UMich come to mind as being great at both, but I'm sure you already have those on your radar.

[bayessays]

Absolutely agreed with above. Biostat departments are in schools of public health - if you want to make a difference in health policy, it's the place to be.