Profile Evaluation - Stats PhD Fall 2022, Atypical Resume

[Leif Drewdal]

Hi all. I am coming into this from a fairly unique path and I would be very grateful for any opinions on how crazy or not I am to pursue a PhD in statistics. I absolutely loved math in HS, and was always told I should be an engineer. That's how I went into college and I kind of got tunnel vision towards getting an eng. degree. Courses in the math department seemed intriguing, but with the time commitments of engineering I focused on just taking the ones I needed. 

 Out of university I worked a year at Carnegie Mellon writing reduced-order battery models which involved some data processing and monte carlo experiments. I touched on statistical methods from an applied side. I have since gone to work as a mechanical engineer in hydropower since 2019. I enjoy building things to an extent, but I spend my intellectual free time reading and trying to understand election models, pandemic models, sports performance models, etc. I still love math, but it has taken me this long to realize it's an option to actually study it. I am taking a general stats/probability course at the local university because I have no formal stats instruction yet. 

I would very much like to pursue a PhD in statistics. I think research driving scientific models would be the most impactful way I could contribute to society. I intend to apply for Fall 2022. I know my undergraduate course load is weak and my current job is not in data science/stats, but I'd like to believe I have something to offer in practical experience and perspective. Am I crazy to think I could switch gears like this? Thank you to anyone willing to share their thoughts.

Undergraduate institution: Princeton University

Major: Mechanical and Aerospace Engineering

GPA: 3.82/4.0

Relevant courses: 

Multivariable calc (A-) ; Algorithms and Data structures (A-) ; Linear algebra (A-) ; ODEs (A-) ; Math in ENG II [PDEs/Complex analysis] (B+)

My remaining engineering coursework was mathematically technical, but the above are the only math-specific courses.

Course plan to take in 2021: Currently enrolled in Engineering Stats at local University. Plan to take a course in R this summer.

Coding: Python, Julia, Matlab

Research experience: 1. 2-semester undergraduate senior project in fluid mechanics. 2. 1 year of data-based Li-Ion battery modeling at Carnegie Mellon as post-bach

Work Experience: Mechanical Engineer at water resources/hydropower company (since AUG 2019)

Letters of Recommendation:  I expect to get  one from the professor under whom I worked at CMU. The second would likely come from my supervisor at my current job.

GRE: Not taken(will take it in 2021)

GRE math: Not taken(will potentially take it in 2021)

Honors:  Magna cum laude, Tau Beta Pi

type of student: Domestic white male

Type of PhD: Statistics

Research interests: Probabilistic modeling (climate forecasting, epidemic, biostats)

School I am planning to apply:  Penn state, UVA, Northeastern, UPenn  - tentative list at this point

I really don't know what schools I am competitive in because this is not my field. I have also seen on here people referencing "top programs" vs "20-40 ranked" or similar. I'd be curious to know if this is common knowledge or if there is a specific ranking (USN&WR?) that these types of comments refer to.

[bayessays]

Definitely not crazy.  You went to Princeton with a good GPA, and I assume you can score very high on the math section of the GRE (the general test - you don't have to take the subject test).  Your math background is certainly weaker than ideal, with the biggest flaw being that you haven't taken real analysis, probability, or any upper-level math with proofs, which is going to make it very hard to get into top programs.

The US News rankings are the rankings that everyone is referring to, so definitely look at that.  Your current schools are all over the place in terms of ranking -  UPenn is an elite program top 10 that is unrealistic.  Penn State is also a very very good (top 20-30), and I think your lack of math would keep you out of schools like this but it would be reasonable to apply. UVA is a solid program (top 60ish), and you probably have a shot there.  Northeastern, to my knowledge, does not have a statistics PhD.

I've seen quite a few people sign up for https://netmath.illinois.edu/college/math-444 and take real analysis online.  To be clear, real analysis is not strictly necessary but because your profile is so good in other respects, you could go to a much higher-ranked program if you took a little more math.  Depending on your goals, this might not be necessary - if you want to go into industry, program ranking matters much less.

[Leif Drewdal]

This is incredibly helpful, thank you so much! My preference would be academia or government research, so I expect it will be worth taking Real Analysis.