Opinions on the PhD in Mathematics with concentration in Statistics

[alemao]

My question is not related to "chances of admissions" or "where to apply", but I will add some background that might be useful for context:

Undergraduate institution: Top 25 University in Brazil

Major: Statistics

GPA: 83%/100% (different system), but the department's head said my grade was the highest ever since the program's creation 15 years ago (wowever, I don't think it is common to "graduate with distinction" or those types of honors in my country, at least not in this department)

Type of student: International white male

Graduate institution: Good R1 University but not sure if it's known for math/stats

Graduate major: Applied Statistics

Relevant courses in undergrad: Stats I and II, Probability I and II, Calc I-III, Linear Algebra, Analytic Geometry, Numerical Calculus, Introduction to Measure Theory, Stochastic Processes, Biostatistics, Inference I and II, Linear Models, Regression Analysis, Statistical Computing I and II, Multivariable Analysis, Sampling I, Introduction to Computing (C), Physics I and II, Non-parametric Statistics, Experimental Design, Risk Analysis, Statistical Quality Control, Categorical Data Analysis, Time Series, Longitudinal Analysis

Relevant courses in graduate degree (last semester): Experimental Design, Regression Analysis, Multivariable Analysis, and some electives (Statistical Computing for Biologists, Plant Breeding, Precision Agriculture). * My thesis has to do with agstats/genetics, and I do research with the crop sciences department, so I took some classes related to agronomy.

Research experience: 

- Two years funded as an undergraduate researcher for a government unit (like USDA in U.S.) during my undergrad (my P.I. was a statistician and my co-PI an agronomist).

- Two papers as co-author in very well known agronomy journal, one paper as co-author in a good agronomy journal. One paper as first author in very well known genetics journal is on review

- Other papers as co-author ongoing related to agronomy in ok/good journals

- Some experience assisting my P.I. with agricultural statistics consultancy for other departments

- Summer internship at a Fortune 500 agribusiness company

-------------------------------------------

Now about my real question:

I am about to finish my master's this year (hopefully). The thing is, for personal reasons, I won't be able to move to another city for the next 2 or 3 years, and the university where I am at right now only offers a PhD in Math or PhD in Math with concentration in Statistics. I realized that we do have excellent statisticians (lot of them came from Top 10 stats programs) in the department, although the majority are more math.

So I thought about two ideas:

- Take some graduate math courses during those 2-3 next years to "fill the gaps" (real analysis, for example, and others if possible). As you can see, my master's was not very heavy in math becase I did a lot of research, but in my undergrad I had lot of proof-based courses. So, I thought it would be helpful to use this 2-3 years to really improve my math background. After that, I would apply to a PhD in Stats/Biostats somewhere else. This of course would take much more time.

- Try to enroll into the PhD in Math with concentration in Statistics and stay in the area. I already know some faculty and the department, and I like the place. The thing is, is it worth to do a PhD in Math with concentration in Statistics? Do the academy market see them differently? Would the recommendation letter from my advisor and other faculty be more important than the degree itself?

I plan to work as a researcher (either in academy or industry). I like the idea to be a faculty, but I don't have experience in teaching so far. If I get the opportunity to teach during PhD, I could see if that is for me. And maybe do a posdoc later to help in this sense.

I appreciate any feedback! I am a bit frustrated with this situation but it is out of my control as I cannot move anyway.

[bayessays]

At the end of the day, if the faculty are actually actively publishing in top statistics journals (JASA, JRSS-B, Biometrika, etc) then you should be totally fine if you can do work like that.  It will probably be harder to find a community of fellow students who want to become statistics professors, if that matters to you.  Most people do post-docs before getting faculty positions now, so you could do that in a stats department.

[alemao]

3 hours ago, bayessays said:

At the end of the day, if the faculty are actually actively publishing in top statistics journals (JASA, JRSS-B, Biometrika, etc) then you should be totally fine if you can do work like that.  It will probably be harder to find a community of fellow students who want to become statistics professors, if that matters to you.  Most people do post-docs before getting faculty positions now, so you could do that in a stats department.

Thanks @bayessays . Some faculty are still publishing, but others not. Is this a bad sign? At least one is doing lots of theoretical papers in stats journals.

Do students often publish after the qualifying examinations, or it is more writing a thesis and use the chapters as potential papers?

I wonder whether it is common to publish in stats journals during PhD or only during posdoc.

I appreciate your help!

[bayessays]

The dissertation in statistics is 3 loosely-related papers stapled togethe.  Some papers might be published before you're done with the PhD, some might still be under review.

In statistics, I'd say no publishing at all is a very bad sign because that means you can't work with that person.  I wouldn't be comfortable attending a place with less than 3 very-actively-publishing stats professors because people leave all the time, especially in that type of environment.  For example, I can think of two middle-of-the-road math departments (UIC and Arkansas) that have had amazing statisticians in the last decade, but they were only there for a few years because they got jobs at good stats departments.  It's risky.

[alemao]

4 hours ago, bayessays said:

The dissertation in statistics is 3 loosely-related papers stapled togethe.  Some papers might be published before you're done with the PhD, some might still be under review.

In statistics, I'd say no publishing at all is a very bad sign because that means you can't work with that person.  I wouldn't be comfortable attending a place with less than 3 very-actively-publishing stats professors because people leave all the time, especially in that type of environment.  For example, I can think of two middle-of-the-road math departments (UIC and Arkansas) that have had amazing statisticians in the last decade, but they were only there for a few years because they got jobs at good stats departments.  It's risky.

Thank you @bayessays. This makes a lot of sense. Basically, I can only see two very good professors I would work with, but just one actively publishing. It is indeed risky. Thanks for touching on this.