I analyze data for a living, and I would like to do a PhD in statistics. However, my undergraduate math background isn't ideal. I'm thinking about doing a master's in statistics, on the grounds that it might help prepare me for a PhD, and, if not, would still be a very good terminal degree. On the other hand, if my undergrad math background is really the biggest thing standing in my way, then maybe a master's degree in math would be a shorter, cheaper, and more direct way to improve my profile. On the third hand, if I ended up stopping with the master's and not doing the PhD, maybe employers wouldn't like the math master's as much as the stat master's. Which path is better? Is my diagnosis of the pros and cons of each accurate? Is there a third path that I haven't considered?
I posted my profile below, but I think that maybe this question goes beyond any one profile: what's a reasonable path to a stat PhD if you're starting from a weak math background?
---
Undergrad: UChicago, BA in econ/policy, 3.4 GPA
Grad: UChicago, master's in policy, 3.4 GPA GRE general: 170V, 170Q, 5.5W
Highest math/stat classes: Analysis in Rn 1 and 2 (B, C+), Linear Algebra (A), Mathematical Statistics for Policy (A)
Research experience: All my job titles have had "research" or "development" in them, and I'm pretty good at thinking of hypotheses and testing them, but I have no publications.
Letters of rec: Two from former supervisors, one from a colleague. None have PhDs.
Computer: Fluent in R and Stata, pretty good at Python, know a little C#, generally able to find my way a new language pretty quick.
I expect that my lack of publications and academic recommendations will work against me. I also suspect that a master's in stats would be equally useful in dealing with this as a master's in math, but I'm not sure.
I self-study math in my spare time (using Baby Rudin, among others), and I recently went back and audited Analysis in Rn-1 and did well, so I'm not worried about my ability to do well in any future math classes.