Profile Eval: non-traditional Ph.D Stats applicant

[arborio]

This will be the first (hopefully last) round that I apply for Stats Ph.D programs (Fall 2025 admission cycle)

undergrad school: UC campus (not Berkeley or LA)

undergrad major: Environmental Science minor in GIS

undergrad GPA: 3.1 overall--this was the result of beginning uni over 10 years ago, having an undiagnosed disability, being immature and aimless. I think I can address it reasonably in my SOP without making excuses and showing improvement in study.

Relevant Classes from ugrad:

• Calc 1 >10 years ago, grade (B-)
• Calc 2 >10 years ago grade (B) retaken at a CC after an (F)
• Diff Eq: taken in 2021 grade (B) at a CC

Here's where it gets somewhat interesting, I recently finished my M.S in Statistics

M.S school: California State University, XYZ not well-known outside of the area, definitely a place for upward mobility. Some alumni of the program are doing very well in industry very few have gone on to Ph.Ds. The degree is terminal at the M.S, course-based and has very little room for research. 2 year program with comprehensive exam to graduate.

M.S GPA: 3.91

Classes taken: all A's unless ANOVA which was an A-

• Probability Theory (Hogg textbook) 
• Mathematical Stats (J.A Rice textbook)
• Regression Analysis
• ANOVA (Design of Experiments Montgomery book)
• Electives: R + Advanced R, Data Visualization, Deep Learning, Bootstrapping Methods, Statistical Machine Learning

Research: none in Stats

Softs: published in Environmental Sciences at a think-tank, decent career in that sector pivoted to SWE and learning to program. Given a scholarship at my MS program, ASA club leader. Hopefully a compelling SoP and good LoRs. Under-represented minority, first-gen college graduate from low income (if it matters)

Schools considering (in order of interest):

• UC Santa Cruz
• Northwestern
• UVA
• Texas A&M
• Washington U St. Louis
• Boston U
• Pitt
• Also open to University of British Columbia
• European schools with low barriers to enter and good name recognition

Goal: Industry. Research Scientist at FAANG+, Quant/Financial. I'd like to spend the next 5-7 years of my life working towards the credentials needed for a lucrative career that needs statistical experts. I enjoyed my MS program and would have stayed on if they offered a Ph.D. I am an adult and will likely be on the older side of any cohort I get into, I am aware of the cost-benefit of staying in my position vs. leaving for an unknown.

Weaknesses: should be obvious, I do not have a B.S in Math or Stats, I haven't taken (formally) linear algebra though very much needed it in the M.S, no higher math like Real Analysis or anything, never written a proof for a higher level math class. Low ugrad GPA in irrelevant field.

 

GIVE IT TO ME STRAIGHT!

 

[bayessays]

Most schools will not even look at your application if you haven't taken linear algebra.  It's possible to overcome math grades like yours, but you need to show some evidence that you can handle math by taking a real linear algebra course (and honestly, for someone with your profile, you need to go above and beyond and take something like real analysis to prove you can handle the coursework).  Graduate statistics courses, which usually give almost everybody in the class As, are not going to prove this.

I don't think any of your current US schools are remotely realistic.  In your current state, I'd start your search at biostats PhD programs outside the top 20 on US News that do not require linear algebra.

If you had linear algebra, real analysis, and a little research experience, I think some of the lower-ranked schools like UVA might start to become possible.  I don't see any quick path to you improving your profile enough to get into schools like UCSC/TAMU/Northwestern.

[arborio]

I don't think most graduate-level courses in stats give everyone all A's--if that were the case there wouldn't be hundreds of posts on forums with M.S GPAs below 4.0. Certainly a track-record of graduate statistics courses is valuable for stats research at the doctoral level. I also do not want a degree in biostatistics, so I will not look into them--I doubt it is significantly easier to get into #68 US World News ranked school in biostats than #88 in stats.

I can certainly take linear algebra at the minimum. I would like to take real analysis as a conditional admit or something, or online at the schools I am gunning for if possible--if you know of any such offerings at please do share. I would happily take upper div and MS-level math courses if I could get into MS in Math programs at schools that have Ph.D in Stats degrees, and use that as my "in." Ex: M.S in Math at UCSC --> do not complete program --> change course of study to Ph.D in Stats at UCSC. I am amenable to hearing the "not quick" path to improve my profile though.

I can continue to look at lower-ranked schools without linear algebra as a requirement, I do think personal statements and diversity in programs are important and have yielded me faculty encouragement to apply from one of the schools you claim are not "remotely realistic" but will consider your opinion.

[Sstat]

Yes.. although undergrad gpa weights more, its not uncommon that some grad schools give strict evaluation.. 

[arborio]

On 8/18/2024 at 4:58 PM, Sstat said:

Yes.. although undergrad gpa weights more, its not uncommon that some grad schools give strict evaluation.. 

that's unfortunate given that my ugrad degree is old and irrelevant. something I will keep in mind. any advice on how to get more competitve/ready to apply?