Stats PhD Profile evaluation 2021

[noskillz]

Hi guys, I just wanted to get some insights into whether or not my school selections are proper. I'm Applying for Statistics/Biostatistics Phd programs for Fall 2021. Unfortunately my profile is really quite a mixed bag. Overall my undergraduate grades where quite underwhelming because I absolutely hated being in school until my third year (during which time there was also serious illness in my family). Things turned around for me in my last year when I took a research course which I overwhelmingly enjoyed  and my supervisor encouraged me to just try my luck at a graduate program. I was extremely lucky to have gotten into the graduate master's program in the same school as my undergrad. And, not that this means much given the profiles of other people on this site, I did phenomenal in my Msc in stats program. And that too in taking really difficult math classes. My goal during my master's was to build up on everything that I messed up on during my undergrad. I did about 3 courses more than what was needed for the degree requirement on top of doing a thesis, which we are now working on to submit for publication. 

I have also built up a decent research profile and did extra research on top of my thesis work with one of my supervisors and one of my course instructors. My three LoR's will all be from the professors that I have worked with and all three are fairly well known in their respective fields (2 for Bayesian statistics and 1 in econometrics). I'm just thinking about if my school choices are correct and which schools I could/should realistically aim for. I know my undergrad fuckup will follow me for a while especially into the PhD application, but if there's anyone with some advice on how to mitigate that, or help me adjust my expectations down to a realistic level that would be very helpful.

 

Undergrad:

Mathematics (Top 3 ranked Canadian Institute.)

Gpa: 3.23, my grades are quite a mixed bag. I got an A in measure theory, A real analysis, A ODE, A Complex analysis, A- probability, D in linear Algebra, B- in Intro to Statistics, B- in Statistical methods (graduate courses), A in research course, B in regression, B- topology, C in GLM. 

 

Master's (Thesis based):

Statistics (Same institute)

Gpa: 3.93, Advanced Probability (using measure theory) A-, Advanced GLM A-, Advanced Real Analysis A, Convex Analysis A, Bayesian Inference A, Financial econometrics A, Time Series analysis A, Advanced Computational Methods A;

PDEs A, Asymptotic and finite sample theory A, Numerical Analysis A, Functional Analysis A (hoping to use this to make up for the D in lin alg). Took these last 4 graduate courses after I finished my master's.

 

Gre: 165Q 164V 4.5A (the verbal score was a SHOCK to me as well!)

 

As you can see even after finishing my master's I have tried to compensate for my poor undergraduate courses by taking graduate courses that will build up my fundamental math knowledge and have done really well in them. All the while being involved as a research assistant with my professor. I also have a strong teaching experience having done 4 ta-ships during my masters and was super lucky to actually been able to teach a first year math course in the immediate aftermath of finishing my masters.

 

Applying to:

Reaches: Michigan, UNC, Columbia, Duke, Minnesota, Yale, Toronto, Pittsburgh, UBC, Brown

 (honestly just have a felling won't get into any of these, my supervisor thinks I can get in but I don't think so.)

Safety: Simon Fraser, McGill

 

Let me know if I'm aiming too high and if so what are more realistic options. Thanks a lot. And good luck to everyone with their applications and if you're still in school with your midterms/exams.

Thanks.

 

 

 

 

[StatsG0d]

First, I think you'll have to specify which of these departments that you are applying to are stats and which are biostats.

Second, I am surprised you list Pitt as a reach and McGill as a "safety," as McGill is arguably the superior department.

Thirdly, I really think this is tough to judge. The D in linear algebra looks very bad, and normally I would say this would exclude you from any ranked stats/ biostats program. However, you've received top marks in graduate level math classes. It's unclear how adcoms will view this, and you should address it in your personal statement.

Finally, if your advisor knows your situation (i.e., your performance in undergrad) and you are at an institution with a solid stats department, then I would say you can trust your advisor's opinions about where you have a good shot to get in. 

[Stttat]

I have a bit similar math grades. My overall GPA was 3.3 from one of the top universities in my country. I have completed Matrix Theory and Linear Algebra I with A+ but I then got D in Linear Algebra II in my last semester. I was working full-time at the moment and couldn't take the final. I got A+ in Real Analysis but I didn't take other advanced math classes like Numerical Analysis. I am also curious how adcoms will view mine.. 

[noskillz]

@StatsG0d Thanks for your response. My current supervisor actually has a contact in McGill Stats and said that the contact was looking for a phd student so that's why I listed it as a safety. In terms of a little bit more details, I would be applying to:

 

Bios: Minnesota, Michigan, Brown, Duke, UNC, McGill

Stats:  Michigan, UNC, Columbia, Duke, Minnesota, Yale, Toronto, Pittsburgh, UBC, Brown, McGill, SFU

I was also thinking of adding UCSD stats as an option as well (I wouldnt mind applying to UCSD master's in applied math and then take that as a stepping stone).

There are also some recent phd programs in data science, like in NYU and UCSD also has on. But I'm a bit sceptical as to how strong of a research background some of these 'new' programs will give to their students.

 

Thanks!

 

[StatsG0d]

Based on your profile, I would only apply to the larger programs in both stat and biostat. Duke biostat seems like a waste, as does Yale and Brown Stat. 

You could maybe consider adding Emory and/or Pitt for Biostat. The former would be a solid target and the latter a solid "safety". 

[noskillz]

@StatsG0d, by waste you mean totally unachievable or programs are not good?

Thanks for your input. Yeah I thought about emory. Will look into it.

[StatsG0d]

16 hours ago, noskillz said:

@StatsG0d, by waste you mean totally unachievable or programs are not good?

Thanks for your input. Yeah I thought about emory. Will look into it.

All of those programs are much more competitive relative to their respective rankings, so I feel like it's a waste of resources.

[Casorati]

I would assume you attended McGill for undergraduate and master's. If you are pretty sure you can get into McGill's PhD program, it only makes sense to apply to better schools than McGill. With that said, I would not apply to SFU and Pitt. Biostat at Duke and Brown are relatively new and despite their low ranking, it would still be difficult to get in because of their general prestige. Other schools in your list are reasonable targets I think. Given your strong performance in advanced math courses, I think most schools will overlook your D in linear algebra and B- in intro statistics so I wouldn't worry too much about it. 

[noskillz]

Thanks a lot for all the advice. In case any one is wondering I have reduced my final list choice to

McGill, Waterloo, Toronto, UBC, Penn State, Michigan, Minnesota, UNC, Columbia, Rutgers, Northwestern and UCSD (Stats PhD for all).

 

I decided against going in to bio stats because I just didn't see myself really enjoying being in bio-stats for the next 30 years of my life.

I also added Rutgers and Northwestern , since they are in 30-50 range and seem like good schools within my range which I would also enjoy going to. Especially Rutgers, since it has a decently sized stats department from what I saw on their website.

[Casorati]

On 11/15/2020 at 6:09 PM, noskillz said:

Thanks a lot for all the advice. In case any one is wondering I have reduced my final list choice to

McGill, Waterloo, Toronto, UBC, Penn State, Michigan, Minnesota, UNC, Columbia, Rutgers, Northwestern and UCSD (Stats PhD for all).

 

I decided against going in to bio stats because I just didn't see myself really enjoying being in bio-stats for the next 30 years of my life.

I also added Rutgers and Northwestern , since they are in 30-50 range and seem like good schools within my range which I would also enjoy going to. Especially Rutgers, since it has a decently sized stats department from what I saw on their website.

Northwestern has a very small statistics department and it might be hard to get in because of its general prestige. Rutgers has a good statistics department with a wide range of research areas. However, I still think McGill is a better option than Northwestern and Rutgers. 

[StatsG0d]

I agree with @Casorati. Northwestern's program honestly is not great. Evanston is a really nice place and near Chicago so location it's awesome, but the program prestige is pretty lacking.

[noskillz]

@Casorati and @StatsG0d thanks for that. Do then think that instead applying to UIUC, Texas A & M and Purdue would be better? As @Casorati said, I should be aiming above McGill and I believe this schools would satisfy that while also being within my reach (maybe).

[Casorati]

I don't think these schools are much better than McGill either. Maybe Purdue/UIUC are comparable to McGill. I would say schools like Wisconsin/UNC/Penn State are a tier above McGill.

[StatsG0d]

On 11/18/2020 at 6:48 AM, noskillz said:

@Casorati and @StatsG0d thanks for that. Do then think that instead applying to UIUC, Texas A & M and Purdue would be better? As @Casorati said, I should be aiming above McGill and I believe this schools would satisfy that while also being within my reach (maybe).

TLDR: All those programs are pretty similar. If McGill is a sure thing, it's probably only worth it to apply to other places if you would for sure go there over McGill.

These rankings might help (but they include OR so there's some noise). All of those programs are great. The ones @Casorati mentioned are all really good. I would put them in the same tier.

Provided you get into at least two of these, I would say go wherever you feel you would be happiest. If it were my choice, I would say McGill simply for the location. Sure, it's cold, but so is West Lafayette, Urbana, and State College. At least McGill is in a large city and there will be a lot to do / easy to catch a flight somewhere. UNC is kind of a nice middle ground because while Chapel Hill is a college town, a major airport (RDU) is only a 20-minute drive, and Durham and Raleigh are both nice cities with quite a lot to do relative to their size.