2019 Biostatistics/Statistics PhD Profile Evaluation

[merp123]

Hi guys, I am applying to Ph.D. programs for fall 2019 cycle. I was hoping if I can get some feedback on the schools I am considering given my profile. 

Undergrad Institution: Large state school (ranked between 150 - 200 nationally)
Major(s): Statistics and Applied Math
GPA: 3.8

 

Type of Student: Domestic White Male
GRE General Test: 

V: 162

Q: 164

W: 5

 

Programs Applying: PhD in Biostatistics/Statistics

 

Research Experience: Three years as a RA in statistics and applied math, two summers doing research with a professor at a top 3 biostat department, two first author papers submitted where one is a methods paper in statistics, four papers submitted total. 

 

Letters of Recommendation: One from each of my research advisors 

 

Relevant Courser Work: 
Math and Stat Courses: Calc 1 - 3 (A), Intro to Linear Algebra and Diff Eqns (A+), ODES (A), PDES (A),  Numerical Analysis I (B), Numerical Analysis II (A+), Data Analysis (A), Mathematical Statistics (A), Probability (A+), Stochastic Processes (A), Linear Algebra (B-), Topology (A), Advanced Calculus (B+), Independent study in Bayesian Statistics

 

Computing Skills: R, Python 

Applying to: 

Biostatistics

• UNC
• Minnesota  
• Emory
• UPenn
• UW (a man can dream lol)

Statistics

•  Florida
• NC State 
• Duke
• UConn
• NYU

Any forms of advice or suggestions would help. Thanks! 

[bayessays]

A lot will probably come down to your letters - you're in a unique position with the B- in LA, middle of the road GRE score, not a big name school but your research sounds far beyond what most people have and you have a letter from a prof at a top department.  Assuming your letters are very good, my guess is that Duke, UW and NYU (so small) are still probably reaches, but that you'll get into a few of the others.  But if you have a glowing letter from a Prof at a top 3 department with whom you wrote a stats methods paper, I could see you conceivably getting in anywhere.

[GoPackGo89]

For biostats I would throw Harvard,  Hopkins, and Michigan on your list if you like their programs/location at all. As long as your letters are strong  (the prof at a top 3 school I'll assume is at UW and a stellar letter from them will go a long way. If you've done 2 summers and published research with them it seems they like the work you do)

Is this list made up of bayesian focused programs? If so I'm not much help for the stats programs but you'll get an offer from one of the top 6 biostats programs.

What math/stat/computing courses are you in now? Crush those as they will be the last coursework the adcom sees.

[Bayesian1701]

I agree with @bayessays if you rec letters are great you have a good chance at getting into a top department.   I might consider applying to Texas A&M as well if their application is still free because it is a Bayesian program and less of a reach than Duke.   But since the sample size of people with tons of research experience like you have is relatively small its hard to be certain about what would happen.  It's going to come down to whether or not they are willing  to gamble that your mathematical maturity is good enough when your math grades aren't amazing (relative to most people in top programs) and you don't have any graduate classes.

[bayessays]

To give you an idea, I had similar research, significantly worse grades, slightly better GRE scores and was waitlisted at some top programs like UW and Chicago, admitted to many top 20 stat/top 5 biostat, but also rejected at some programs outside the top 40. It's a mixed bag. Your profile is far beyond UConn level though, so unless you have personal reasons, I'd add some middle range schools to your list. If you're shooting for the best possible school, you will get into multiple top 25 programs I think. 

[merp123]

I'm taking measure theory this semester as in independent study with my research advisor and she said she would talk about it in her letter. Do you guys think that would make any difference in mitigating my grades in LA and Advanced Calculus? 

 

[bayessays]

Yes that will help.  I don't think it's a huge deal.  I was worried my B- in linear algebra would torpedo my applications too, and it was a lot of worry over nothing. Keep doing what you're doing and you'll have a lot of success.

[BL250604]

Hi guys, I am looking at programs for the fall 2019 admissions cycle, for a statistics Ph.D.

Undergrad Institution: Small liberal arts school (~70 for all LAC in the U.S.)
Major(s): Mathematics, and Finance
GPA: 3.4 (will be a 3.6 or so when I graduate)

 

Type of Student: Domestic White Male, NCAA Athlete (Conference honor roll, rookie of the year, player of the week)
GRE General Test: Taking soon

 

Programs Applying: PhD in Statistics

 

Research Experience: 

• RA in sabermetrics (baseball statistical research), from which I will have 1 presentation, poster and a publication (first author)
• Research over the last 2 years in applied spatial statistics (on gerrymandering, and its prevalence in PA)
  • 7 presentations, 2 more upcoming. Poster presentation (won "best presentations" distinction at JMM), and have on more upcoming. 
  • Published an op-ed in a national newspaper on my research.
  • Have 2 (hopefully) publications on this topic, 1 in a mathematics journal, and 1 in a statistics journal (both refereed)
• One more research project in predictive analytics to predict voter behavior
• Independent study in Fall '18 in applied spatial statistics, and time series

 

Letters of Recommendation: One from my research advisor, one from my advisor and one from the research coordinator who was the director of DIMACS at Rutgers for over a decade.

 

Relevant Course Work: 
Math and Stat Courses: Calc 1 - 3 (B+), Intro to Linear Algebra (C+, medical condition), Mathematics statistics, I, II (B+) , Transition to abstract mathematics (B+, proofwriting, logic, etc.), Combinators & GT (B+), Real Analysis (currently enrolled), Experimental Mathematics (A), Independent study (A)

 

Computing Skills: R (fluent), Python, JavaScript

 

I went through some neurological problems the beginning of my sophomore year, into this semester. This can hopefully explain some of my struggles in the classroom, and will be mentioned by my advisor, and myself.

Applying to: 

Statistics

• Florida
• GW (have a faculty with tenure/endowed chair in political science writing a LOR for me)
• NC State (my advisor went there for pure math)
• PSU (my research advisor went there for statistics)
• UVA
• VT
• UConn
• South Carolina
• Clemson
• University of Maryland
• Michigan
• Rutgers
• Temple (one recommender went there for a M.S. in stat, and I go to undergrad in PA)
• OSU
• George Mason
• UNC

I know the list is long, but hopefully I get one on my list! Any forms of advice or suggestions would help. Thanks! 

 

Ben

[Bayesian1701]

@BL250604 how independent and involved was your research experience?  Your grades aren’t good and your not coming from a top program but good research experience with letters to verify that you are the best undergrad in the department (or something like that) can help mitigate that a lot.  How much of the publications you are on did you actually write?  Did you make some research decisions or did you just do what you where told? Publications and research experience look good but you would need to set your self apart.  You would be a gamble on whether you could handle PhD coursework because you don’t have amazing grades at a great institution like some other applicants.   

Do you like gerrymandering research?  I am interested in social science applications so I have done a lot of reading CVs to find possible advisors and you might only have 1 or 2 advisor options at some of those places.  If you know what you want to research check for advisor matches because that is going to be a big factor.  The more the better.   

NC State is a reach but they are huge so that helps with their competitiveness a bit for a program of that caliber.  UNC, Michigan and Ohio State are reaches as well.  South Carolina, Virginia Tech, and UVA and similar schools  would be doable with great letters and above average  experience.  There probably aren’t any less competitive programs that do a lot of social science research except maybe Kansas State which has someone doing gerrymandering.   It’s not impossible  for places to gamble on relative preparedness vs independent research experience because that is how I think I cracked the top 10 and did so well.  Use your SOP to show your passion and experience and be prepared to answer questions about your research.  

[BL250604]

11 hours ago, Bayesian1701 said:

@BL250604 how independent and involved was your research experience?  Your grades aren’t good and your not coming from a top program but good research experience with letters to verify that you are the best undergrad in the department (or something like that) can help mitigate that a lot.  How much of the publications you are on did you actually write?  Did you make some research decisions or did you just do what you where told? Publications and research experience look good but you would need to set your self apart.  You would be a gamble on whether you could handle PhD coursework because you don’t have amazing grades at a great institution like some other applicants.   

Do you like gerrymandering research?  I am interested in social science applications so I have done a lot of reading CVs to find possible advisors and you might only have 1 or 2 advisor options at some of those places.  If you know what you want to research check for advisor matches because that is going to be a big factor.  The more the better.   

NC State is a reach but they are huge so that helps with their competitiveness a bit for a program of that caliber.  UNC, Michigan and Ohio State are reaches as well.  South Carolina, Virginia Tech, and UVA and similar schools  would be doable with great letters and above average  experience.  There probably aren’t any less competitive programs that do a lot of social science research except maybe Kansas State which has someone doing gerrymandering.   It’s not impossible  for places to gamble on relative preparedness vs independent research experience because that is how I think I cracked the top 10 and did so well.  Use your SOP to show your passion and experience and be prepared to answer questions about your research.  

Thanks for the advice, and I do recognize the faults in my application and record. As for the research, it was an independent project. I made nearly all of the decisions, however they were under the guidance and suggestion of my advisor, as he is a co-author. 

As for the publication, I am doin the entire write up, except for some of the theory on our slratial methods. I am working with my advisor to try and learn how to write the deep theory that goes into these journals, to better understand the process for my next publication. 

The saber metrics research will also result in a publication, and am hoping that will help as well. 

I am hoping that my research experience helps set me apart, and my LOR/SOP, will be very good, I think. All of the schools on my list are places I would go to, so I’m hoping that I can get into one of them on my list. Maybe I’ll toss in UNH as well, for a lower, less competitive school. I am hoping for GW, as I am getting a very well regarded recommender writing on my behalf. 

 

Thanks! 

[BL250604]

14 hours ago, Bayesian1701 said:

@BL250604 how independent and involved was your research experience?  Your grades aren’t good and your not coming from a top program but good research experience with letters to verify that you are the best undergrad in the department (or something like that) can help mitigate that a lot.  How much of the publications you are on did you actually write?  Did you make some research decisions or did you just do what you where told? Publications and research experience look good but you would need to set your self apart.  You would be a gamble on whether you could handle PhD coursework because you don’t have amazing grades at a great institution like some other applicants.   

Do you like gerrymandering research?  I am interested in social science applications so I have done a lot of reading CVs to find possible advisors and you might only have 1 or 2 advisor options at some of those places.  If you know what you want to research check for advisor matches because that is going to be a big factor.  The more the better.   

NC State is a reach but they are huge so that helps with their competitiveness a bit for a program of that caliber.  UNC, Michigan and Ohio State are reaches as well.  South Carolina, Virginia Tech, and UVA and similar schools  would be doable with great letters and above average  experience.  There probably aren’t any less competitive programs that do a lot of social science research except maybe Kansas State which has someone doing gerrymandering.   It’s not impossible  for places to gamble on relative preparedness vs independent research experience because that is how I think I cracked the top 10 and did so well.  Use your SOP to show your passion and experience and be prepared to answer questions about your research.  

Social science isn’t an end all be all for me whatsoever, so are there any other schools which you think I should consider? Preferably ones which might be on the safer end (even though that’s not that possible for me...) thanks! 

[GoPackGo89]

37 minutes ago, BL250604 said:

Social science isn’t an end all be all for me whatsoever, so are there any other schools which you think I should consider? Preferably ones which might be on the safer end (even though that’s not that possible for me...) thanks! 

I think getting an A in real analysis would help you tremendously given your other grades. If you've impressed your letter writers then I think your list is reasonable. 

[BL250604]

4 hours ago, GoPackGo89 said:

I think getting an A in real analysis would help you tremendously given your other grades. If you've impressed your letter writers then I think your list is reasonable. 

Agreed! So far so good in that class. 

[Bayesian1701]

@BL250604 You should be in good shape then.  Submit one the publications as a writing sample not that they will actually read entirely it but some professors definitely looked over the one I submitted.  Not everyone will take a writing sample but some will.  Try a contest or two (like this one).  If you win or an get an honorable mention your odds will increase and if you have something written already they are worth a try.  Doing very well in real analysis this semester is important.  If you get an A in real analysis that will help alleviate the concerns about your mathematical maturity.   Doing well enough to have a rec letter come from that professor might help too.     Focus on your research experience and you should be fine. There is a difference between run of the mill undergraduate research and independent undergraduate research so I wanted to know which is was.   With an A in real analysis and good independent research you should be a strong contender for most middle and lower tier program (not top 20) with at shot at the reach schools.  15 places is a lot of place to apply to and you can certainly apply to all of them if you want to but I would consider cutting it down a bit (8-10) to save time and money. 

[superk60]

Hi guys, I am looking at programs for the Fall 2019 admissions cycle (stats Ph.D):

Undergrad Institution: Top 25 Private (top 10 for stats)
Major(s): Mathematics
GPA: 3.3 (3.4 expected)

 

Type of Student: Domestic Asian Male
GRE General Test: Taking soon. Will probably take math GRE too. 

 

Programs Applying: PhD in Statistics

 

Research Experience: Will be doing independent study/research in the fall on numerical linear algebra. Worked on applied stats research project with professor. Did quant research at a hedge fund last summer.

 

Letters of Recommendation: One from research advisor, one from real analysis instructor (very strong up-trend in performance), one from independent study advisor.

 

Relevant Course Work: 
Math and Stat Courses: 4 stats courses: A's, Linear Algebra: B, Real Analysis 1/2: C/A, Probability: B, Numerical Analysis: A, Computational Bio: A

 

Computing Skills: Python (software engineering experience), R, Matlab

Applying to: 

Statistics

• Penn State
• USC
• UConn
• Yale
• Michigan
• Ohio State
• UCLA
• Rice
• Brown
• UT Austin

Probably way too ambitious, feel free to recommend some substitutions/deletions  

Thanks in advance. 

Edited April 7, 2018 by superk60

[Bayesian1701]

@superk60

I am going to assume that your undergraduate institution is a place where a 3.3 is not a bad GPA and math classes are rarely curved, and that your real analysis professor will be like:  ”he can handle the math”.   But if most people make As in real analysis/probability and you didn't  that's going to look bad. That being said your list is probably to reach heavy.  To consistently get into top programs (ie get into like 3+) you probably need some combination of luck,  a great profile, and a wow factor that makes you unique. Yale and UT Austin may not be highly ranked, but they are both small and still good (Yale being a better program).  Your list is mainly (but not completely) top universities (which can be harder to get into even if they aren't highly ranked) and top statistics programs at state schools which means on an as-is basis you might not do well.   Safety schools are complicated (some of the lower tier programs are small and/or have funding issues) but you are probably lacking realistic options.   I don't really have suggestions on what to drop and what to add because it depends on your preferences (location, program size, research, etc.) but probably fewer places like Yale and Michigan and more places like UConn.  A good strategy in general might be (for domestic students) be: 1-4 places you don't think you will get into (group 1),  2-4 places where you expect to get into 25-50% of this group (group 2), and 1-4 places you expect to get into more than 50% of this group (group 3).  How you structure your list will be based on how much risk you think you can tolerate.     A lot of the programs you are interested are popular (Michigan, Ohio State) enough so that by reading old profiles you can get a vague estimate of what your chances are.   But it's probably realistically going to be +/- 20% with the small amount of biased data available.  But if you would be happy with applying to all those places and only getting into UConn and maybe one or two other programs, and a slight chance of striking out than go for that list.  It's not that you are not a good applicant, its that there are a lot of good applicants. 

[superk60]

Thanks Bayesian, this is very helpful. Let me add that I had a rough freshman year (health issues), so my ex-freshman GPA rounds to 3.6. No idea how I'll address that on the application (!). I'm also taking an grad course in the fall with one of my potential recommenders, which might help a bit. 

To give you an idea of risk tolerance,  I'm targeting Pr(getting into >= k schools) = 1-0.15*k for k ∈ {1,..,5}. Given what you said, here are some more UConn/less Yale schools:

• UCSB 
• University of PIttsburgh
• Boston U
• Rutgers
• UC Irvine
• Colorado State

So my list might be...

Reach/Group 1: UT, Yale, Michigan, UCLA  

Group 2:  UConn, Boston U, Rutgers, Rice

Group 3: Colorado State, UC Irvine, UCSB

Reasonable? Also, if I were to put in for biostat (at schools that have it), would that help?

Edited April 15, 2018 by superk60

[Biostats doer]

 Greetings everyone! I am planning to apply for Biostatistics Ph.D./statistics master programs this fall.  I am humbly and sincerely asking for your gracious advice!

Below is my background:

Institution: Top UC's (Statistics program ranking: 31) 

Major:   Applied statistics (concentrating field:  mathematics) I know it is weird... My intention is to learn more math since I switch my major from applied mathematics. 

GPA: 3.5

 

Type of Student: International Asian Male
GRE General Test:  Not taking yet, will do soon

 

Programs Applying: Ph.D. in Biostatistics/Statistics

 

Research Experience:    1 year of RA in biostatistical research experience and currently working on Large-scale biostatistical data mining. 

Teaching Experience:     One-quarter of Learning assistant for calculus

 

Letters of Recommendation: One from my research advisors, two from my statistics professors

 

Relevant Courser Work:  (P means Pass, some are optional)
Math and Stat Courses:  Cal 1-4: All (A's) | Linear Algebra: (B) | Elementary Statistics: (A) | Intro to Abstract Math: (C+) (with 70% of the class below C- range) | Differential equations for: (P) |                  Real analysis_1: Pass | Real analysis_2: C | Regression analysis: B | Statistical Data Science: B+ (right after I switched to statistics from Applied math major, without any R experience beforehand, worked really hard) | Probability theory: (P) | Time series: (C-) (was taking 5 upper divs at one quarter, trying to allay from my tuition burden, trying to finish in 4 years, therefore admitted sacrifice) | Nonparametric statistics: (C) (medical condition) | Analysis of categorical data: (C+) (medical condition) | Analysis of Variance: (B+) (mental medical condition)| Mathematical statistics: (A-) (medical condition) | Euclidean Geometry: (B+) (still in love with proof writing, want to prove myself) | Applied statistics: (B+) | Advanced statistics undergraduate independent study: (P) | Programming and problem solving: (B) (C language)

 

Computing Skills: R (Proficient, but still learning and exploring), Matlab

 

Applying to:

• UC Berkeley-biostatistics (master)
• Stanford - statistics (master)
• UCR - statistics 
• UCSC -  statistics 
• SMU - biostatistics
• Duke - Biostatistics
• Columbia - Biostatistics
• University of Michigan - Biostatistics
• UIUC - statistics
• Notre Dame - Statistics
• University of Minnesota - Biostatistics
• CMU - statistics in  Public policy
• University of Pittsburg - biostatistics
• NCSU - statistics
• Uccon - biostatistics
• ISU - Statistics
• USC - Biostatistics
• UCLA - biostatistics
• Yale -  Biostatistics

I am pretty sure that  I am not competitive enough and both my math/statistics classes rarely curved (Statistics department is known for the epitome GPA deflation  )

I am preparing for the worst to come. 

Please enlight me and guide me through this, thanks!

Please be critical for my school selection!

 

Edited July 18, 2018 by Biostats doer