2020 STAT/BIOSTAT PHD profile evaluation

[carrrll]

Hi all,

This is my first post in Gradcafe. I appreciate if you guys can give me some suggestions on choosing schools for my PHD.

 

Type of Student: Asian Male

 

Undergrad Institution: UIUC
Major: Statistics
GPA: 3.32

Math/Statistics Grades:  Applied Linear Algebra, Applied Regression and Design, Basics of Statistical Learning, Advanced Data Analysis, Intro to Mathematical thinking, Method of Applied Statistics, Statistical Computing, Statistical Data Management, Statistics and Probability I, Statistics and Probability II, Calculus I, II, III.

 

Graduate Institution: University of Warwick(UK)
Major: Statistics
GPA: 4.0

Math/Statistics Grades:  Medical Statistics, Designed Experiment, Advanced topics in Data Science, Bayesian Statistics and Decision Theory, Statistical Method, Statistical Practice, Multivariate Statistics, Data Mining.

 

Additional Info:  6 months dissertation doing research on Bayesian model selection

 

GRE General Test:
Q: 167
V: 155
W: 3

Programs Applying: Biostatistics/Statistics 

 

Research Experience: none
 

Letters of Recommendation:  One from dissertation research supervisor, one from course lecturer, one from Internship supervisor 

Coding Skills: R , python, SAS, SPSS
 

Concerns: My math background is really bad, and I did not take any high level math course since my undergrad department is more applied and job oriented. And I was going to work after I graduate from undergrad, and I took a lot of applied course. I was hoping I can go to work in pharmaceutical industry as a SAS programmer. But I realized it is not what I want.  And I am working on a 20000 words dissertation doing research on Bayesian model selection. That would be my only real research experience. Lack of math background and research experience would be a big minus. I studied a lot of data mining, machine learning and deep learning techniques. But I do not think that is useful.

 

Any help would be greatly appreciated !!!

[hnn12]

I think it will be very, very tough for you to get into a reasonably good PhD program in Statistics. Your Master GPA is good, but it cannot make up for your Undergrad GPA. This is more so because your Undergrad was done in the US whereas your Master is in another country. I am an applicant this year, and I also got my education in the UK. My impression is that the admission committees are usually not familiar with our grading system. I know that Warwick has a fantastic Maths department and it is a great achievement to get 4.0 from them, but the adcoms may not have that knowledge. Furthermore, advanced Maths courses, especially Real Analysis and Linear Algebra, are much more important than Applied Statistics courses. You are an international student, so it is even tougher for you, given that you have no research experience. 

My suggestion is that you can try some PhD programs in Biostatistics. The required Maths background for PhD Biostatistics is much lighter than PhD Statistics. You can consider taking the GRE Maths Subject Test and try to absolutely ace it. To some extent, it can help alleviate the adcoms' concern about your Maths background. Another option for you is to do another Master degree in the US and try to get the most out of it (advanced coursework, research experience and connection with profs). 

[Stat Assistant Professor]

OP: Unfortunately, I think the above poster is correct in his assessment of your chances. However, I don't see any real analysis in your list of classes you have taken, and the course titles for the classes you took in your MS program do not give any indication whether you had to take theoretical/mathematical statistics (e.g. "Data Mining," "Medical Statistics," "Designed Experiment," and "Statistical Method" suggest to me that these courses were more on the applied and computational side and not of the Casella & Berger flavor).

I think you may also struggle to get into a Biostatistics program without real analysis and a few more mathematical classes. Even if you ultimately decide to do research that is heavily applied/computational, you still need to pass theoretical courses like Theory of Linear Models, Statistical Inference, Large Sample Theory, etc. in a Biostatistics program. And the adcoms willl have concerns about this with your record. I don't think a stellar Math Subject GRE score will help much (since Biostat departments don't seem to care much about this, and scoring decently on this standardized test can be done through a lot of practice). I think your best chance may be to either enroll as a non-degree seeking student and take a handful of advanced math/stat classes and do well in them, OR obtain a Masters in Math or (Bio)statistics in the U.S. In some Biostat MS programs, you can transfer directly to the PhD program if you perform well in the Masters-level classes and the first semester of PhD-level classes (you could take one or two of these in the first semester of your second year).

[carrrll]

2 hours ago, Stat PhD Now Postdoc said:

OP: Unfortunately, I think the above poster is correct in his assessment of your chances. However, I don't see any real analysis in your list of classes you have taken, and the course titles for the classes you took in your MS program do not give any indication whether you had to take theoretical/mathematical statistics (e.g. "Data Mining," "Medical Statistics," "Designed Experiment," and "Statistical Method" suggest to me that these courses were more on the applied and computational side and not of the Casella & Berger flavor).

I think you may also struggle to get into a Biostatistics program without real analysis and a few more mathematical classes. Even if you ultimately decide to do research that is heavily applied/computational, you still need to pass theoretical courses like Theory of Linear Models, Statistical Inference, Large Sample Theory, etc. in a Biostatistics program. And the adcoms willl have concerns about this with your record. I don't think a stellar Math Subject GRE score will help much (since Biostat departments don't seem to care much about this, and scoring decently on this standardized test can be done through a lot of practice). I think your best chance may be to either enroll as a non-degree seeking student and take a handful of advanced math/stat classes and do well in them, OR obtain a Masters in Math or (Bio)statistics in the U.S. In some Biostat MS programs, you can transfer directly to the PhD program if you perform well in the Masters-level classes and the first semester of PhD-level classes (you could take one or two of these in the first semester of your second year).

Thank you so much for your suggestion. It is very helpful. I was thinking do more theoretical courses in UK. But, it turns out I can not do that here...

[carrrll]

14 hours ago, hnn12 said:

I think it will be very, very tough for you to get into a reasonably good PhD program in Statistics. Your Master GPA is good, but it cannot make up for your Undergrad GPA. This is more so because your Undergrad was done in the US whereas your Master is in another country. I am an applicant this year, and I also got my education in the UK. My impression is that the admission committees are usually not familiar with our grading system. I know that Warwick has a fantastic Maths department and it is a great achievement to get 4.0 from them, but the adcoms may not have that knowledge. Furthermore, advanced Maths courses, especially Real Analysis and Linear Algebra, are much more important than Applied Statistics courses. You are an international student, so it is even tougher for you, given that you have no research experience. 

My suggestion is that you can try some PhD programs in Biostatistics. The required Maths background for PhD Biostatistics is much lighter than PhD Statistics. You can consider taking the GRE Maths Subject Test and try to absolutely ace it. To some extent, it can help alleviate the adcoms' concern about your Maths background. Another option for you is to do another Master degree in the US and try to get the most out of it (advanced coursework, research experience and connection with profs). 

Thanks for your post. Is that possible for me to get into non-prestige biostat PHD programs? Is there any universities you may think of  I have some chance to get in? Thank you so much

[Geococcyx]

I'm just another applicant, so take my suggestions lightly.  That said, I'm given to understand that epidemiology programs are more forgiving of light math backgrounds.  I'm a little unclear as to what your endgoal is for your career (other than "not a SAS programmer in pharmaceuticals", of course), but depending on what you want, an epidemiology degree might be a way to get there.  

If you do want to stick with biostatistics/statistics, though, Stat PhD Now Postdoc's suggestions are probably the best.  I'm pretty sure I outperformed my evaluations from this forum during this year's application cycle, so I'm not inclined to stop you from applying to a few lower-ranked biostatistics programs that you're interested in if you really want to, but I wouldn't spend too much money on that vs. enrolling in some theoretical math classes.

Edited March 17, 2019 by Geococcyx

[carrrll]

3 hours ago, Geococcyx said:

I'm just another applicant, so take my suggestions lightly.  That said, I'm given to understand that epidemiology programs are more forgiving of light math backgrounds.  I'm a little unclear as to what your endgoal is for your career (other than "not a SAS programmer in pharmaceuticals", of course), but depending on what you want, an epidemiology degree might be a way to get there.  

If you do want to stick with biostatistics/statistics, though, Stat PhD Now Postdoc's suggestions are probably the best.  I'm pretty sure I outperformed my evaluations from this forum during this year's application cycle, so I'm not inclined to stop you from applying to a few lower-ranked biostatistics programs that you're interested in if you really want to, but I wouldn't spend too much money on that vs. enrolling in some theoretical math classes.

Thanks

[Stats951]

Can I ask why you put your ethnicity? Do PhD admission really look at this? I assumed it was the case only for undergraduate schools. 

[Geococcyx]

@Stats951 I'm interpreting "Asian" as meaning that they are a citizen of a country in Asia, rather than a citizen of the US (I'm not a lawyer, I'm unclear how permanent residents or DACA fit into this).  Domestic students have a much easier time in the admissions process than international students, so domestic/international status is definitely important.  I might be incorrect in that making that interpretation, though.

Ethnicity might factor into grad school admissions in the form of a diversity statement, wherein some schools will ask how you contribute to diversifying the school/field -- I recall UCLA had this, maybe also Wisconsin.  

[hnn12]

10 hours ago, carrrll said:

Thank you so much for your suggestion. It is very helpful. I was thinking do more theoretical courses in UK. But, it turns out I can not do that here...

In addition to what @Geococcyx said, you can try some Bioinformatics programs, which emphasize the computational aspect more. I have a friend who got into a reasonable PhD Bioinformatics program without any research experience and knowledge in biology. She has a light Maths background and come from an institution that is not known outside my country. That said, she got a higher GPA than you, so I am not sure how your chances are at PhD Bioinformatics, but you can give it a shot if you want. Perhaps more experienced members of the forum can give you more insights.   

[Stat Assistant Professor]

12 hours ago, Geococcyx said:

@Stats951 I'm interpreting "Asian" as meaning that they are a citizen of a country in Asia, rather than a citizen of the US (I'm not a lawyer, I'm unclear how permanent residents or DACA fit into this).  Domestic students have a much easier time in the admissions process than international students, so domestic/international status is definitely important.  I might be incorrect in that making that interpretation, though.

Ethnicity might factor into grad school admissions in the form of a diversity statement, wherein some schools will ask how you contribute to diversifying the school/field -- I recall UCLA had this, maybe also Wisconsin.  

Permanent residents and green card holders are considered domestic applicants since they don't require student visas to study in the U.S. And there is less competition among domestic students for sure. Either way, though, "Asian male" is definitely not an underrepresented group, so it wouldn't help. It does help in admissions to be a highly qualified domestic female or underrepresented minority, though.

As it stands, the OP's math background is still too light for Biostatistics (you don't need a *ton* of upper division math for biostat, but you do need real analysis -- and the amount of upper division math/stat needed on the transcript seems to be inversely proportional to the undergrad GPA. If your undergrad GPA isn't stellar, then you certainly need more math as evidence of mathematical ability to compensate for that).  

Edited March 18, 2019 by Stat PhD Now Postdoc

[carrrll]

7 hours ago, hnn12 said:

In addition to what @Geococcyx said, you can try some Bioinformatics programs, which emphasize the computational aspect more. I have a friend who got into a reasonable PhD Bioinformatics program without any research experience and knowledge in biology. She has a light Maths background and come from an institution that is not known outside my country. That said, she got a higher GPA than you, so I am not sure how your chances are at PhD Bioinformatics, but you can give it a shot if you want. Perhaps more experienced members of the forum can give you more insights.   

That's very helpful. Thanks!

[carrrll]

13 hours ago, Stats951 said:

Can I ask why you put your ethnicity? Do PhD admission really look at this? I assumed it was the case only for undergraduate schools. 

I dont really know. This is my first post. So I may have made some mistakes