Statistics PhD profile evaluation

[WChew]

Hi! I'm really unsure where I stand among other applicants for Statistics PhD programs and want to know how competitive I will be in applications.   

Undergrad Institution: Kansas State University (graduating December 2016) 
Majors: Statistics, Economics
GPA: 3.73 overall, 4.0 in Statistics courses and 4.0 in Economics courses

Honor & Awards: Possibly graduating Cum Laude.  

Member in Omicron Delta Epsilon honor society for Economics and Mu Sigma Rho honor society for Statistics.  

Type of student: Domestic White Male
GRE: Taking very soon, expecting Q:160+ V: 150+

Quantitative Courses: Calc 1:A, Calc 2:B, Calc 3: A, Business and Econ Stats: A, Probability and Statistics 1: A, Probability and Statistics 2: A, Regression/ANOVA: A, Design of Experiments: A, Nonparametric Statistics: A, Econometrics: A, Economic Forecasting: A

Programming: Proficient in R and SAS

Experience: Interned in a Research Hospital in KC, Tutor for upper level Macroeconomics

Letters of recommendation: 

1. Boss/Mentor from internship
2. Statistics academic advisor
3. Economics academic advisor  

List of schools:

• Penn State
• Iowa State
• Colorado State
• Ohio State
• University of Michigan
• University of Florida
• University of South Carolina
• probably Kansas State as a safety school as well

Edited November 7, 2016 by WChew

[Karoku_valentine]

In my Personal opinion, your profile lacks some of the basic mathematical training the committees are expecting. There are no higher level Math courses like Analysis or linear algebra (if these were graduate level that would be even better). Even if you took all of those econometric courses, the committees need to know you have enough mathematical training to be able to understand and produce your own proofs because the applied part will come easily later.

Good things are that you have internship experiences and that you have some interest in applied statistics, and I think even being an American may give you a slight advantage in admissions. You need to prepare very well for the gre to get a score that above 165, but I don't know if this would compensate the lack of math.

i would consider doing a master in one of those schools, get math and measure theory necessary for the PhD and apply for PhDs.

[WChew]

Thanks for your reply!  I just realized that I forgot to include Matrix theory in my courses, I got an A in Matrix Theory. This probably substantially changes my profile and if you don't mind, a re-evaluation would be greatly appreciated.

Edited November 8, 2016 by WChew

[windwhispered]

Matrix Theory / linear algebra could certainly be helpful, but my personal perception is that real analysis is a more important qualification.

[cyberwulf]

Yeah, I think you're going to have a tough time getting into a lot of the schools on your list. When you combine your relative lack of mathematical preparation (e.g., no real analysis) with the fact that Kansas State doesn't have a particularly strong student body (about.com says 95% of applicants are accepted there), there are going to be some legitimate doubts about your ability to do the math required in a good Stat PhD program. 

If it's feasible for you, you might want to consider starting in a good Masters program to get some more advanced coursework under your belt, then use that as a stepping stone to the PhD.

[WChew]

Thanks for all replies, I chose to go for admission to PhD programs at the advice of my Statistics advisor.  My first thought was to do a masters before entering into a PhD program.  Her advice was that if I wanted to get a PhD eventually (which I do), to apply directly for a PhD, which is why I will continue to seek a PhD.

 I am struggling to understand the concerns about my math background and how that relates to my admissions chances as I had strong grades in calc and matrix theory.  From what I gather from the admissions websites, a background in calculus and matrix theory is the baseline for acceptance, with real analysis being less important for admission.  

I also feel as though other parts of my application make up the weight of not having real analysis.  My letters of recommendation are mostly strong.  My boss from my internship is a biostatistician and the hospital I worked in has a good reputation in the midwest.  My statistics advisor should be strong as well as she knows a member of the faculty at most of the schools on my list and received her PhD from Penn State.

I do really appreciate the feedback and I probably should have done a better job with my original post (it was a somewhat rushed job).  

[Karoku_valentine]

On November 14, 2016 at 1:58 PM, WChew said:

Thanks for all replies, I chose to go for admission to PhD programs at the advice of my Statistics advisor.  My first thought was to do a masters before entering into a PhD program.  Her advice was that if I wanted to get a PhD eventually (which I do), to apply directly for a PhD, which is why I will continue to seek a PhD.

 I am struggling to understand the concerns about my math background and how that relates to my admissions chances as I had strong grades in calc and matrix theory.  From what I gather from the admissions websites, a background in calculus and matrix theory is the baseline for acceptance, with real analysis being less important for admission.  

I also feel as though other parts of my application make up the weight of not having real analysis.  My letters of recommendation are mostly strong.  My boss from my internship is a biostatistician and the hospital I worked in has a good reputation in the midwest.  My statistics advisor should be strong as well as she knows a member of the faculty at most of the schools on my list and received her PhD from Penn State.

I do really appreciate the feedback and I probably should have done a better job with my original post (it was a somewhat rushed job).  

This is the opinion of the people who have some knowledge of graduate programs in statistics. 

Even if Real Analysis is not "required", the school pre requisites are what they expect as a minimum. The strongest applicants will have graduate level real analysis and linear algebra plus some research (even if it is not super specialized), excellent grades and recommendation letters, and GRE higher scores.

No one is saying you can't do or you won't be admitted. However, the admission committees will have this concern. certainly departments in Penn, Iowa St and Michigan will have it too. This concern is very valid considering, for example, how heavy in probability theory is the qualifier at Iowa State, or how hard classes can get.

In my experience, if I consider into account all the people who are in the PhD of Stat and Biostatistics AND people who were admitted into the master program, only one person had not taken Real Analysis, but this student was admitted to Biostatistics and already had a MS in Biostatistics. In addition, at least one third of the entering class had already master degrees in biostatistics and statistics, I am also certain they all got A's in all or most of their undergraduate Analysis classes.

In other words, you will be competing against people with: 1) Master degrees in statistics and some in mathematics, or just people in the honors program of undergraduate mathematics, 2) people with excellent or good grades in more advanced classes and excellent GRE scores, 3) people with excellent recommendation letters. Even the "weak applicants" are still excellent in most of the schools you mention above.

[Warelin]

FWIW: University of Michigan says they received 341 applications to its statistics program and had 16 new enrollees. Typically universities extend 1.5-2.0x the number of spots to achieve their desired number.  Let's say they extended 32 offers for those 16 spots. This means there's less than a 10 percent chance of being offered a spot. The chances decrease and increase depending on a number of other factors. However, from studying the process, it seems that only individuals coming from a very prestigious undergraduate institution will get offered a spot directly to the PHD. There are exceptions but it tends to be a very rare individual who has a killer resume.
 

I'm not sure how accurate the information is (it grabs data from the schools, but the updatedness depends on the schools themselves) but Kansas State accepts 42 percent of students into their program. They state that 66 people apply and they extend offers to 28 of which 14 accept their offer. Prior to applying, I'd inquire into the program first. Some universities don't like accepting undergraduates into their own program. I believe Stanford is one of those programs.

University of Florida says they receive 394 applications and accept 23 of which 15 enroll.

I'm not saying any of this to scare you but I'm hoping this paints a more accurate picture on why it takes some people several rounds to gain acceptance into a Graduate Programs.

[WChew]

Thanks for the input,  I am somewhat concerned that my GRE will be a little low and I was going to use my score to decide my Masters vs PhD debate (the powerprep practice tests put me at 150V 161Q).  So, how does my profile look for a masters? 

[StatsG0d]

I think you might have a shot at places like South Carolina and Kansas State. I agree with everyone else saying that the other programs might be difficult. Florida may be a possibility because they have a real analysis course in the first year curriculum for those who haven't taken it yet, and I think a lot of other schools admit qualified students who haven't taken real analysis, forcing them to take it in their first year. That being said, I agree that school reputation matters and that not having taken real analysis is a serious detriment taking this into account.

A 161Q is pretty sufficient for a MS, but I'd say this score given your math background will almost certainly exclude you from a PhD program. I'd probably advise against applying to some of those schools, as some do not really distinguish between Master's and PhD admissions (everyone is admitted to the PhD program if admitted, and the first year curriculum is identical). You should probably research this to find out which schools do this.

Edited November 18, 2016 by footballman2399