Statistic Program Advice // Which Schools let me in?

[TrynnaDoStat]

Hello everyone. I am new to this forum and signed up because I would like advice on which statistics programs best fit my goals and which of the top programs I could realistically get into. My goal is to get a PhD in statistics and work in industry after graduation. Although Stanford is known to be more theoretical, I would still like to know what you all think my chances of getting in their would be. The #1 ranking of Stanford and it's beautiful location really makes me want to go there.

Here is my information:

Undergraduate School: ranking is in the 20-25 range of all undergraduate schools in the US.

Year: Junior

Overall GPA: 3.8

Math/Stat GPA: 3.95

Classes (Grade): Calc I (AP Credit), Calc II (AP Credit), Calc III (A), Linear Algebra (B+), Business Statistics (A), Into to Mathematical Probability (A), Intro to Mathematical Statistics (A+), Applied Linear Models (A), SAS Programming (A), Basic Real Analysis (currently enrolled - probably will get an A), Mathematical Statistics - Graduate Level (currently enrolled - probably will be in the B+ to A range), Longitudinal Analysis - Graduate Level (currently enrolled - probably will get an A)

Classes I am planning to take senior year: Stochastic Processes, Categorical Data Analysis - Graduate Level, Survival Analysis - Graduate Level.

Classes I am considering taking senior year: Advanced Real Analysis - Graduate Level, Probability - Graduate Level. Any other recommendation for classes to take?

GRE: 700 Q, 690 V. Obviously I need to retake this as my Q score is pretty bad. I plan on studying my ass off and getting an 800 the next time around. I know why my Q score was so low - I was too slow. If I work on my speed, getting a 780+ shouldn't be too difficult. The questions are not difficult - I am just too slow.

Research Experience: I have been involved in a psychometric lab all throughout my junior year. We are currently working on a publication.

Internship Experience: I had an internship at a government agency where I didn't really do anything (I don't even know why they paid me) and I have an internship at a consulting firm this summer.

Letters of Recommendation: I can get a good letter from the professor I do research under and I plan on just getting the other letters from professors I have taken a few classes with.

Thanks everyone. Any advice or suggestions would really help.

Edited March 22, 2011 by TrynnaDoStat

[kash]

Hello everyone. I am new to this forum and signed up because I would like advice on which statistics programs best fit my goals and which of the top programs I could realistically get into. My goal is to get a PhD in statistics and work in industry after graduation. Although Stanford is known to be more theoretical, I would still like to know what you all think my chances of getting in their would be. The #1 ranking of Stanford and it's beautiful location really makes me want to go there.

Here is my information:

Undergraduate School: ranking is in the 20-25 range of all undergraduate schools in the US.

Year: Junior

Overall GPA: 3.8

Math/Stat GPA: 3.95

Classes (Grade): Calc I (AP Credit), Calc II (AP Credit), Calc III (A), Linear Algebra (B+), Business Statistics (A), Into to Mathematical Probability (A), Intro to Mathematical Statistics (A+), Applied Linear Models (A), SAS Programming (A), Basic Real Analysis (currently enrolled - probably will get an A), Mathematical Statistics - Graduate Level (currently enrolled - probably will be in the B+ to A range), Longitudinal Analysis - Graduate Level (currently enrolled - probably will get an A)

Classes I am planning to take senior year: Stochastic Processes, Categorical Data Analysis - Graduate Level, Survival Analysis - Graduate Level.

Classes I am considering taking senior year: Advanced Real Analysis - Graduate Level, Probability - Graduate Level. Any other recommendation for classes to take?

GRE: 700 Q, 690 V. Obviously I need to retake this as my Q score is pretty bad. I plan on studying my ass off and getting an 800 the next time around. I know why my Q score was so low - I was too slow. If I work on my speed, getting a 780+ shouldn't be too difficult. The questions are not difficult - I am just too slow.

Research Experience: I have been involved in a psychometric lab all throughout my junior year. We are currently working on a publication.

Internship Experience: I had an internship at a government agency where I didn't really do anything (I don't even know why they paid me) and I have an internship at a consulting firm this summer.

Letters of Recommendation: I can get a good letter from the professor I do research under and I plan on just getting the other letters from professors I have taken a few classes with.

Thanks everyone. Any advice or suggestions would really help.

Looks like your grades are good make sure you get an A in real analysis, and definitely an A in your Mathematical statistics classes, some more classes you can take could be a course on the theory of the linear model, multivariate analysis, time series analysis, graduate probability is good. Having a B+ in linear algebra might hurt a bit since you need to be really solid in that area. How much work did you actually do in your psychometric lab, because that will affect how good of a recommendation you will get. To get into Stanford you will have to get three really good recommendations, usually getting a recommendation from a professor you just had one class with doesn't get you a great recommendation, try working with other professors and get to know them. And you will have to take the GRE Math Subject test if you want to go to Stanford, looking at your coursework you probably do not have the background to get a score above the 90th percentile and it will probably require a lot of work and self study (which in my opinion is not worth it) to get above the 80th percentile without classes in topology, abstract algebra, complex analysis, differential equations, etc. Most schools do not require the subject test besides UChicago, Stanford, and Columbia. Im not sure you really want to spend your time studying for this test, since you dont really even have a compelling reason to attend Stanford, besides the ranking. What is your area of interest by the way.

[TrynnaDoStat]

Thanks for the reply. I do a good amount of work in my psychometric lab - let's just say that if/when our project gets published, other than my professor in the lab, I'll be the primary author. I needed a letter of recommendation for some of the internships I applied to this summer and when I hear back from the programs they always mention her glowing letter. So, I'm not too worried about her letter of recommendation. For the other letters, I was just going to get some of the professors I have had multiple classes with and whose office hours I frequent write the letters (the Stat department at my school is very small so it's pretty close knit). You suggest that I "try working with other professors and get to know them". How should I do this without actually getting involved in research? I feel as if I have a pretty heavy load with classes/research and that taking on more responsibility would wear myself too thin.

It looks like the schools that require the math subject GRE are out of the picture, since I do not have the background required for the test. However, if several of the best statistics schools require such a background (by virtue of requiring the subject test), I am worried that my background is not mathematical enough and that perhaps I should stop taking so many stat classes and take math classes instead.

I don't know what my area of interest is in. I do like the theory of statistics, but I do not want to work as a professor. On the applied side, I am really enjoying my longitudinal data analysis class so far.

[kash]

Thanks for the reply. I do a good amount of work in my psychometric lab - let's just say that if/when our project gets published, other than my professor in the lab, I'll be the primary author. I needed a letter of recommendation for some of the internships I applied to this summer and when I hear back from the programs they always mention her glowing letter. So, I'm not too worried about her letter of recommendation. For the other letters, I was just going to get some of the professors I have had multiple classes with and whose office hours I frequent write the letters (the Stat department at my school is very small so it's pretty close knit). You suggest that I "try working with other professors and get to know them". How should I do this without actually getting involved in research? I feel as if I have a pretty heavy load with classes/research and that taking on more responsibility would wear myself too thin.

It looks like the schools that require the math subject GRE are out of the picture, since I do not have the background required for the test. However, if several of the best statistics schools require such a background (by virtue of requiring the subject test), I am worried that my background is not mathematical enough and that perhaps I should stop taking so many stat classes and take math classes instead.

I don't know what my area of interest is in. I do like the theory of statistics, but I do not want to work as a professor. On the applied side, I am really enjoying my longitudinal data analysis class so far.

Well there are plenty of great statistics programs that dont require the subject test, such as Harvard, Berkeley, Washington, and all biostats departmnets. You can also apply overseas to Oxford, and Toronto for example (both of which are great schools). The only great schools that require the subject test are Chicago and Stanford (Columbia is not considered a great school in the world of statistics). As to your question on how to get to know your professors better, well visit their office hours and talk about statistics, outside of your class assignments. Show them that you have a real interest in the subject, talk to them about journal articles you have read, ask intelligent questions after you have read them, and ask them for recommendations on articles to read. This can also help you figure out which field you are interested in, and what research in that field looks like. Try taking an independent study class if that is at all possible. In your personal statement it will be best to talk about areas in which you have experience in, rather than areas you only have a vague idea about.

[TrynnaDoStat]

Hypothetically speaking, lets say I took the math subject GRE and did well. Do you think I would have a realistic chance at Stanford?

Back to the topic, what are the recommendations for schools if my goals is to go into industry. I saw the below in another thread from 2009 and wanted to know if others agree:

Top theory-focused Programs: Stanford, Harvard, UChicago, UCBerkeley etc.

Top Mixed Programs (strong in both theory/applied): UIUC, UMinn, UWisc, UMichigan, Texas AM, Florida, UNC Chappel Hill

Top Applied Programs (generally not "highly ranked" but have excellent industry connections): NCSU, VTech

I think what I want is a good "mixed program". Although I want to apply my statistical knowledge, I believe having a firm understanding of the theory is important if you want to be successful in your applications.

[stansfield]

I think stanford is the only one that require the gre math, uchicago recommends it but it's not required. you need maybe 65+ on gre math for stanford stat since they recommend 85+ for math phd. You should start contacting some professors from stanford now and establish a fit, that could go a long way in offsetting standardized test scores.

Edited March 25, 2011 by stansfield