Statistics PhD Questions

[dino88]

Hi,

I'm kind of a newbie to statistics and come from a different discipline. I've googled the net, but didn't come up with much. So, can someone please explain the following to me:

1) If you applied to a Statistics PhD program and got accepted to an MS/PhD, is that the same thing as getting accepted to a PhD? Pros? Cons?

2) Rankings. I am a bit confused as to what Stats PhD programs are top 10 or top 20 versus top 50 or 100. I suppose this would affect your placement in the end. Can anyone be so kind as to list some schools that are not listed on the U.S. News and World Report for the Math:Statistics (http://grad-schools.usnews.rankingsandr ... statistics)? For example, where do these Ivy League schools fit in that are not on that list? (the ones missing are Yale, Cornell, UPenn...I didn't even know Princeton had a Stats Dept. - I searched their site) Or places like CMU or even international schools like Oxford? I don't really care about rankings, but I'd hate to walk into a program that was way out there.

3) Is there a distinction between Bayesian and Classical?

4) What is the typical size of an incoming PhD Stats class? Any thoughts about that?

Thank you.

[Tinyboss]

I'm applying to mathematics programs, but I think I can answer some of your questions at least:

1) If you applied to a Statistics PhD program and got accepted to an MS/PhD, is that the same thing as getting accepted to a PhD? Pros? Cons?

Very few mathematics programs combine the MS and the PhD programs, but some do. UIC, for example, requires an MS in mathematics in order to apply to their PhD program. More commonly, a school may have an MS program primarily for PhD applicants who have gaps in their preparation. In that case, someone might be admitted to the MS program with the understanding that with satisfactory performance in certain classes, they can advance to the PhD program.

2) Rankings. I am a bit confused as to what Stats PhD programs are top 10 or top 20 versus top 50 or 100.

This site may be helpful: http://graduate-school.phds.org/rankings/statistics

3) Is there a distinction between Bayesian and Classical?

Can't help you here. :wink:

4) What is the typical size of an incoming PhD Stats class? Any thoughts about that?

This depends entirely on the school; some are big, and others are small. If you know what your specialization will be, a small school with faculty who are into that same area could be great. If you aren't sure where you want to concentrate, then a large school with lots of diverse faculty research can give you more options.

[ralphhumacho]

Hi - I've applied to Phd Programs in Statistics and so far got into 4 places I applied (waiting to hear from one more)

1. Not the same thing. MS/PhD requires you to first go through their MS route, pass the master's exam, etc., then go into their PhD program (you MAY not get in the PhD program, even after getting the MS, but you have a huge advantage.) Often times, the MS portion is NOT funded. Many Biostatistics PhD programs require a master's, and this route is common for Biostats, but all Statistics PhD programs I know of (which is most) prefer that you apply for the highest degree you intend to get (so you either apply to an MS program OR a PhD program.)

2. Depends on if you're looking to go into Theoretical Statistics (if you want to get into Academia), General Applied Statistics (if you want to work in industry, finance, consulting, at a lab etc.) or Biostatistics (if you want to work in pharma or a research lab.)

I would highly recommend against biostats. Almost every job that a Biostats PhD can do can be done by a Stats PhD, maybe even a Math PhD with a concentration/grad minor in stats or and MS in Stats. However, a Biostats Phd may lock you out of finance, insurance, etc.

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

3. Kind of - will discuss later

4. For PhD programs' incoming class: smaller programs: 5-6 med size programs: 6-12 (I'd say ~80% of schools are in this category), large programs >12 (very few programs are this large)

Sorry the grammar sucks, im in hurry!

[hedgie]

I had a question regarding your comment "maybe even a Math PhD with a concentration/grad minor in stats or and MS in Stats". How practical is it to do a PhD in Applied Mathematics and become a statistician? If one were to go the route of PhD in Applied Math would they be more employable in industry rather than just the PhD in Stats? I guess along the same lines as your emphasis on not doing a PhD in Biostats, because you can do it with a Stats PhD. I ask because I am interested in Stats PhD but am wondering if Applied Math isn't more practical.

Thanks in advance!

Hi - I've applied to Phd Programs in Statistics and so far got into 4 places I applied (waiting to hear from one more)

1. Not the same thing. MS/PhD requires you to first go through their MS route, pass the master's exam, etc., then go into their PhD program (you MAY not get in the PhD program, even after getting the MS, but you have a huge advantage.) Often times, the MS portion is NOT funded. Many Biostatistics PhD programs require a master's, and this route is common for Biostats, but all Statistics PhD programs I know of (which is most) prefer that you apply for the highest degree you intend to get (so you either apply to an MS program OR a PhD program.)

2. Depends on if you're looking to go into Theoretical Statistics (if you want to get into Academia), General Applied Statistics (if you want to work in industry, finance, consulting, at a lab etc.) or Biostatistics (if you want to work in pharma or a research lab.)

I would highly recommend against biostats. Almost every job that a Biostats PhD can do can be done by a Stats PhD, maybe even a Math PhD with a concentration/grad minor in stats or and MS in Stats. However, a Biostats Phd may lock you out of finance, insurance, etc.

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

3. Kind of - will discuss later

4. For PhD programs' incoming class: smaller programs: 5-6 med size programs: 6-12 (I'd say ~80% of schools are in this category), large programs >12 (very few programs are this large)

Sorry the grammar sucks, im in hurry!

[statsguy]

I had a question regarding your comment "maybe even a Math PhD with a concentration/grad minor in stats or and MS in Stats". How practical is it to do a PhD in Applied Mathematics and become a statistician? Not very. If you want to become a statistician, get a PhD in Statistics. An Applied Mathematics curriculum will include many courses which will be useless for statisticians. Your thesis will also likely not have much to do statistics. This is the portion of grad school where you really learn the most too. If one were to go the route of PhD in Applied Math would they be more employable in industry rather than just the PhD in Stats? Depends on the type of work I guess. For any type of biostatistics work, a PhD in Statistics or Biostatistics is almost always required. For finance and quant work, I'd still say the PhD in Statistics is a better choice, but then again, there are Physics PhDs doing this type of work too. I guess along the same lines as your emphasis on not doing a PhD in Biostats, because you can do it with a Stats PhD. I ask because I am interested in Stats PhD but am wondering if Applied Math isn't more practical. In my opinion, the Stats PhD is superior in terms of employability and salaries. That doesn't mean you can't do great with a PhD in Applied Mathematics (you can),.

Thanks in advance!

Edited October 1, 2010 by statsguy