Strengths and Weaknesses of PhD Statistics: Iowa State, NC State, and VA Tech

[ducky500]

If anyone is familiar with any three of these programs: Iowa State PhD Statistics, NC State PhD Statistics, and VA Tech PhD Statistics, could someone tell me their strengths and weaknesses as far as placement in terms of academia and industry. Do they have strong placements for Academia jobs (post-docs or Assistant Professor) and industry?

My worries (perhaps wrongfully): NC State and Iowa State are not particularly known for having 'strong' name recognition.  Do employers who are hiring statisticians know that both NC State and Iowa State have reputable statistics programs.  Or would a person who attends any of these three schools be at a disadvantage to someone, for example, attends a PhD of Statistics at UCLA, even though UCLA is of 'lower' rank, as UCLA has more name recognition or 'prestige'.  Are NC State and Iowa State, and Virginia Tech also recognized for academia or is that not the case.  Thank you.

[Stat Assistant Professor]

Those are all good programs, and NCSU is definitely a top tier program for Statistics. That is reflected in the USNWR rankings as well. These schools may have less "name recognition" overall than say, the Ivies, Stanford, Berkeley, Chicago, etc., but in the field of Statistics, they are well-regarded. When it comes to certain academic disciplines, individual programs can be more reputable than the school as a whole. For instance, University of Pittsburgh is a top 3 school for Philosophy and has academic placements on par with Ivy schools or UChicago, even if UPitt's overall rankings are not as high.

 

Also, the PhD advisor's reputation is more important. Students of any distinguished professor from any program should be in good shape to get a good postdoc. The top programs will have, on average, more professors who are leaders in their field, but the mid-to-lower ranked programs also have some good distinguished professors who are leaders in their respective fields (off the top of my head, I can think of Dr. Marianna Pensky at UCF, Cun-Hui Zhang at Rutgers, Peter Mueller at UT-Austin, etc.)

[ducky500]

Thanks for the response.  Do know by any chance any good Stat PhD Programs that are strong in applied probability models: stochastic processes and applications, markov chains, etc.. ? I've been struggling to find good programs in this field.  Thanks.

[Stat Assistant Professor]

Most of the top-tier Statistics programs have at least a couple very strong faculty who work in probability theory: Penn Wharton, Stanford, Berkeley all have some faculty who work in probability, e.g. Diaconis at Stanford, Bhattacharya at Penn Wharton, and Pitman at Berkeley. 

Below that, UNC-STOR is particularly strong in stochastic processes and applied probability (e.g. Bhadimi and Budhijara). For Markov chains, University of Florida is a world leader department in the area of MCMC theory, with world-renowned faculty who work on convergence rates and complexity analysis of Markov chains (e.g. Hobert and Doss).  

Edited January 4, 2019 by Stat PhD Now Postdoc

[Stat Assistant Professor]

Also wanted to add:

1) You can look at Alumni Placements for most schools that interest you to see how their PhD graduates fare. If this information is not available on the Departments' websites, you can e-mail the PhD program coordinators who should be happy to provide you with this info. 

2) If you're interested in stochastic processes, it may be worthwhile to look into Applied Mathematics and Operations Research & Industrial Engineering PhD programs as well. Some graduates of Math or ORIE programs do become faculty in Statistics departments (e.g. Princeton's Operations Research PhD program and Cornell's ORIE program have both placed their alumni in Statistics postdocs/Assistant Professorships).

[ducky500]

Hi, Stat PhD Now Postdoc Thanks for all the help.  Your information has been really valuable.  I have applied to both Phd in Stats and Operations Research. I think I will apply to University of Florida as well as they seem to have a lot of faculty working on topics related to stochastic processes after looking through their department website.

[CarolinaSmash]

16 hours ago, Stat PhD Now Postdoc said:

Most of the top-tier Statistics programs have at least a couple very strong faculty who work in probability theory: Penn Wharton, Stanford, Berkeley all have some faculty who work in probability, e.g. Diaconis at Stanford, Bhattacharya at Penn Wharton, and Pitman at Berkeley. 

Below that, UNC-STOR is particularly strong in stochastic processes and applied probability (e.g. Bhadimi and Budhijara). For Markov chains, University of Florida is a world leader department in the area of MCMC theory, with world-renowned faculty who work on convergence rates and complexity analysis of Markov chains (e.g. Hobert and Doss).  

Bhamidi is the GOAT.. Fav professor at UNC 

[Bayesian1701]

I had a bad experience with VA Tech but you may not.   I’ve written about it before but there completion rate is low (they admitted most people have to master out because they failed exams) and they do funding on a year by year basis. They said it was because they had to accept students who where really qualified and assured me I would be ok but I couldn’t take the chance of failing out when I had stable funding packages at places with higher completion and better placements. 

[bayessays]

NC State is by far the strongest and most well-known of those programs, and any statistics job worth having will know that it is among the best programs in the country. UCLA has a strong but small department. Iowa State is historically very strong, but has declined somewhat and the quality of the very large faculty varies more widely than any program I've ever seen, though some are very good. Virginia Tech is definitely the weakest of those programs by far, especially for an academic job, although they might have some industry connections. 

[ducky500]

Thanks guys.  How would you say Iowa State does as far as academic placements? Also although not mentioned in the title of thread, do you have any opinions on University of Florida's Program?

 

 

Edited January 5, 2019 by ducky500

[Stat Assistant Professor]

2 hours ago, ducky500 said:

Thanks guys.  How would you say Iowa State does as far as academic placements? Also although not mentioned in the title of thread, do you have any opinions on University of Florida's Program?

 

 

UF lost a lot of bigshot professors to retirement/death in the early 2010s, e.g. Casella, Agresti, Randles, Linda Young, Daniels, etc., so it became a smaller department whose main strengths were in niche areas (specifically Markov Chain Monte Carlo theory).

 However, recently, it has been on the rise again. Daniels returned to be the Department Chair, the department hired Michailidis from Michigan who heads the UF Informatics Institute, and the department has been hiring a lot of new faculty in machine learning and data science. I would put UF currently in the same neighborhood as schools like UCLA, UIUC, and Rutgers. 

[galois]

Sorry to hijack ducky, but while we're on the subject, I would love to hear takes on UT Austin - specifically its strengths, weaknesses, and reasons for the discrepancy between its rank # 50 and its perceived prestige (at least according to this forum). I've heard previously this is only due to its size, but there are other small programs that score higher than 50.

[bayessays]

49 minutes ago, galois said:

Sorry to hijack ducky, but while we're on the subject, I would love to hear takes on UT Austin - specifically its strengths, weaknesses, and reasons for the discrepancy between its rank # 50 and its perceived prestige (at least according to this forum). I've heard previously this is only due to its size, but there are other small programs that score higher than 50.

They are small and new (only have a couple graduates) and some of their professors aren't really members of the department officially, which might have some impact. They have some of the best Bayesians in the world (James Scott, Scott Walker). The rest of the faculty isn't as renowned, but there are some other top people like Carvalho and there's nowhere with nearly as many Bayesian non-parametric folks.  If coursework is important to you, they don't have the advanced theoretical stats classes you would get at a more traditional program. I would say if you want to go and have Scott or Walker as your advisor, you're in as good of shape.l as being a Duke grad, though.

[Stat Assistant Professor]

On 1/5/2019 at 1:36 PM, bayessays said:

They are small and new (only have a couple graduates) and some of their professors aren't really members of the department officially, which might have some impact. They have some of the best Bayesians in the world (James Scott, Scott Walker). The rest of the faculty isn't as renowned, but there are some other top people like Carvalho and there's nowhere with nearly as many Bayesian non-parametric folks.  If coursework is important to you, they don't have the advanced theoretical stats classes you would get at a more traditional program. I would say if you want to go and have Scott or Walker as your advisor, you're in as good of shape.l as being a Duke grad, though.

Agreed with this assessment. Carvalho, Scott, and Walker are all very well-known and renowned in the field of Bayesian statistics, so having any of them as an advisor gives you a very good shot at landing a top postdoc at a world-class institution (e.g. Duke or Columbia).

Edited January 7, 2019 by Stat PhD Now Postdoc

[Teaching Faculty Wannabe]

On 1/3/2019 at 3:19 PM, ducky500 said:

Are NC State and Iowa State, and Virginia Tech also recognized for academia or is that not the case. 

I am not in statistics, but I know two people who did their PhDs in Stats at NC State and I also did my undergrad at NC State. NC State has strong connections with SAS, which is a large software company. SAS funds a lot of projects at NC State and I believe there are a lot of collaborations with the two (https://news.ncsu.edu/2014/08/sas-agreement/). The Statistics Department is located in SAS Hall. The program might not be super well-known but I think having that connection with SAS is good and NC State is a top research school in general. The two graduate students I know seemed to like the program and they were able to work on multiple projects while they were there. Additionally, NC State is part of the Research Triangle, which allows for a lot of collaborations with various companies, universities, and government agencies in the area. Also, Raleigh is a really cool place.