UNC vs Penn State for PhD in Statistics

[Joanna Tran]

 

Hello all, 

I am considering between penn state and unc for statistics PhD. For Penn State, I got university fellowship, which means I don't need to do TA/RA job in the first year and do TA/RA duty in the second year with $32000 a year (year 1 and year 2 only). Penn state also guarantees that I will get summer stipend ($8000) if I stay on campus. For UNC, they don't have university fellowship for statistics department this year. Thus, I only get 21000 for 9 months with TA jobs. They also give me the first year research fellowship for $9000 and I have to work in the lab. That means, at UNC, I have to do TA and lab work in my first year and only do TA (only 21000) for upcoming years. 

I am considering these two schools because they are both top 20 in stats (UNC tie 11th and Penn State tie 19th). However, I am not sure which school is better for statistics. I am also thinking that Penn State gives me really good fellowship, while UNC does not. 

Can you all please give me the opinions of Statistics programs at these two schools? Although Penn State has lower overall ranking than UNC, these two schools seem to be quite similar and equal in statistics program. Am I correct? Should I choose the school with higher ranking or the school that really wants me and gives me good fellowship? Does Penn State lean towards applied and computing things and UNC lean towards theory more? 

P/s: I prefer to do research about data analysis and computational statistics since I prefer to work as a data scientist in industry. Thank you!

[bayessays]

I think your assessment seems accurate - there is not a big enough difference between these departments where there is a clear choice to make based on their ranking alone.  Especially if you're interested in doing applied work and working in industry, I don't think you'd see a difference in your outcomes at the two programs.  For doing computational statistics, I know PSU has some good people like Murali Haran.  They also have some cool applied areas (in the past they had astrostatistics, personalized medicine groups, etc).

$11k+ a year is a lot of money - I'd look at apartments in the area too and see about cost of living.

I think UNC could make sense if you want to set down roots, since you could theoretically stay in the area with the industry jobs in the triangle.  The weather is also a little warmer. 

[Joanna Tran]

21 hours ago, bayessays said:

I think your assessment seems accurate - there is not a big enough difference between these departments where there is a clear choice to make based on their ranking alone.  Especially if you're interested in doing applied work and working in industry, I don't think you'd see a difference in your outcomes at the two programs.  For doing computational statistics, I know PSU has some good people like Murali Haran.  They also have some cool applied areas (in the past they had astrostatistics, personalized medicine groups, etc).

$11k+ a year is a lot of money - I'd look at apartments in the area too and see about cost of living.

I think UNC could make sense if you want to set down roots, since you could theoretically stay in the area with the industry jobs in the triangle.  The weather is also a little warmer. 

Thank you for your opinion! How about UNC? Would it be better to do computational and applied stats at UNC? Another thing is that UNC seems to have smaller stats department, and I am not sure if it's good or bad side. You are right, $11k/year is a lot and PSU pays me more money for less work in the first two years, which is really great. However, if UNC is better in stats and UNC could give me more job opoortunities, I probably choose UNC. But, I heard some people saying that PSU has huge alumini network, leans towards applications, and I can collaborate with CS or EE people as well (eventually, PSU is a great school for engineering). So even after looking at research areas and faculty of these two schools, I am still not really sure. 

[bayessays]

No, Penn State's placements are great.  It seems like you for are putting too much stock in the rankings.  Is there a single reason you can think of that would make you want to go to UNC?  Because "jobs" or "it's a better program" are not valid reasons, because they're not true.

Weather, location, smaller program, unique research interests (UNC has lots of probability people, someone doing object-oriented data analysis, and fiducial inference, which are unique research areas), wanting to live in the Research Triangle forever - these are clear reasons to go to UNC.

Go to each school's website, go to every single faculty member's website, and look to see what they're working on and make a list of how many profs at each school sound interesting to you.  

[Joanna Tran]

Just now, bayessays said:

No, Penn State's placements are great.  It seems like you for are putting too much stock in the rankings.  Is there a single reason you can think of that would make you want to go to UNC?  Because "jobs" or "it's a better program" are not valid reasons, because they're not true.

Weather, location, smaller program, unique research interests (UNC has lots of probability people, someone doing object-oriented data analysis, and fiducial inference, which are unique research areas), wanting to live in the Research Triangle forever - these are clear reasons to go to UNC.

Go to each school's website, go to every single faculty member's website, and look to see what they're working on and make a list of how many profs at each school sound interesting to you.  

I was chosen to work in EPIC project (brain network analysis) at UNC in the first year for extra $9k, which is interesting to me. However, they said they cannot guarantee that I can work in that project in the coming years depending on funding and professors. So, I am not sure if I should risk it. I found some profs that I would like to work at UNC, but I found slightly more profs I would want to work at PSU as well. Moreover, there is one postdoc stats student telling me that both programs have kinda old research directions in applied and computational stats (i am not sure if it's true). So, right now, I am a bit leaning towards PSU. But it would be great if I know more about bad cases from both programs, like are there any reasons to not choose psu over unc or to not choose unc over psu in my interesting fields (computational stats and data scalability),