Prestige vs. Fit

[Cavalerius]

In anticipation of the receipt of application decisions, I have a question concerning how to regard the prestige of a school versus the value of research fit. To begin, I think that I am a fairly strong candidate for most doctoral statistics programs. To determine the schools to which I applied, I searched for schools where at least three or four faculty members were engaged in a line of research that I would be interested in pursuing. Accordingly, the list of schools to which I applied included a number of schools not ranked in the top tier of schools—at least according to comments that I have seen on people’s applicant profile evaluation posts—although they are, by virtue of my selection process, schools that are a solid match for the research that I would like to do. While I have been told that once you enter a program, you can attempt to dictate the course of your research even if your research interests are not exactly concordant with those of faculty at the school, I am nonetheless skeptical of this advice; however, when I read comments on this board saying that a certain school should be beneath the consideration of an applicant given the strength of the applicant's profile, I wonder whether it would behoove one to prioritize the reputation of an institution above all. Has anyone else experienced this dilemma? If so, assuming you were admitted to both, which did you choose—the more prestigious school with a weaker research fit or the less prestigious school with a stronger research fit?

Edited December 29, 2018 by Cavalerius

[Stat Assistant Professor]

Prestige is something you should certainly take into account but it is not the only thing. It is probably a good idea to visit the programs you are admitted to in order to get a sense of where you would be most comfortable. In addition, even if a program is ranked "below" the top tier, it likely has niche strengths that are very strong, e.g. Rice University and UCSB Statistics are very strong for financial mathematics, while OSU and University of Missouri are both strong in spatial statistics. So if you are interested in a niche area, it makes more sense to go with the program that has particular strengths in that area.

In addition, your PhD advisor matters more than the program. I attended a mid-tier program (in terms of overall USNWR rankings), but I worked with one of the most distinguished professors who has great PhD placements... former PhD students of his are now faculty at Duke, TAMU, University of Georgia, etc., and from 2012-present, this PhD advisor has had his students taks postdoc appointments at Stanford, Carnegie Mellon, Columbia, Penn, etc. So while the prestige of the institution is something to take into account, your success is ultimately up to you. Your publications and recommendation letters are what carry the most weight in academic hiring for Statistics. 

[theduckster]

Stat PhD's answer is great.

I want to add that if you are set on joining industry, prestige matters a lot more than it does in academic hiring. 

Edited December 29, 2018 by theduckster

[cyberwulf]

Here are a few good reasons to choose a "prestigious" program:

1) It's very difficult for students (and even faculty) to gauge the quality of a faculty member "at a glance". Having a large number of faculty working in one area doesn't necessarily mean a department has strength in that area. Faculty at higher-ranked departments are, on average, "better" than faculty at lower-ranked places, so you don't have to rely as much on your own judgment.

2) Targeting a particular faculty member (or small group of faculty) when deciding on a program is risky. Faculty may leave/retire/change research areas/go into senior administration/etc. This is a particular concern at a lower-ranked place; if a faculty member is truly excellent, they are more likely to be "poached" by a higher-ranked place, asked to serve in higher admin at their current institution, etc.

3) Following from the first two points, higher-ranked programs generally offer better "fall-back" options if your original research plan doesn't pan out. At top places where most faculty are field leaders, switching advisers and research areas won't typically have a big impact on your future prospects.

4) The student cohort matters. Not only is it motivating and inspiring to be around other top-tier students, but an underappreciated benefit of going to a top program is the academic network that you build from being there. Knowing people in academia means you are more likely to be invited to give talks at other institutions and at conferences, more likely to be asked to be a reviewer or associate editor for a journal, etc.

And, a few reasons not to:

1) Sometimes it's better to be a big fish in a small pond. If you're one of many excellent students at a good place, it's easier to get lost in the shuffle.

2) Top places tend to be pretty big, and you may prefer a smaller environment.

3) Faculty members at prestigious programs are under a lot of pressure to maintain a high level of productivity. This can create an intense environment that may be sub-optimal for student learning/advising. For example, advisers may feel that they can't be as patient with you if they need to get stuff finished and published quickly to be competitive in their field.

[Gauss2017]

Those were two excellent posts by cyberwulf and stat now.  Faculty at prestigious institutions may be under a lot of pressure to publish for reasons other than tenure.  It could be because of ego, salary bumps in the future or for getting. consulting gigs in the future. Also top professors may travel a lot or simply have a lot of commitments.  I would generally say going with the prestigious institution is a good move but there are always exceptions .

[Stat Assistant Professor]

cyberwulf makes good points. If I were hypothetically applying to Stat PhD programs again and were admitted to a Harvard, Berkeley, UChicago, Penn Wharton, CMU, etc. (or a JHU, Harvard, UWashington for Biostat), I would probably be inclined to go with one of those. Below the very top tier, I would probably be a little bit more choosy though. For instance, I quite admire some of the smaller programs which are quite strong, in my opinion, like UIUC, Rutgers and Yale -- they have some truly pioneering faculty like John Lafferty and Cun-Hui Zhang. I do not know for certain that I would pick a more "prestigious," large state school program over one of these smaller but very solid programs.

Just my opinion though. Others may disagree, and of course, it is always easier to evaluate things in hindsight after you have gained more experience and knowledge of the field. ?

[Cavalerius]

Thank you all for the responses, which have been immensely helpful in illuminating the significance of prestige in choosing a doctoral statistics program. It still seems strange to me when people make judgements based purely on rankings, such as expressing consternation at being accepted by higher ranked programs and rejected by lower ranked ones. In general, however, I can now see why, ceteris paribus, a higher ranked program might be preferable to a lower ranked one. I especially appreciate the detailed and measured responses of cyberwulf and Stat PhD.

[galois]

On 12/29/2018 at 8:08 AM, Stat PhD Now Postdoc said:

e.g. Rice University and UCSB Statistics are very strong for financial mathematics, while OSU and University of Missouri are both strong in spatial statistics

Is there a thread or some singular location where I can read about which departments have which niche strengths? Is this something that just changes year to year?

[Bayesian1701]

On 12/28/2018 at 11:16 PM, Cavalerius said:

While I have been told that once you enter a program, you can attempt to dictate the course of your research even if your research interests are not exactly concordant with those of faculty at the school, I am nonetheless skeptical of this advice; however, when I read comments on this board saying that a certain school should be beneath the consideration of an applicant given the strength of the applicant's profile, I wonder whether it would behoove one to prioritize the reputation of an institution above all. Has anyone else experienced this dilemma? If so, assuming you were admitted to both, which did you choose—the more prestigious school with a weaker research fit or the less prestigious school with a stronger research fit?

 

I just finished the first semester of my Ph.D. program.  I have a very narrow research interest and had no great advisor options really anywhere I got in.  I was in a similar spot and while I won't say what I chose in the end, I'll offer some advice on what to consider.  I thought about everything: prestige, stipend livability,  location, happiness, qualifying exam process, and required coursework.  No one knows what the best decision for you is like you do and there really isn't a blanket answer for this.  However, forging your own research path is hard.  It was incredibly stressful to have two advisor options and one of them is only an ok fit.  I considered transferring multiple times or dropping out entirely.   I had the impression that I wouldn't be accepted if I was a bad fit,  but I wonder if my research interests were not taken seriously. 

1 hour ago, galois said:

Is there a thread or some singular location where I can read about which departments have which niche strengths? Is this something that just changes year to year?

Not that I know of.  if you have an area in mind you can look at profile evaluations people did in your area and look at where people suggested.  Some users are much more knowledgable than others. More experienced users @Stat PhD Now Postdoc @bayessays know many niches, but other newer people like me know only certain niches (mine is Bayesian and southern programs).

[Cavalerius]

Thanks for the advice, Bayesian! I am sorry to hear that you are having a tough time in your chosen program. Interestingly enough, it seems that you and I had very similar approaches to the process of selecting programs, so your insight is particularly germane to my situation.

[Bayesian1701]

11 hours ago, Cavalerius said:

Thanks for the advice, Bayesian! I am sorry to hear that you are having a tough time in your chosen program. Interestingly enough, it seems that you and I had very similar approaches to the process of selecting programs, so your insight is particularly germane to my situation.

I think I have it figured out but it was an incredibly stressful time advisor searching for my research credit in the spring.  If you are deadset on a particular area make that clear at your visits and interactions with faculty.  That was one thing I didn’t do but I regret. Once you are accepted don’t be afraid to ask tough questions and know what you really want. I was way too shy. 

[Cavalerius]

Okay, thanks for the advice, Bayesian! 

I also have an additional question related to my previous one about prestige, one that touches on another recent post about the prestige of an individual department versus the prestige of an institution. One of the things that I noticed when looking at different programs was that at certain schools, many of the departments related to statistics--such as departments of math, comp sci, econ, industrial engineering, etc.--received higher rankings than did the statistics department itself. Some of these programs, moreover, feature interdisciplinary seminars and dissertation advising across the departments, with a few even having specializations that directly combine the two graduate programs. What is the general opinion of such programs, especially if the particular area of statistics that one wishes to study has a direct interface with a higher ranked program at the institution? I know that every institution is different, but in general, would this be something that should affect one's decision in choosing a program?

Edited January 10, 2019 by Cavalerius

[Gauss2017]

I think some of the posters may be able to give much better advice if you wouldn’t mind sharing some more specifics about your qualifications. 

[Cavalerius]

Well, when you look at, say, NC State and UNC, both have programs in their statistics department related to financial statistics, but UNC's business program is ranked considerably higher than NC State's. I realize that the professors that one works with are a crucial aspect of one's graduate record, but if one will be taking classes and potentially working with professors in both the business program and the statistics program, would choosing a program based on the school's joint reputation across these fields be reasonable, rather than just focusing on the school's statistics ranking?

A more pointed comparison would perhaps be between Rice and NC State, where the business school of the former has the higher ranking while the statistics department of the latter has the higher ranking. So if, for instance, one wished to focus on financial statistics, would such a situation give Rice the edge over NC State, despite the latter's prestige in statistics. 

Edited February 8, 2019 by Cavalerius