Jump to content

Soliciting opinions on the relevance of school ranking.

Terry O'

Recommended Posts


I am applying to graduate school after an extended period working in the computer industry. I have pretty good GREs( 168Q,163V,4.0A), 13 years relevant work experience but a low undergrad GPA 3.1 (athletics and women were my real major). I'd like to teach at the university level and thus will be pursuing a PhD. It seems the university you do your graduate work at greatly effects your chances of getting that first prof. gig. 

My question is, do I have to attend a top tier university (ranked top 15 US news) in order to succeed? It seems likely I'd be excluded from these programs based on my GPA. Honesty, I'm not sure I'll be admitted to a top 30 school either. If I attend a program outside the top 50, is the paper even going to be allow me to pursue my goal?

On a related note, I'd be really nice to find a repository of GPA/GRE distributions for those admitted to grad programs, not just averages. It'd really help some of us make informed decisions about where to apply.

Thanks in advance, 


Link to comment
Share on other sites

This is a post taken from Academia StackExchange, a user named JeffE who is a professor at UIUC.



Let me answer as a theoretical computer scientist with former PhD students in tenure-track academic positions and many years of experience on faculty hiring committees. (However, my understanding is that the selection process at industrial research labs like IBM T.J. Watson, Microsoft Research, Google Research, AT&T Research, etc., is really not that different from academic recruiting.) As always, take my advice with a grain of salt; I'm as guilty of confirmation bias as any other human being.

Nobody in theoretical computer science cares where you got your degree. Really. We. Do. Not. Care. We only care about the quality and visibility of your results. Publish strong papers and give brilliant talks at top conferences. Convince well-known active researchers to write letters raving about your work. Make a good product and get superstars to sell it for you. Do all that, and we'll definitely want to hire you, no matter where you got your degree. On the other hand, without a strong and visible research record, independent from your advisor, you are much less likely to get a good academic job, no matter where you got your degree.

(This is less true in more applied areas of CS, in my experience, mostly because it's significantly harder for PhD students in those areas to work independently from their advisors.)

But. Faculty candidates are necessarily judged by people who are not experts in their field. Without the expertise to judge whether your work is really good, those people must look at secondary data that correlate strongly with successful researchers. One of those secondary characteristics is "pedigree". Did you get your degree at MIT, Berkeley, Stanford, CMU, another top-10 department, or somewhere else? (What's an "Ivy League"?) How good/famous is your advisor? If they're really paying attention: Where did your advisor's other PhD students get jobs, and how well are they doing now?

Fortunately, most good departments do make a serious effort to understand the quality and impact of applicants' results, instead of relying only on secondary data. Also, secondary data matters considerably less once you actually have an interview.

And. In my experience, where you get your degree is strongly correlated with successful research. I got my Master's degree at UC Irvine in 1992 and my PhD at UC Berkeley in 1996. The biggest difference I saw between the two departments was the graduate-student research culture. Every theory student at Berkeley regularly produced good results and published them at top conferences. When the FOCS deadline rolled around each year, the question I heard in the hallways from other students was not "You know the deadline is coming up?" or "Are you submitting anything?" but "What are you submitting?", because "nothing" was the least likely answer. Everyone simply assumed that if you were there, you were ready and able to do publishable research. Publishing a paper wasn't exceptional, it was just what you did. That cloud of free-floating confidence/arrogance had a huge impact on my own development as a researcher. I've seen similar research cultures at a few other top CS departments, especially MIT, Stanford, and CMU. (Caveat: This is an incomplete list, and there are many departments that I've never visited.)

tl;dr: Yes, getting a PhD from a top department definitely helps, but more by helping you become a better researcher than by making you look better on paper.


So, to answer: No, your work ethic is what matters. Top schools will help foster you to become a better researcher than most schools, but they do not dictate your success. 

Edited by complexbongo
Link to comment
Share on other sites


Excellent find! Thanks for the input. I'll take heart for my admittance prospects and keep my head down once accepted. Good luck in your search.


Link to comment
Share on other sites

I think the answer is a bit more complex than "no, your work ethic is what matters."


I think it depends on where you want to work.  If you want to be a professor at one of the top universities, then you'll need a PhD from one of the top programs.  I took a look at the first 25 professors (alphabetically) in MIT's computer science and electrical engineering department:


Berkeley, MIT, Stanford, Stanford, Stanford, Stanford, Stanford, Minnesota, Berkeley, MIT, Oxford, Columbia, Princeton, MIT, MIT, MIT, Harvard/MIT (MD/PhD), MIT, MIT, Stanford, MIT, Berkeley, Columbia, MIT, Berkeley.


Out of this 25 (which took me through the Cs), literally 20 out of 25 went to one of three schools - MIT, Stanford, or Berkeley.  Those schools are ranked in the top 5 CS programs.  The other programs are all within the top 30. (For all the "what is the Ivy League?" talk, Princeton, Harvard, Penn, Cornell, Columbia, Dartmouth, Yale, and Brown are all in the top 25-30 CS programs, as ranked by the NRC.)

But I then took a look at the CS departments at Case Western (middle of the pack), Temple (bottom of the doctoral programs), my own mid-ranked SLAC and a lower-ranked SLAC in the same city, and the results were more mixed.  Case Western and Temple both had mixed departments, with a few professors from top 30ish programs but most professors came from lower-ranked programs.  My alma mater (the mid-ranked SLAC) is located in a desirable city, so both of the professors in that small department come from the same institution - a top 20 program that happens to be in the same city.  The lower-ranked nearby department has a larger department and also more of a mix.  Also took a peek at a directional state U nearby my hometown, and that was a mix too.
So it appears that if you want to teach in a top department, like MIT, Stanford, Caltech, Berkeley, Ohio State, Chicago, Arizona State, etc. you'll probably need a degree from a top department.  If you want to teach in a nice solid mid-ranked department, like RPI, Virginia Tech, Oklahoma, JHU, Mississippi State, etc., then you can probably have a PhD from a mid-ranked department.  If you're just interested in teaching in general - especially if you want to, or wouldn't mind, teaching at smaller rural schools, directional state comprehensive, 2-year schools, etc. - then a PhD outside of the top 50 or so probably won't hurt you in your quest.  It's just that the pool of schools that are realistic for you to shoot for is smaller.
Caveats, of course, are subfield, accomplishments, and advisor.  If you go to UConn but have a superstar advisor and publish your butt off, you may be competitive for jobs within the top 30-40, especially if you got a prestigious postdoc or something.  Or perhaps UConn is mid-ranked for comp sci overall but really highly ranked for *your* specific subfield of comp sci.
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Create New...

Important Information

This website uses cookies to ensure you get the best experience on our website. See our Privacy Policy and Terms of Use