Advice on MIT, Georgia Tech, and Duke for grad school

Philip Greenspun's Homepage : Philip Greenspun's Homepage Discussion Forums : Ask Philip : One Thread
Notify me of new responses
Hi Philip,

I am currently trying to decide on a grad school for my PhD and could
really use some advice! I have been accepted into Duke and MIT's
mechanical engineering program and Georgia Tech's Robotics program. I
realize I have really good options but each one seems to have its
pros and cons which is making my decision difficult. I received
prestigious fellowship offers from both Georgia Tech and Duke. At
Duke, I know who I would be working with and have a pretty good idea
at Georgia Tech. I like my potential advisers and research projects
at both schools. However, at MIT I have no idea who I would be
working with which concerns me. Overall I liked the atmosphere and
people at Georgia Tech and Duke better.

My professors tell me I should choose MIT without a doubt and would
be crazy not to. However, I have heard from so many other people
that adviser is the most important consideration. What advice can
you give me in this situation? Should I go to the MIT because of the
prestige of the school even though I have no idea who I would be
working with or what I would be doing? Or should I go to either
Georgia Tech or Duke since I am pretty certain about the
research/adviser situation at these schools? Does the reputation of
the school matter as much since I am planning to go into industry
eventually? Would I be crazy to turn down MIT? Thank you for your

-- Mariah S, April 3, 2018


Congratulations! This is a great problem to have.

You're right that grad school is all about which advisor. There are plenty of great researchers who are bad advisors, i.e., who can't tell a graduate student "You can be successful in Academia by doing X, Y, and Z."

The fellowship is also nice. If you're at MIT and your professor runs into funding difficulties, a risk that you can't control (you don't work at a government funding agency), you might be cast out into a teaching assistantship.

Getting a job in Academia is all about the job talk and presenting what you accomplished. Also about papers that you might publish during graduate school. So if the Georgia Tech faculty can lay out a road map for you to publish that is a strong indication that you'll be successful there.

Oddly I think that your aspiration of a job in industry might favor the MIT choice. A company is more likely to be impressed by the MIT name than another university (which will ask "What has she published?").

MIT Mech E. certainly has some great people, but I'm sure that you can find some folks stuck in dead ends at MIT. MIT in the old days was a "sink or swim" atmosphere and the assumption was that graduate students could take care of their own careers. The result was a lot of protracted PhDs. Did you ask at Duke and Georgia Tech how long they expected your program to last? That's one important factor. If you get out a year earlier because of more intensive mentoring that's $150,000 in your pocket (difference between grad school stipend and industry salary). Well, maybe $85,000 after tax, but you get the idea.

If you come back to MIT to visit with some more labs and professors, let me know and I'll take you on a helicopter tour!

-- Philip Greenspun, April 4, 2018