<p>As is obvious from my handle, I am a class of 2006 graduate of the MIT mathematics department (UG).</p>
<p>From my experience, it is unusual for an MIT math undergraduate to to be accepted into MIT’s math Ph.D. program. I only saw a handful of students do it each year. It isn’t that MIT feels its undergraduates are ill prepared for the graduate program, or that MIT does not “want” them. As I understand it, the rational is a relatively common one among certain departments (and not only at MIT): The faculty wants their graduates, by and large, to go off and experience mathematics under a different regime, so to speak. They feel that a differnt perspective and culture of mathematics is the best way for their students to grow as mathematicians.</p>
<p>I think the practice is justified. I woud wager that not <em>some</em> but <em>most</em> MIT mathematics undergraduates who have serious aspirations of pursuing math at a graduate level already have significant exposure to both MIT’s graduate curriculum in their field as well as its research activities by the time they graduate. You can never run out of things to learn at MIT, but I do think it is fair to say that going elsewhere for graduate school exposes MIT math grads to a greater variety of genuinely “new” experiences.</p>
<p>From my exposure, it seems that the students who were selected to continue on into MIT’s Ph.D. program were not selected for being the “best” undergraduates. Some of the most brilliant students that I knew went elsewhere and were encouraged to do so. The ones who stayed were generally students who had strong promise but, more importantly, had research interests that were very well aligned with their desired advisors’. I should go further to say that in perhaps <em>all</em> cases, these students had existing, strong research ties to their advisors (existing joint papers, projects, etc.). It seems that the math department allowed students to stay only if MIT was fairly conclusively <em>the</em> best place in the world for them to carry on with their existing research.</p>
<p>So, to get to my point: If you want to do a Ph.D. in math at MIT, I would recommend going elsewhere for undergrad…simply because doing “your best” is not really the formula for transitioning from MIT UG to MIT G (in math). And, you certainly wouldn’t want to keep your interests from developing naturally such that you would be a strong candidate for the Ph.D. program. Let your mathematical interests guide you, not your graduate school aspirations.</p>
<p>This is just my two cents from experience. And, lest people get the wrong impression, this is certainly not the way that most MIT departments operate. Not withstanding a few select departments, it is very common for MIT undergraduates to continue into MIT graduate programs.</p>
<p>*In discloser, I pursued graduate study in a field not offered a MIT (and not in mathematics). So, the above was a moot issue for me. Having not participated in the MIT grad application process, my perspective is from observation, not direct experience. So, take that for what it is worth.</p>