An interesting post from the political science blog The Monkey Cage. Some folks at Syracuse University looked for the best predictors of graduate student success in their Economics program, using as data the records of their past students.
The upshot: the best predictor of passing comps: your GRE scores, having an M. A. degree, and having an economics major.
However, the best predictors of completing the Ph. D. program were different. They combined something very concrete–a student’s preparation in mathematics, with something far more intangible–the strength of their research motivation, as gleaned from the personal statements in their application to grad school.
My take, for what’s it worth.
In most programs, your written and oral comprehensive exam tests your mastery of a set of material, and, indirectly, the work habits that allow you to teach yourself.
But there is a reason grad school is often described as a process of transformation from someone who reads to someone who is read.
Because the second hurdle is the ability to design, and analyze novel research. This takes a mastery of logic (and mathematics is, at its heart, logical calisthenics) and a stubborn drive to see a project through. Those students who work on mastering the literature, but avoid the often lonely process of designing, collecting, and cranking methodically through data and manuscripts, risk the scarlet ABD of academia–All But Dissertation.
We’ll be spending some time in the upcoming posts working on your analytical chops. But in the meantime, don’t skimp on math. Even if your dissertation never requires matrix algebra or integration and differentiation, studying math trains your mind to think clearly.
Posted by pheidole 
Survive and Thrive in Grad School 