Each fall I teach a mini-course for our physics majors who want to take the Physics GRE. The trouble is that you can't do well on the test if you just go in and

Each fall I teach a mini-course for our physics majors who want to take the Physics GRE. The trouble is that you can't do well on the test if you just go in and try to solve as many questions as you can: it's 100 questions in 170 minutes, and many (most?) of the questions look like full fledged homework questions (that would take even a pretty good student 5-10 minutes to solve, or sometimes much longer). Instead, you need to learn all of the common tricks and shortcuts that the test writers expect you to know and use (including things like checking the units on all of the multiple choice answers to eliminate some choices, or testing that the choices have the right limits like time->0 or mass->infinity). It's a pretty obnoxious test, to be honest. Once you've learned the types of strategies to use, the GRE is testing something meaningful about physics knowledge and intuition, but I don't know how well that something correlates with either "successful in classes" or "successful in research".