I'm glad Soluyanov that you can distinguish between real intellect and not and that you know the difference. I'm not sure a lot of
people know the difference. I'm happy that guy is doing
so well, but sad as an American that we lost the potential to have
him in the US just on account of that test. That seems stupid
on the part of the US schools to set up a system where he
would be discouraged from applying.
That's interesting Soluyanov would expect high scores from a native language verbal test. For the Americans, lots of the words on the verbal are not frequently used even for well-educated people. The funny part of that (and this is only my theory): I think Pres. Bush must have hired speechwriters that had that kind of GRE verbal vocabulary normally whereas he did not. So, every time he made a speech he stumbled on the big words and it made him look like an idiot repeatedly. Talk about getting creamed on verbal!
Everybody seems to claim the GRE physics is for uniformity, but
everybody is also saying how NOT uniform it is: the test can
differentiate between foreign and American, liberal arts, gender,
good time planners, multiple choice training and how much time you had to prepare. Maybe it forgot to test for physics since its testing all these other things
I think though the American's are not that used to multiple choice
as one might think. The multiple choice we always had was
4 answer choices and two are nonsense and the choices are
widely separable. On this test, 5 answer choices, not nonsense
that is immediately recognizable, and the choices are in some
cases numerically close together (.254, .255, .256 ...). There
are some questions that even had a 2 in the numerator, the
denominator, or in the square root factor and when we learned
to 'handwave' and derive formulas we might have handwaved the
2 out of there. So, its really the mother of multiple choice tests
even for Americans.
Paul said: Knowing how to discriminate between realistic and unrealistic solutions, take limits, look at units, devise quick ways to solve problems, and use mathematics to decompose a question whose physical basis is completely unfamiliar to you, all seem to be good skills that most physicists employ at times, even if these skills are secondary to actually knowing the advanced material and understanding it.
approximations, back-of-the-envelope calculations, dimensional analysis, and physical intuition,
I think the "use mathematics to decompose a question whose
physical basis is completely unfamiliar" is exactly the formula
pushing a good professor would have trainined us not to do -
that some people can do this really well and not understand
the physics. Regarding "back-of-the-envelope calculations" -
it's a true statement to say: When I do research I really do
20 back-of-the-envelope calculations per day. But the
converse is not true. The essence of research or doing
physics is not in the back-of-the-envelope calculations. To me its
more in the physical insight and concept understanding. So, testing the mechanics of the things we do while doing physics isn't the same as testing our physics understanding, insight, and intuition.
I hope people saying the best foreign students are not
designated the best on the basis of just this test because
then you are doing what you want the schools not to do -
just base things on one test. I think the schools and
foreign students like to create images of superiority. I
want very much to work with great scientists from all over,
but will not bow down to anybody being superior by some
sort of class structure we just made up and above all else
not from this test!
My personal belief is that the system of acquiring repositories of previous questions is much more efficient in group oriented societies - the brain dumping of questions after the test
for years and years is done better in certain places. I don't
think that accounts for all the perfect scores, but at least some
of them. And if you had to have seen the questions before,
how is that testing our ability to do physics?
So, my real question is, no matter what your score is, does the
score accurately reflect your ability to do physics and/or
physics research, and/or be successful in graduate school?
I still have to say I'm not convinced.