Hi Annette,
I just got some stats from ETS on predicting graduate GPA.
Correlation
between graduate GPA and each of the three sections of the GRE (verbal,
quantitative, and the now discontinued analytic) is about .3. Use all three
and R = .34. Use undergrad GPA alone, r = .37. Use GRE and undergrad GPA
and get R = .46. Interestingly, the subject test is a better predictor than
the general test, but few schools require the subject test (I was in the
99th percentile on the subject test, by the way, but only 75th on general).
In Psychology, for
example, the r between graduate GPA and the subject test is .37.
Several years ago my department participated in a validity study.
Graduate students' performance was best predicted by their verbal and their
analytical scores, with quantitative scores adding next to nothing. When
predicting performance in a three-semester sequence in statistics and one
semester of psychometrics the results were almost the same, with the
analytical scores being the best predictor. Since the stats courses were
taught by me, the predictive value of the analytical scores may say more
about me than about my students. :-)
Cheers,
Karl W.
----- Original Message -----
From: "Annette Taylor, Ph. D." <[EMAIL PROTECTED]>
To: "Teaching in the Psychological Sciences" <[email protected]>
Sent: Saturday, January 21, 2006 11:26 AM
Subject: Re: GRE Norms
Good Morning Karl:
I don't know the answer to the norms question but I wouldn't be at all
surprised
if the annual mean has shifted. Over the years there has been much more
emphasis
on quant achievement as a better estimate of 'aptitude'.
I do much advising of undergrads on getting into grad school and over the
last
20 years have attended as many panel discussions and forums at professional
conferences as I could, to find out what it is that grad schools are looking
for. Without exception, when the GRE comes up, it is the quant score that is
emphasized as more important than the verbal.
Thus, I suspect students are preparing better for the quant exam, AND I
believe
that it is actually easier to "prepare" for that part of the exam. You can
certainly cram a good review of basic and advanced college algebra, by
taking a
good review course; but the verbal skills, I think (no data) would
require much
more time to build. For example, there are tricks for vocab, like learning
latin and greek prefixes and suffixes, but bottom line: correct grammar and
syntax are more likely to reflect life long habits; and much vocab, as
assessed
on such tests, is idiosyncratic.
Of course I am fairly certain that Robert Sternberg has published data
showing
that the GRE is, in fact, not a very strong predictor of success in grad
school. Other factors, such as creativity, are more important.
Annette
---
You are currently subscribed to tips as: [email protected]
To unsubscribe send a blank email to [EMAIL PROTECTED]
