Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: Effective HS Physics (was Statistics / more ...)



According to my references the Binet-Simon 1905 was credited as being the
first "modern" test designed to predict performance. 1916 Stanford-Binet,
followed by Otis in 1917 with the Army Alpha Test for the literate and Army
Beta Test for illiterate. Wechsler was 1939. This may or may not be
accurate. Otis is probably more widely used in schools, and the Wechsler is
confined to professionals such as psychologists and developmental
pediatricians (children's shrinks).

I am sure with quite a bit of iteration it is possible to make a test that
predicts success in an average sense for a group of children in a given
school system. Whether such a test is accurate over a longer time frame, or
with a wide sampling of students is certainly more problematic. Applying
such a test to an individual student is again difficult. It may be possible
to predict success within a range over a limited time span.

There are some tests that are actually fairly good predictors, but only for
limited results. Hestenes VASS has proven to be a fairly good predictor of
success in the introductory physics courses. The Lawson pretest is a good
predictor of the maximum achievable gain on the FMCE or FCI for an
individual student, but it does not predict the actual gain nearly as well.
David Meltzer has shown that some math tests have some predictive value, but
again individual student effort and attitudes can have a big effect. VASS
is measuring attitude, while the Lawson test is measuring thinking skills.
My data seems to show that the Lawson test puts a very strong limit on
achieving gain on some of the conceptual tests. While a good attitude can
push the student to achieve the limit, it seems that they can not achieve
more within the parameters of my course.

John M. Clement
Houston, TX



Actually, IQ tests were originally developed to measure the "intelligence"
of army recruits to determine what area they were most suited
for. The story
is long and somewhat convoluted as to how it became a measure of something
that we can't even agree what it is.

By the way, I'm a high school teacher who doesn't use a text. I guess my
students do OK at the college level, but I sure work way to hard. I'd like
to see a well written text with appropriate problems to solve and lots of
conceptual type questions. Oh well, I guess I'll have to writ one
of my own.
;-)

Steve Clark, Ph.D.


From: John Clement <clement@HAL-PC.ORG>
Reply-To: "phys-l@lists.nau.edu: Forum for Physics
Educators"<PHYS-L@lists.nau.edu>
Date: Sat, 05 Jan 2002 15:53:33 -0600
To: PHYS-L@lists.nau.edu
Subject: Re: Effective HS Physics (was Statistics / more ...)

I think that is right, as I read that IQ was originally
designed to predict
success in school. One of the big fallacies about IQ is that
it is fixed.
The research of Feuerstein has shown that it is possible to dramatically
change IQ, even starting as late as age 15. If the sort of
treatments that
are useful in changing thinking level were generally available,
the effect
of SES (Socioeconomic Status) might be lessened.

John M. Clement
Houston, TX



I thought IQ tests were a good predictor of college success --
that's how they're
designed. The problem is they're misnamed. i.e. that's all they do.
(indirectly [concomitantly] they may measure socio-econ,
status, english
comprehension, etc., but not I.Q. -- unless I.Q. is another
phrase meaning the
former.)

bc

P.s. What's SES?


Brian Whatcott wrote:

At 09:16 PM 1/2/02, Dan MacIsaac responded to this:
It is not PC to mention it, but an important predictor of
academic test results is that old faithful, the IQ measure.
This measure is designed to be normally distributed.
A high school may expect to see student results in
general that are
reasonably normal, even if test results are not 'marked on a curve'
[i.e. transformed to a normal distribution] for this reason alone.

Brian:

Check out Gould's book "The Mismeasure of Man" IQ is a very
cooked statistic
(as you say, "designed" for statistical reliability), but
after reading
Gould's book I rather believe it measures very little about actual
intelligence (whatever that means). If you really want an
indicator for
predicting student success, try SES. If you are looking to
predict success
in college physics, look for whether the HS physics teacher
used no text.
(after Sadler's recent _science education_ article)
Dan M

Dan MacIsaac

I hope that if one party mentions a non-PC predictor of
academic success (IQ)
and another party responds with a PC predictor of academic
success (SES)
then a reasonably conscious third party may see a statistical
connection
between the two measures.
Moreover: what red-blooded American could be against socio-economic
status? And a relevant question: Is SES normally distributed?

But I really hoped someone would pick up on your no textbook
proposition -
it is so intuitively appealing. The *next* time someone asks for a
physics book recommendation (an almost monthly event, I
sometimes think:)
I hunger to read that someone is advising:
- "dump it in the round file: list the physics topics you
have something
interesting to talk about - and go for it!"

Brian Whatcott
Altus OK Eureka!