NEW STANDARDS TO BE LINKED TO STANFORD 9 TEST
\doc\web\98\06\newstan.txt
Date sent:      	16 Sep 98 10:09:23 EDT
From:           	Gary.M.Degasta@Dartmouth.EDU (Gary M. Degasta)
Subject:        	Re: Barcoding
To:             	arthurhu@halcyon.com


Arthur, I was referring to the Harcourt-Brace efforts to link the New
Standards Reference Exams to the Stanford 9 Achievement tests. They used a
control group of 3500 students, and gave them both exams. They are trying
to validate and compare the two so that our Governor can say that this is
how we compare on a national level. They did an experiment, basically,
taking 1200 4th graders, 1800 8th graders and 500 or 600 10th graders....
somewhere in the nation. Our DOE contact could not tell us how many states
participated in this national sampling. Then they took Vermont's scores on
the New refernece Exams and linked them to this study. This gave our
Governor the ammunition to say that on the New Standards Refererence
Exams, our state did performed better in a national comparison. What we
are attemtping to investigae is how realiable this sampling is and its
validity. When I told Dr. Rosenkrantz this, he called this attempt by the
NCEE voodoo psychometrics. I was,'t referring to the SAT's or anything

else. I thought Fran Rice had outlined this for you.


Anyway, we are still researching this and hope to expose it if it is
bogus.


Fran




I was wondering how Vermont could claim a national comparision.
That's a big problem with the state sponsored tests that don't give a
national comparison. It's conceivable that New Standards is different
enough that there isn't a high correlation, though most studies show
"performance-based" tests give exactly the same rank-order results as
traditional tests for 10 times the cost and inaccuracy and racial
inequity.


New Standards is based on tossing comparisons anyways, this shows
that they're already giving in to sanity.


I doubt anything NCEE does would qualify as psychometrics, it's all
based on Tucker's gobbledygook garbage view of ed reform.


From my evaluation standpoint, the examples given are all
ridiculously hard at every grade level and bear no resemblence to any
written standards, any claim to be based on "world class" standards
is simply a lie. It only means that Tucker took a plane trip
overseas, designed a badly mangled version of the German system, and
came up with a test to go with it. German standards aren't this high
either.




Date sent:      	16 Sep 98 10:09:23 EDT
From:           	Gary.M.Degasta@Dartmouth.EDU (Gary M. Degasta)
Subject:        	Re: Barcoding
To:             	arthurhu@halcyon.com


> Arthur, I was referring to the Harcourt-Brace efforts to link the New Standards
> Reference Exams to the Stanford 9 Achievement tests. They used a control group
> of 3500 students, and gave them both exams. They are trying to validate and
> compare the two so that our Governor can say that this is how we compare on a
> national level. They did an experiment, basically, taking 1200 4th graders,
> 1800 8th graders and 500 or 600 10th graders.... somewhere in the nation. Our
> DOE contact could not tell us how many states participated in this national
> sampling. Then they took Vermont's scores on the New refernece Exams and linked
> them to this study. This gave our Governor the ammunition to say that on the
> New Standards Refererence Exams, our state did performed better in a national
> comparison. What we are attemtping to investigae is how realiable this sampling
> is and its validity. When I told Dr. Rosenkrantz this, he called this attempt
> by the NCEE voodoo psychometrics. I was,'t referring to the SAT's or anything
> else. I thought Fran Rice had outlined this for you.
> 
> Anyway, we are still researching this and hope to expose it if it is bogus.
> 
> Fran