Date sent:        Tue, 27 Jan 1998 12:26:22 -0500
From:             "Richard G. Innes" <70224.434@compuserve.com>
Subject:          RE: Florida repeats WA's test mistakes
To:               ed-consumers main listing <education-consumers@tricon.net>


To: Ed. Consumers,


<<Florida leaders say they are bracing themselves against any public outcry
that develops after students take a tough new assessment test next week.>>


How does the line go about those who don't know history are doomed to
repeat it?


<<The FCAT, which will measure reading and writing, was developed
specifically for Florida and approved by the governor and Cabinet almost
two years ago.>>


Kentucky's KIRIS was approved in 1990 and first administered in early 1992.
 Three years later it was well-known that the test had problems, and two
major reports said so in 1995.  Why did Florida ignore this and rush into
the same mistake a year later?


<<The test...includes multiple-choice questions and questions that require
students to write explanations or show how a math problem has been
solved.>>


This is the same promise made in Kentucky in the early 1990's.  In the next
day or so our legislature will start taking votes on repairing the mess
that has ensued.  It's nice to know Kentucky is leading the country in this
area.  Other states are just starting to make the mistakes made here nearly
a decade earlier!


<<Results, which will be available in May, will provide a basis from
which officials this summer will develop standards for scores.>>


Using just one year for a baseline was the start of many of our problems. 
We didn't have trend data to examine before the test actually started
counting.  That was dumb mistake number one, in my opinion -- not looking
at trends.  Florida didn't learn a thing from us!


<<The public may be dismayed at the first results....>>


May?  You can count on these results being dismal.  First of all, first
time testing always results in low scores because teachers and students are
not familiar with the formats.  


In Kentucky, we had an added problem that scorers apparently were directed
to score low.  I talked to one of them, and that was the briefing -- score
low so we can show rapid improvement later.  


Perhaps Florida's testing will be more above board.  Of course, Florida
didn't learn much from Kentucky so far, so maybe there isn't much hope this
obvious problem will not appear in the Sunshine State.  It will be
interesting to see if cheating and all the other Kentucky problems show up,
too


By the way, does anyone know how the Florida tests were created?  If they
used testing contractors, who were they?  How will the open response
questions be graded?  What scale will be used?


Richard Innes, E-Mail: 70224.434@compuserve.com 






  
EDUCATION CONSUMERS CLEARINGHOUSE




Richard, I got your packet of information and zip disk, and it's 
dynamite for anyone else that wants an excellent newsletter that 
shows that KIRIS is a massive bust, and while they brag about KIRIS 
scores going up, every other indicator of progress is going DOWN. 
I'll be posting as much as I can off of my education website shortly.


I'll also forward this to WA press and educators so they a preview
of the hell that Terry Bergeson will soon be getting once other
people get wind of the 7th grade tests they've been peddling
as a 4th grade test. Heck, even the first page of the college level
SAT had more 4th grade level problems than the WA test.


All we need is loop people from all of these states to raise 
hell about this performance based testing garbage. I'm going
to testify in favor a bill which will attempt (yeah, fat chance, 
even our republican legislature is sold on Tucker Based
Edcuation) to repeal the performance based reforms, 
text is at 


http://leginfo.leg.wa.gov/pub/billinfo/senate/5875-5899/5890_021897


For all this talk about kids who are good at "guessing", can
anyone compute the probability of getting 40 out of 40
mulitple 4-choice questions by guessing? Try 2^80. Can 
you count that high??? A 64 bit computer can't even count
that high!




> 
Thanks, you're data will be good for waving at WA
legislators. How come Business Week and Ed Week
don't write about you, Cunningham, and your friends? 


And what the heck is the deal behind this Pritchard
Committee? Seems you've got your own home grown
NCEE taking gobs of corporate money to spread Tucker's
poision.


Do you have data on how races fare? If it's like CLAS or
WA's test, minorities should be way the heck down in the
basement on these tests. If it's not broken out, has anyone
tried to get them to do a breakout? Do you have 
minorities complaining yet? At least the article on the
national test indicates at least one other person knows
about minority impact besides Cunningham and me.


> Date:          Tue, 27 Jan 1998 17:21:46 -0500
> From:          "Richard G. Innes" <70224.434@compuserve.com>
> Subject:       RE: Florida repeats WA's test mistakes
> To:            "INTERNET:arthurhu@halcyon.com" <arthurhu@halcyon.com>


> Arthur,
> 
> <<I got your packet of information and zip disk, and it's 
> dynamite for anyone else that wants an excellent newsletter that 
> shows that KIRIS is a massive bust, and while they brag about KIRIS 
> scores going up, every other indicator of progress is going DOWN.>>
> 
> I'm glad you found this helpful and interesting.  Did you have any problem
> reading the Zip disk?   
> 
> If you need more ammunition for your legislature, let me know.  I am pretty
> well connected to all the Northern Kentucky members of our state
> legislature's education committee.  If words from them would help, let me
> know that, too.  We all have major problems with this program, as the data
> I sent obviously made clear.

> 
> <<For all this talk about kids who are good at "guessing", can
> anyone compute the probability of getting 40 out of 40
> mulitple 4-choice questions by guessing? Try 2^80.>>
> 
> Let me add to that.  An 8th Grade teacher I talk to frequently says that on
> Kentucky's KIRIS assessment our kids have to score above 85 on a 0 to 140
> scale to get a grade above "Apprentice."  She is furious.  A student's
> score can rise from 15 to 85 and there is no change in the grade!  Add your
> probability data above, and there is virtually no "random" chance a kid
> will score above "Apprentice" just by guessing.  
> 
> However, if the kid is a special education student, he gets "accomodations"
> on KIRIS, which means special teacher help.  Our Special Ed kids OUTSCORED
> regular students on KIRIS science a year or two ago!  That, however, WAS
> NOT random.
> 
> By the way, the KIRIS grades are "Novice," "Apprentice," "Proficient," and
> "Distinguished."
> 
> Dick  
> 


It's also funny that while it is impossible to get all right
answers simply by guessing on a multiple choice test,
on many of the WA questions, you can only get the
"right" answer by making a guess - in some instances,
you must chose between equally valid or nearly equally
valid answers. There must be cases like this on the
KIRIS test. 


Have you had a chance to compare an actual 4th
grade test against the 4th grade benchmarks? WA
gets a high failure rate by making 35% of the
questions material from the 7th grade benchmarks,
and knowing that no one will know the difference.