Re: CA State results shed new light on wealth vs. poverty deba...

[Peter Farruggio <pfarr@cal.berkeley.edu>]

BINGO!  You hit the nail on the head.  Why do they refuse to reveal 
raw scores?  And exactly HOW do they calculate their scales from 
these raw scores?  The mysterious scale numbers are used to decide 
where the lines are drawn for the levels (advanced, proficient, 
basic, etc).  The public sees an even line of numbers for students, 
like child #1 has a 258, child #2 has 315, etc, and assumes that 
these scaled numbers accurately represent correct answers.  But how 
do we know?  If I get 18 of 20 correct on a bunch of subtests 
(spelling, language, vocabulary, etc), and you get 16 of 20 correct, 
will you be that close to me on the scaled score for Language Arts, 
or will we be spread out more?  In the latter case, you will appear 
to know a lot less than I do, when in reality we are pretty close (at 
least in the logic of standardized tests as accurate measures of knowledge)


At 05:50 PM 8/18/2007, Le1212@aol.com wrote:
> The questions about the validity of poverty information is important.  I have
> a greater interest in how items are evaluated by CDE and result in a test
> score.
>
> I question the accuracy of this year's test scores.  Each year the state
> changes items and decides if the new items are appropriate to include in that
> year's score.  When I checked two years ago, the number, or 
> percentage, of correct
> answers needed for "proficient" performance varied from subtest to subtest
> and from year to year.
>
> This variability seems to be due to the ratings of items by mysterious forces
> within CDE.  They use various statistical measures to evaluate items.  It is
> quite likely that these statistical measures impact how much scores change
> from year to year.  Unless these mysterious forces are put into clear view,
> comparison of scores from year to year is more politic than 
> education.  If they
> truly want teachers to embrace this assessment model, teachers 
> should be allowed
> to see the test's item analysis.
>
> Marc Barlow
>
>
>
> ************************************** Get a sneak peek of the all-new AOL at
> http://discover.aol.com/memed/aolcom30tour