Re: Push for 'value added' - or valueless addition?

["Horn, James" <jhorn@monmouth.edu>]

Bill Sanders, the silver-back of of this group of growth-modelers, has
always said that it is not his business to say what should be tested or
even to say if it is a good test or a bad one.  He is concerned only
with applying his proprietary algorithm to any quantitative set of
longitudinal data.  His being the top shelf approach, one can expect the
zealots of efficiency to choose something much cheaper. Besides the
concerns Monty expresses, I have always been concerned that "growth
modeling" and looking at gain scores serve to make the effects of
poverty even more invisible to this color-blind crowd than it is now--if
that is possible.

________________________________

From: arn-l-owner@interversity.org [mailto:arn-l-owner@interversity.org]
On Behalf Of Monty Neill
Sent: Wednesday, May 03, 2006 10:25 AM
To: care@yahoogroups.com; RScriticalteach; ARN-state@yahoogroups.com;
ARN-L; arn2-strategy
Subject: [arn-l] Push for 'value added' - or valueless addition?


The push for so-called 'value added' testing is, I think, growing. See
this piece by Tom Toch of Education Sector at
http://www.educationsector.org/analysis/analysis_show.htm?doc_id=336629
 
He notes that Sandy Kress who helped design NCLB is for it. 
 
Indeed, there is research suggesting that value added is likely to be
more fair.
 
Toch is also correct that a major obstacle is the shared political
illusion that all children will score proficient by 2014. He notes many
of the harmful side effects of this approach.
 
Ironically, Toch also authored for Ed Sector a report showing how weak
most state tests are, in substantial part because states won't pay for
better tests. That argument is not in this piece. 
 
Aside from numerous technical problems and that it is nearly impossible
to understand the technology without a substantial statistical
background, the deep flaw in "value added" is the lack of value being
added: the extant designs rest on current standardized, mostly
multiple-choice tests. While some of value is measured, much of what is
measured is included because it is easy to measure even if not very
important; and great amounts of what is important is not measured.
Worse, with such tests are used for high-stakes purposes, whether used
as NCLB does or in a value-added scheme, the result is narrowing
instruction to teach to the test. This phenomenon has been reported in
numerous studies - the only questions are how severe it is and how
intense its effects are by different sorts of schools - the evidence is
that it is more common and more intense in schools that have lower
scores (e.g., mostly serving low-income kids). For those willing to
accept a dumbed-down education for poor kids, this is fine. For the rest
of us, beware valueless addition. 
 
Monty 
 
Monty Neill, Ed.D.
Executive Director
FairTest
342 Broadway
Cambridge, MA 02139
617-864-4810    fax 617-497-2224
monty@fairtest.org
http://www.fairtest.org
Donate: https://secure.entango.com/servlet/donate/MnrXjT8MQqk