Re: In Defense of Testing

[Monty Neill <monty@FAIRTEST.ORG>]

Problem is that NRTs measure a limited slice of what is reasonable to expect
kids to learn, while testing much that is not so relevant, and doing so in a
way designed to sort kids on a rough bell curve in 99 percentile ranks. Bad
measures badly conceived usually put to ends more harmful than helpful
(e.g., tracking with many kids on dead end tracks). The info is usually too
little and too vague to be of use to teachers -- a good teacher might
occasionally find a small nugget of info, hardly worth imposing the test. As
to value added, see article by Jerry Bracey in last FairTest Examiner posted
to our website www.fairtest.org.

Monty Neill
-----Original Message-----
From: Juanita Doyon <Jedoyon@AOL.COM>
To: ARN-L@LISTS.CUA.EDU <ARN-L@LISTS.CUA.EDU>
Date: Wednesday, February 14, 2001 2:39 PM
Subject: In Defense of Testing


>Any opinions on this? (yeah, right)  It seems to me a pretty comprehensive
>and practical plan, that could be a big improvement over the current
testing
>mess.  If it replaced rather than added to the state mish mash of
assessments.
>
>http://www.educationnews.org/In%20Defense%20of%20Testing.htm
>
>In Defense of Testing
>
>I am a public school educator, and I believe that a valid and reliable
>method of testing students
>Marilyn Keller Rittmeyer, Northwestern University
>
>I am a public school educator, and I believe that a valid and reliable
>method of testing students is necessary to adequately inform parents and
>educators about the children's intellectual development. I favor
value-added
>assessment, but suggest a simpler approach than Tennessee Value-Added
>Assessment.
>
>Many nationally normed standardized testing instruments offer to provide
>either grade level equivalencies or percentile ranking for reports. Though
>grade level equivalencies can be misleading to parents, as well as to
>educators, and clarification would have to be provided, using grade level
>equivalencies allows for some interesting analysis.
>
>I recommend that all students in Grades K-10 take a survey battery within
>the first few weeks of the school year, so that beginning grade level
>equivalency can be established. The time required for such a test is
>approximately 1.5 hours. At the end of the school year, a parallel form of
>the survey battery or a comprehensive core subject battery (longer and more
>informative) should be administered. If the comprehensive test is
>administered, the time for testing approximately doubles. Gain in grade
>level equivalency can be determined for each student. Testing experts at
>Riverside Publishing Company have told me that a student who starts the
>school year below grade level equivalency can be realistically expected to
>gain 6-8 months after a normal 10-month school year. Students who start the
>school year at grade level are expected to gain 10 months grade equivalency
>after 10 months instruction. Students who start the school year above grade
>level are expected to gain 12-14 months grade level equivalency after 10
>months instruction.
>
>Students who have a handicapping condition or limited English proficiency
>can be tested using a lower level form of the nationally normed test than
>their same-age peers would be administered.
>
>Though nationally normed standardized tests are not tied to specific state
>standards, research has proven that students have greater achievement on
>nationally normed standardized tests if the school they attend is using
>rigorous standards (See John Hopkins Study on Core Knowledge Schools at the
>Core Knowledge Foundation website). Though criterion-referenced tests based
>on the state's standards are not essential to improving classroom
>instruction, standards can still guide the instructional program.
>
>Some people will claim that too much testing is being done. But my
>recommended testing program requires only 1.5 hours in the fall and a
>maximum of 3 hours in the late spring. It is important to administer the
>survey test in the fall so that summer achievement gain or loss can be
>monitored. Also, no one will accuse the teacher of "teaching to the test"
>for the fall administration; the fall test will be a true reflection of the
>students' long-term achievement. Parents and educators need objective test
>information on an annual basis so that trends in achievement can be
watched.
>Wise parents take their children to the pediatrician on an annual basis to
>have their physical development evaluated against norming charts, thus, any
>problems are identified and dealt with early on. Children's intellectual
>development deserves to be tracked annually throughout a child's early
years
>for similar reasons. If we only have objective test data every few years
>instead of annually, we may lose valuable time during the child's schooling
>when intervention was needed.
>
>Test data should not be used for individual teacher evaluation or merit
pay,
>because many factors inpact student achievement. The test results can be
>used, along with other criteria such as teacher input, to guide
>instructional decisions for individual children and the results can be used
>to develop a school improvement plan.
>
>As for reporting to the public, I suggest that each grade and each school
>and each district be evaluated by the average gain made by three categories
>of students: those who start the school year below grade level, those who
>start the school year at grade level, and those who start the school year
>above grade level. By focusing on the gain in achievement for three
>categories of students (below, at, above grade level), the staff of the
>school will focus on the instructional needs of individual students. We
will
>be able to compare schools more honestly, because we will be comparing
>similarly achieving students in each school. If the average gain is not
>equal or better than the gain recommended by the publisher of the test, the
>school has some explaining to do. On the contrary, if the gain is
>significantly higher than recommended, exemplary programs of instruction
>will be identified as models for other schools to use.
>
>I hope that decision-makers will give serious consideration to my
>recommended testing program. The current one-size-fits-all approach to
>testing either guarantees a test that is too easy for most students or is
so
>difficult that many students fail. Wise educators and wise parents know
that
>all students benefit from a rigorous instructional program, but that they
>don't all achieve the same verbal or quantititative literacy at the same
>pace or at the same time. Tracking students' annual gain in intellectual
>development on a nationally normed standardized test provides important
>information to parents, educators, and to students.
>
>Marilyn Keller Rittmeyer
>Northwestern University, Evanston, IL. USA mkr@nwu.edu
>
>--------------------------------------------------------------------------
>To unsubscribe from the ARN-L list, send command SIGNOFF ARN-L
>to LISTSERV@LISTS.CUA.EDU.
>

--------------------------------------------------------------------------
To unsubscribe from the ARN-L list, send command SIGNOFF ARN-L
to LISTSERV@LISTS.CUA.EDU.