Home | Top of Index | Contents | Education Reform Wars Main Page
url: http://www.arthurhu.com/index/washtest.htm Assessment Links

FIX OR KILL WASL!: Assessment for Education De/Reform

News!

WASL Assessment State Page http://assessment.ospi.wednet.edu/ASMNT99/WASL-789.html \clip\99\19\wasl99.htm 1999 test results

Partnership for Learning Dumps corporate money into slick publicity for a faulty test which harms our kids. Poll purports showing voters support "raising standards". AT&T - 1-800-550-kids

Headlines

10/10/99 Seattle Times reports 4th grade test is too hard 4/26/1999 4th grade test is easier than 10th grade test, sample questions used on the "secure" acual test!

Introduction

"The key is an assessment system, not a list of goals. The assumption: The high stakes system will cause the desired change in teaching and goals".

Marc Tucker, presentation to educator in Seattle 1991 leading to WASL

Washington State is following other states in implementing what is called "performance based testing". The purpose of these test is not to measure effectiveness of current curriculums, but deliberately fail current skills in order to enforce the drive towards a radical reform of education.

It is the cornerstone, and first step in reforms as envisioned by people such as Marc Tucker of the NCEE who first proposed such tests to Washington state in 1991. Although lip service is paid that the test only covers the Essential Acdemic Learning Requirements, in fact, most of these tests ignore any written standards documents. Nationwide, 22 states are tying graduation to these tests

Marc Tucker started this movement with his proposal to have a Certificate of Mastery backed by a new generation of tests. He has stated that we need to insure that all meet "world class" standards to be globally competitive. But in fact, nations like Germany and Britain do not require all students to meet the same performance standard. Students are ranked, and graduate with whatever grades that they get, which may be passing or not.

The trend is away from rote checking for memorized skills towards free response and partial answers for questions with no one correct answer. They are graded more on prose than actually getting a correct answer. They are sold as being carefully checked for cultural bias. Yet they are written to impossibly high cognitive standards that are more difficult than IQ tests in terms of impact on minorities and the poor, so that 4th grades get problems that look like middle or high school and high school expectations exceed college placement exams.

Conservatives charge that using the same passing standard for all will result in lower standards. But in actuality, standards are being placed so high that only the top students can actually demonstrate this performance, and in some cases, no schools have been able to meet state goals of 70% or 80% passing.

Washington's 1997 4th grade math test placed half of the population in the lowest "doesn't know anything" score category, failed 80% of the general population, and 97% of the minority population. By 1999, 80,000 4th, 80,000 7th and 80,000 10th graders were taking the test. [about $10M per year]

In comparison, traditional norm-referenced tests are designed to discriminate between the best and the worst between wide extremes. While norm-based tests were designed primarily with statistics and scientific research in mind, the new tests are based almost entirely on politics, with no research to back up the assertion that the new tests are improvements over the old tests in terms of cost, reliability, or fairness. Tests are deliberately piloted with known failure rates of 75% or higher, without any evidence that any diverse group of students can be expected to be taught to pass at higher rates.

These are proving more difficult than the old tests. In the fall of 1997, Washington was reviewing the results of a new test widely heralded as "raising standards" so that "all children will master essential skills". No one in the education business questioned the idea that a test that 80% of students "flunked" wasn't too hard. Nearly half of problems required knowledge beyond 4th grade curriculum in any current textbooks, and up to 10th grade level measured by the state benchmarks the test was allegedly designed to assess. In every other state with similar tests, similarly low pass rates are celebrated as proof of the need for reforms, and examples show similar usage of over-grade level problems.

Other Experts on Testing and Assessment Problems

What's wrong with the new "performance based" tests?

Read Sander's paper for the the best case against the new tests: (4) "Educational Assessment Reassessed: The Usefulness of Standardized and Alternative Measures of Student Achievement as Indicators for the Assessment of Educational Outcomes," Educational Policy Analysis Archives, Vol 3 No 6, March 3, 1995

It's the best summary of what's wrong with these tests, it amounts to what I was able to figure out on my own with the CLAS test.

A Point By Point Critique of "Best Features"

Feature: They don't rely on multiple choice.

Problem: Essay problems are much more difficult for poor and minority students. Most students aren't taught to answer math problems by explaining in terms of words and pictures instead of simply computing a corrrect answer. Essay questions are much more difficult to score and grade because they require subjective judgements. Machine scoring is much less error prone, and much less expensive. Either they're right or wrong, and no bone-headed idea that you should get full credit even if the result is wrong.

Feature: The new tests are tough

Problem: Tough? They don't test what the students are taught, and in the WA assessment, they cover topics that according to the national NAEP frameworks shouldn't even be assessed at the 4th grade. For example using proportional logic to measure things like flagpoles based on their shadow isn't until the 8th grade. Reading scale drawings is an 8th grade skill. NAEP guidelines say to not do conversions, even within the same system, like converting 60 inches to 5 feet. Yes, they're supposed to be able to read simple bar charts, but not frequency histograms with unlabeled minor tics. They're supposed to know about central tendencies, but not formal mean or median, and not pick out the one with the most items UNDER a certain value like the battery life question. And they're not supposed to be exposed to experimental statistics. They're not supposed to be tested on trick questions - the battery life question was posed as which one should you pick which lasts at least 9 hours, but this answer was different from the one with the longest average life, which is what most students are taught to look for in battery ads.

See these sample problems: Problems with comments | Official samples | impressions of 1999 test

WASL 4TH GRADE TEST IS NOT 4TH GRADE BUT 7TH AND 10TH GRADE OR COLLEGE LEVEL

The following are NOT essential 4th grade skills. None of these is explicitly mentioned in the 4th grade Washington state essential skills, only some are in higher grades. They are not mentioned in the answer key, which only gives the relevant strands, despite statements by officials by the Commision on Student Learning, or the Superintendent of Instruction. The only 2 answer rubrics released were both wrong (measuring height by shadow, bird feeder cost list) In fact, many of these tasks ARE in the benchmarks, but at a much higher grade level. Similar skills can be found in other state tests like Kentucky, Exemplars, New Standards and other Tucker-inspired tests. George Cunningham and others have noted that these tests are way beyond any current standard expected of students, and often pay no attention to the standards (usually similar to Washington's). Similarly, the claim of "world class" benchmarks is fictitious since Japanese and German standards don't put these 7th and 10th grade skills into G4 either.

Grid showing grade level of sample problems Only 40% of problems are clearly within 4th grade 31% are clearly at 7th grade or higher level by benchmarks

Two education experts, Gregory Cizek of the University of Toledo, and home schooling advocate Richard Bray agreed that the WA 4th grade test would be difficult for a 6th grader if not higher, and some skills are on the 8th grade NAEP agenda. It is claimed that these skills are in the new Essential Learning Skills 4th grade guide, and that the best schools teach these skills, but they are NOT on anybody's curriculum or text book, they're not in the Essential Academic Learning Requirements/Benchmarks (EALR), and the only places that scored relatively well (70% passing, still 30% failed) was in the best and wealthiest schools under the old tests.

Feature: They're not racially biased

The WA test had blacks score 0.8 Standard Deviations lower than whites on the math test, virtually the same gap as IQ tests where blacks outside the south generally score about 90 IQ, or 0.8 SD less than whites. California's CLAS, and every other state test show minorities doing just as bad, and usually worse on the new tests than thge old fashioned ones

Therefore, if inequality is bias, the WA test is just as bad if not worse than IQ tests. Written response tests are even more difficult for minorities than multiple choice tests. Multicultural questions are not easier for minorities, passages by authors like Alice Walker are unintelligible even by college level readers. Questions with partial credit for process over answers are not easier. It is all an illusion that getting rid of these features will make them easier for minorities.

Feature: They don't grade students on a curve.

Problem: Criterion based tests base based on largely arbitrary choice of a standard with no apparent relationship to current curriculum. For example the 4th grade test assumes knowledge of how to build a bird feeder out of wood set into the ground in cement, or how to interpet frequency histograms. Such tests CAN be converted to a curve, but even so, is not standardized to allow comparison with other states. In fact, it is highly deceptive when the press in WA, RI and Colorado all have headlines about how the new scores are much lower than old norm-referenced tests when in fact, the level of performance is the same. All they did is change the level of expectation from 50th percentile to meeting some new arbitrary state standard which falls well beyond national average.

Feature: Tests are graded on a 5 or 6 point scale "mastery" based or criterion-referenced instead of norm-referenced scale.

Problem: Mastery _is_ relative. Some kids can solve a much higher percentage of advanced problems than others. There is no absolute measure of what a student "should" know without a measure of what actual kids can do. Distributions of such tests typically bear no resemblence of the percentages of students who get A or B letter grade, which normally correspond to equal to or better than expected. In the 1994 CLAS, 0% got a "6" which corresponds to an "A", and only 5% got a "B", which was like a "B" grade. Where curve-based tests are criticized because it shows some grades scoring above average, the new tests can and sometimes do "flunk" almost everybody. Such a coarse scale is almost useless for ranking students compared to a 1-100 percentile scale or the typical 200-800 SAT style scale which can discriminate between students even in the top 0.1% of the curve.

Feature: Problems are graded such that good explanations are more important getting a correct answer.

Problem: When questions don't have just one correct answer, students can be confused as to what the correct answer is, and grading is very subjective.

Feature: Problems are scored by a human expert, not a computer

Problem: Grading is subjective, with no guarantee that the same result will be scored the same by different scorers. Computers can automatically score multiple choice questions with no problems with repeatability, and provide very fine resolution between ability with many problems.

Feature: New tests look for "thinking skills" instead of rote memory

Problem: Critical thinking was never the problem with students who did well or poorly with the old tests. Questions which resemble an intellectual obstacle course are only even more difficult for the poorly prepared, as the California's CLAS showed the poor and minorities did even worse. New tests often trade demonstration of hard computational or verbal skills for this "thinking skills" instead. If you want to add thinking problems, add them, don't replace facts and skills based problems, which can result in lower, not higher math performance.

Features: All students should be able to learn

Problem: You can't simply state that we're going to set a standard that only the top 85th percentile can hack right now, and say that everybody is going to pass this, and then require 85th percentile performance for a "certificate of mastery" to proceeed on the next grade or get a high school diploma when currently performance above the 15-20th percentile is good enough. These tests are really no different than IQ tests, and nobody's even been able to demonstrate an educational program that takes a general population and makes them pass any test at a 85th percentile / 115 IQ level. It just can't be done. If you can show me a research project that shows it's possible that's one thing, but it's nuts to run a real school system that way with no evidence that it's possible. @@Contents

Summary of 1997 WASL Data

District Rankings on "new performance based 4th grade test"


washasrk.xls
Scatter Chart showing
correlation between new test and old CTBS
scatter chart

	WASL Spectrum, ranked by math score

	CTBS = math over 50 percentile 
    FL= Free/Reduced Lunch%
    CTBS= percent above 50th percentile on traditional test
    Math etc. = percent above "standard"

	Math  Read Listn Writ CTBS FL
	 92   100   96    92  NA    Challenge Edmonds
	 83   100  100   100        St. Thomas Bellevue Cath
	 80    88   92    80        Bertschi Seattle Priv
	 80   100   93    53        Brighton Edmonds Priv
----------------- best public school = 72 ------------------------
	 72    82   87    83  91  2 Somerset Bellevue
	 68    95   91    74        Assumption-St Bridget Seattle Priv
	 66    83   93    83  77 16 Wedgewood Seattle 
	 63    90   90    90        Evergreen Academy (Bothell Priv)
	 62    79   85    79  82  1 Cougar Ridge Issaquah 
	 61    78   85    73        Bush Seattle Priv
	 60    81   88    67  88  1 Lakeridge Mercer Island
------------ best district = 60 ---------------------------------
	 58    76   84    71        Mercer Island Dist
	 58    79   71    79  78  1 Sunrise Northshore 
	 57    68   85    77  72 26 Laurelhurst Seattle
     53    74   86           25 Kennewick Ridge View
	 51    75   86    83  84 21 Whittier Seattle
	 50    70   74    35  80  9 Shelton View Northshore
	 50    67   71    82  62  7 B.Franklin Lk Wash
					  68        Black Diamond
	 50                         Franklin Elem (LWSD)
	 44    71   76    52        Bainbridge Island
	 33               61        Lake Washington
	 36    64   74    65        Issaquah
	 34    62   71    57        Bellevue
	 34    57   73    59        Northshore
------------very good suburb 31-34 ------------------------
	 32    62   74    39        Riverview
	 24    47   51    50        Auburn
------------average 22-30 --------------------------------
	 22    48         42        Statewide
	 22    59   75    46        Snoqualmie
	 21    58   65    46        Enumclaw  
	 17    48   62    49        Renton               
	 14                         Bell Elem (LWSD)
----------- worst inner city schools = 3-7 ---------------------
	  7    28   40    31        South Central
	  6    47   61    57 51 41  Cascade Renton
	  6    31   52    40 40 56  Thorndyke South Central
	  3     3   28    28 40 73  Dunlap Seattle
	  3     39  77    36        Holy Family Auburn Priv
	  3    26   46    31 33 56  Campbell Hill Renton
	Math  Read Listn Writ

Poor urban schools do the worst, wealthy and private
schools do the best.


Statewide Performance by Race / Gender

\doc\web\97\08\waasrace.wk1

ASIANS AND BOYS MATH AVERAGE LOWER THAN WHITES, GIRLS
DIFFERENCE OF 0.67SD MEANS IT IS JUST AS DISCRIMINATORY
AGAINST BLACKS AS IQ TESTS.

Washington Assessment of Student Learning
Grade 4 Spring 1997
Analysis by Arthur Hu Oct 20, 1997
Data: Office Washington State Superintendent of Instruction

Mathematics                                                         Interpolate
        Not     Level                           State Std   SD      Est
Group   Tested         1       2       3       4Mean  Deviatdiff    Percentile
Total          3    46.3    28.9   14.95    6.85 374.5 33.95    0.00      50
Male         3.2    47.2    27.8    14.7     7.1 373.7  34.9   -0.02      49
Female       2.8    45.4    30.0    15.2     6.6 375.3  33.0    0.02      51
White        2.5    41.1    31.3    17.1     8.0 378.9  32.8    0.13      55
Black        4.0    73.4    17.1     4.0     1.2 351.6  30.8   -0.67      26
Hispanic     7.2    71.7    15.8     4.3     1.0 350.3  31.4   -0.71      25
Asian        2.5    45.8    27.4    15.6     8.7 376.8  34.9    0.07      52
Native A     3.4    70.1    19.7     5.3     1.6 355.0  31.5   -0.57      30

Note - percentile estimated assuming 1 SD = 35% difference in percentile
25th percentile = 90 IQ

Index White=1.00
        Not     Level                           State Std   State
Group   Tested         1       2       3       4Mean  DeviatMean
Total          3    1.13   -1.08   -1.14   -1.17 -1.01  1.04   374.5
Male         3.2    1.15   -1.13   -1.16   -1.13 -1.01  1.06   373.7
Female       2.8    1.10   -1.04   -1.13   -1.21 -1.01  1.01   375.3
White        2.5    1.00    1.00    1.00    1.00  1.00  1.00   378.9
Black        4.0    1.79   -1.83   -4.28   -6.67 -1.08 -1.06   351.6
Hispanic     7.2    1.74   -1.98   -3.98   -8.00 -1.08 -1.04   350.3
Asian        2.5    1.11   -1.14   -1.10    1.09 -1.01  1.06   376.8
Native A     3.4    1.71   -1.59   -3.23   -5.00 -1.07 -1.04   355.0

Chart


Ranked by % scoring 4 (highest range)
                                                % of
        Not     Level                           State
Group   Tested         1       2       3       4Mean
Asian        2.5    1.11   -1.14   -1.10    1.09 -1.01  1.00
White        2.5    1.00    1.00    1.00    1.00  1.00
Male         3.2    1.15   -1.13   -1.16   -1.13 -1.01
Total          3    1.13   -1.08   -1.14   -1.17 -1.01
Female       2.8    1.10   -1.04   -1.13   -1.21 -1.01
Native A     3.4    1.71   -1.59   -3.23   -5.00 -1.07
Black        4.0    1.79   -1.83   -4.28   -6.67 -1.08
Hispanic     7.2    1.74   -1.98   -3.98   -8.00 -1.08
Blacks and Hispanics only 1/7 or 1/8 as likely as whites to score
"4", nearly twice as likely to score "1"

Ranked by Mathematics Mean
        State   White
Group   Mean    Ratio   Rank
White      378.9    1.00       1
Asian      376.8   -1.01       2
Female     375.3   -1.01       3
Total      364.4   -1.04       4
Male       373.7   -1.01       5
Native A   355.0   -1.07       6
Black      351.6   -1.08       7
Hispanic   350.3   -1.08       8
Asian mean is lower than White, Male mean is lower than Female
Discussion:

State Superintendent Terry Bergeson insists that the tests are not too difficult. But since 46% of students scored only a "1", which corresponds to "knows or understands almost none of what was expected", and 50% is generally understood to be the definition of "grade level", endorsement of the new standards is equivalent to declaring that grade-level achievement corresponds to a zero-achievement level on the new scale. This was not released in most news reports which only reported that "only 20%" of students achieved the state standard, but not that that the halfway point was at the lowest score. Now the Bergeson is cooperating with McDonalds to publicize the new standards asking "are you as smart as a 4th grader" without mentioning that the 50th percentile student completely flunked.

Ranked by top scores, Asians are most over-represented, but by far less than on SAT test, Native Americans, Blacks and Hispanics are under-represented by a factor of 5 to 8 times at the level of a "4"

The Asian average, however is actually about 1% worse than whites, compared to minority scores, which are 7-8% worse. This is contrary to most tests where the Asian mean is also higher than whites, not just the percentage at the top.

Similarly, boys are over-represented at the top (5) and bottom (1), but their average is lower than for girls. Like the CLAS test, girls outscored boys in every category in terms of their median/mean score, which puts more emphasis on writing. It has been observed by many that boys are more likely found at either extreme

A difference of .67 Standard Deviation between white and black scores would make the black score about 90 on an IQ scale, or about the same as IQ tests. Since the most difficult tests have the biggest spreads between races, this puts the new WA assessment on the level of an IQ test in terms of difficulty and disparate racial performance. Blacks statewide generally peform at the 30th to 40th national percentile on tests like the CTBS instead of the 15th to 25th percentile equivalent 85 to 90 IQ on most IQ test results.

Oddly, though organizations like Fair Test claim IQ and SAT tests are biased because results are so disparate, they support new style assessments even though racial differences are even more prounounced than tests like the CTBS.

Tests which require written responses are MORE DIFFICULT and will therefore, MAGNIFY any differences between the races.


Consumer's Guide Ratings of Standardized Tests

Excellent Norm-Referenced Tests
-------------------------------
CTBS
SAT
ACT
TAAS (Texas)


So-So
-------
NAEP - Criterion referenced, higher order thinking
Voluntary National Test - NTCM fuzzy math standards,
  constructed response, criterion referenced
Terra Nova - also NTCM compliant

Awful Performance-Based Tests
-----------------------------
Colorado
Connecticut
CLAS (California) Tested for subjective values
WASL (Washington) Math tested 4th graders at 7th and 10th grade benchmarks
FCAT (Florida)
KIRIS (Kentucky) No multiple choice
New Standards (NCEE) NY City, Vermont
Exemplars (Vermont)
Massachusetts



Who is Against High-Stakes Testing




New Articles on WASL 

======================================================================================


Note how many articles are front page news.

zip#28\images\972\1209\p01.bmp Students Need To Show they can reason
on tests Seattle Times 12/7/97 Education Q&A Only 2 of 40 questions
are straight math, says you won't get full credit without correct
answer (but that's not what the sample bird feeder rubric says!)
they cut out my source of test being too hard. Asks what happens if
you don't pass CIM test.

\clip\98\09\prep.txt Seattle Times
http://www.seattletimes.com/news/editorial/html98/tested_051998.html
Tuesday, May 19, 1998 Don't let preparation blur state test results
TERRY Bergeson, superintendent of public instruction, is worried that
problems with the new state tests could seriously erode public
confidence in schools. She should be. 

"State tells schools: Don't Cheat. Warning issued after reports of
test rules being broken" Jolayne Houtz, Dick Lilly Seattle Times May
16, 1998. Front Page. Terry Bergeson and Chuck Colson, chairman of
the state Commision on Student Learning issued a warning memo after
getting reports of testing rules being broken, and improper coaching
at Seattle's T.T. Minor Elementary school.

 for all practical purposes the EALR
movement will destroy the junior and senior years of our high schools.
Rod Roberts


From: "Roxanne Sitler"  Date sent: Fri, 24 Apr 1998
07:19:00 -0700 This year HB2849 moved the CTBS from 4th grade to 3rd.
We tried to get an amendment to that bill to leave it in the 4th but
were unsuccessful.  Now they have no comparison for the
performance-based and CTBS on same grade level.  Also, CTBS was
normed for 5 years and that has expired which means they are
searching for a new norm-referenced test.  The RFP's will go out this
year.



April 23 Eastside Journal Claims no
other state is doing the CIM or tests like this, CIM is back to year
2000-01

CTBS 4th grade tests to be
cancelled in exchange for reading bill. That's just great. Now we
won't be able to tell that our kids can't read.

Spokesman review article - testing educational reforms. Says that 

"This is just a test: 4th graders go for heavy mental mettle." Debera
Carlton Harrell Seattle Post Intelligencer April 14, 1998 p. 1 (front
page) Achievementof minority students a major concern. [Solution to
height of flagpole from shadow is incorrect - it says to count the
!@#$% bricks!] 71% of 10th graders would flunk if they were given the
test today.
files \clip\98\08\test\p01-p04.tif

http://www.seattletimes.com/news/local/html98/test_032698.html
\clip\98\07\cramtest.txt Seattle Times March 26, 1998 Schools cram
for state test by Jolayne Houtz Seattle Times staff reporter More
students are writing in math journals these days, reading nonfiction
materials and being asked not only what the right answer is, but also
why it's right.  "Parents at Cougar Ridge Elementary in Issaquah, one
of the region's highest-scoring schools last year, are being asked to
help children learn multiplication tables and other basics at home,
so teachers can focus on math concepts and more sophisticated
computation"


Seattle Times March 24, 1998 Time to put your commitment to the
test by Gary Locke and Terry Bergeson Special to The Times [Blantant
sales pitch for "higher standards" and new test]
\clip\98\07\commtest.txt
\clip\98\07\timetest.tif

\doc\web\clip\98\01\testtest.txt
\doc\web\clip\98\01\testtest.htm
[Tacoma WA] News Tribune Putting
the tests to the test Washington students about to take a
different type of exam that already is raising controversy March
24,1998 Debbie Cafazzo; The News Tribune ; Education writers Mary
Schneiter, Skip Card, Susan Gordon, Kris Sherman and Debby Abe
contributed to this report 

" Parents like Terry Olive of Federal Way worry that the grading
system may be too subjective. But others support the new tests and
what they represent.  "

" Arthur Hu is a Kirkland parent of preschoolers who has made an
avocation of critiquing the new state tests and education reform. He
maintains an Internet mailing list about reform here and in other
states.  And he thinks educators are headed in the wrong direction.
Hu believes the Washington tests, particularly the fourth-grade
tests, are too hard. He points to dismal results from last spring's
testing of fourth-graders - the first of the new-style test scores to
be reported.  "

Letter to Gov Locke by Arthur Hu

http://www.seattletimes.com/extra/browse/html97/lati_121297.html
\clip\97\29\latinsc.txt Seattle Times Company Friday, Dec. 12, 1997
Groups seek better Latino test scores by Tamra Fitzpatrick Seattle
Times staff reporter [only 5% passed new math assessment, emphasizes
multiculturalism, latino history, not academics]

11/14/97 Representative from Commision on Student Learning says that
since 1994, they have been independent of the New Standards
organization.  They contract with Riverside to take questions
developed by committees. Items and scoring are kept confidential, and
"you just have to trust us", they have been field tested and are
based on the essential learnings requirements. She notes that the
example test did not have any examples of item-specific rubrics, but
says that they do exist on a real test, didn't have any to show. Not
familiar with national testing standards. Says the bird feeder
question was re-designed, "you know you always have a problem in the
developmental stage". "we tested them with real kids" Yeah, and 80%
flunked. Then she said bye.

\clip\97\25\edclip7.txt 11/2/97 Seattle Times Commentary If we set
the bar high, our children will compete by Mindy Cameron Seattle
Times editorial page editor ARE national tests necessary to improve
public schools and student learning across America?


Seattle Times Wednesday, Sept. 17, 1997 Test scores reinforce link to
family income

"Most Eastside scores higher than state average" Seattle Times Sept
13, 1997 p. A10 Katherine Long eastside bureau "Bainbridge district
sees room for improvement" (same date and page) Bell Elem was the
lowest scoring school in LWSD (principal told me it was one of the
most "diverse" schools in the district) image:
\images\97\76irs\washtest.tif

Arthur Hu Letter To Seattle Times Editor
on what's wrong with new WA assessment

ritest.txt 44% OF RI 8TH
GRADERS FAIL STATE MATH STANDARD, BUT EQUAL TO NATIONAL AVERAGE
\clip\97\25\ritest.txt New Test in R.I. Sends Shock Waves Through
State By Jeff Archer 10/29/97 Denver Post Scores on new test will be
low By Janet Bingham Denver Post Education Writer Brace for grim and
sobering student test results, the governor and the state education
commissioner are warning.

\doc\web\97\08\busibeat.txt "Businessman beats drum for education
reforms" Eastside Journal (Bellevue Wa) September 24, 1997 p. A2
Molly O'Connor. F101697

http://search.tribnet.com/archive/0997/0915a111.htm
\clip\97\23\misspent.txt VIEWPOINT: Misspent money, untested methods
produce poorly prepared students (Tacoma)News Tribune September 15,
1997 Clyde Ballard and Peggy Johnson

\clip\97\23\muchwork.txt
http://search.tribnet.com/archive/0997/0911b61.htm 'Much work to be
done' in schools Partnership for Learning assesses education reform
and identifies needs Tacoma News Tribune September 11, 1997 Kris
Sherman; The News Tribune

\clip\97\23\werewell.txt
http://search.tribnet.com/archive/0997/0919a143.htm Tacoma News
Tribune We're well ahead of other Washington in testing September 19,
1997 ROBERT MARSHALL WELLS; News Tribune Editorial writer " Despite
such a lackluster performance, complaints about low scores and
unreasonably tough statewide standards have been fairly muted. "

\clip\97\23\mintest.txt
http://www.seattletimes.com/extra/browse/html97/scor_101097.html
Seattle Times Company Friday, Oct. 10, 1997 Minority test scores
`devastating' by Jolayne Houtz Seattle Times staff reporter

Percent of students passing standard
              Math  Reading
Black         5
Hispanic      5     20
NativeAm      7     20

 Seattle Times Reformers could put
teachers to the test by Jolayne Houtz Seattle Times staff reporter
[Assess penalties and rewards for school that don't improve by x
percent on the new test, but no evidence that it has worked in other
states]

to Opinion at Seattle Times
re:\clip\97\23\mintest.txt
http://www.seattletimes.com/extra/browse/html97/scor_101097.html
Seattle Times Company Friday, Oct. 10, 1997 Minority test scores
`devastating' by Jolayne Houtz Seattle Times staff reporter [Of
course test scores are low for minorities]

Loop newsPeter Schrag: The new school
standards: A cheer and a caveat. Setting standards too high, and
rushing into innovations with no track record of success, CA goes
back to old fashioned machine scored state-wide tests.


Julanne Burts is bothered too


Anti: \CLIP\97\97\23\canttest.txt 
Seattle Times Friday, Oct. 3, 1997 You can't legislate higher
test scores by Ralph DeBruler Special to The Seattle Times

Pro: \clip\97\23\testok.txt 
Seattle Times Friday, Oct. 3, 1997 We're on the way to better
test results by Clyde Ballard and Peggy Johnson Special to The
Seattle Times

BRUCE ALBERT WANTS TO TEST FOR THINKING, NOT FACTS Business
Week \clip\97\21\badtest.txt A SCIENTIST ASSAULTS THE SCIENCE OF
TESTING Bruce Alberts wants students to interpret data, not parrot it
Business Week Oct 6, 1997


Telling Bill Pfeifer what's wrong with
the new test


http://ourworld.compuserve.com/homepages/mathman/lesson.htm The problem isn't norm referenced tests,
the new tests are based on a philoshipy that has destroyed math in
some places "The real problem in the continued use of
norm-referenced tests is the perception on the part of many that they
are useful and meaningful."... "The problem with the new assessment
is that it is not assessment at all. It is refusal to assess when
refusing to assess forthrightly would draw well deserved public
outrage. California's CLAS, Kentucky's KIRIS, etc., all share the
same problem in mathematics. They are so unreliable that they can
have no consequences or people will (and rightly will) sue if they
are used in making individual student decisions.  The ITBS, CTBS,
SAT-7, ERB, etc. do not have that problem. " Iowa students math goes
to hell with the new program.

Bill Pfeifer asks Arthur about the WA assessment test \doc\web\97\07\family.txt

EDUCATION WEEK REVIEW \clip\97\21\assess.htm
http://www.edweek.org/context/topics/assess.htm Assessment: What You
Test Is What You Get Liberals say we have to test for skills, not
just rote facts, conservatives say it's just an excuse to toss out
facts, some say these won't discriminate against minorities (but they
do just as poorly in actual results, or worse) 

"Because it's difficult to test a lot of these thinking skills with
the multiple-choice bubble-in test sheet that was such an ideal tool
for testing rote learning, many school systems are moving toward what
they call "alternative assessments"--typically "performance-based
tests." These are more open-ended tests, where teachers judge
students on written essays, on the process they use to solve a math
problem, or even on portfolios of their work over the school year."

SEATTLE TIMES ENDORSES NEW TEST
http://www.seattletimes.com/extra/browse/html97/tested_090797.html
\clip\97\20\edref.txt Seattle Times Sunday, Sept. 7, 1997 Education
reform gets real

\clip\97\20\sking.txt
http://www.seattletimes.com/extra/browse/html97/scor_091297.html
Seattle Times Friday, Sept. 12, 1997 South King County school' test
results on par with state by Dionne Searcey Seattle Times South bureau

Sample Questions for Washington State:
http://csl.wednet.edu/Web%20page/Assessment/Pilot4/Pilot_menu.html

"LOW" SCORES CITED FOR NEW WASHINGTON PERFORMANCE BASED TEST, BUT 
IT'S JUST A BAD IDEA

Annotated copy with samples
http://www.seattletimes.com/extra/browse/html97/test_090497.html The
Seattle Times Company Thursday, Sept. 4, 1997 Education reform's big
test: Held to higher standards, fourth-graders score badly, showing
the challenge ahead

http://csl.wednet.edu/Web%20page/Assessment/Pilot4/Pilot_menu.html
Test samples

The new test costs $29 per student compared to only $2.40 for the
CTBS. Only about 20% met or exceeded "standard" in math, about half
met or exceeded reading though. Not quite as hard as CLAS where not
even 1% met the highest grade, and less than 5% got the 2nd highest.

----------------------------------------------------------------

BRACE FOR ROTTEN SCORES ON REFORM TEST \CLIP\97\20\JUSTTEST.TXT Mindy
Cameron Seattle Times editorial page editor August 31, 1997 p. B6
"Just a test? No, a giant step in education reform" [Comment - new
tests just flunk everybody on standards real kids can't meet, while
dropping basics that the old test were just fine at measuring much
easier, the poor and minority will still do the worse, while dropping
sight of mastering the basics] "In
Maryland, Bergeson said, only 10 percent met the new standards on the
first test. In Kentucky, same thing."

NEW MATH STANDARDS TEST ACTUALLY FLAWED \doc\web\97\07\mathtest.txt

HUNDREDS PROTEST DESIGN OF NATIONAL MATH TEST
Hundreds of mathematicians, academics, teachers, school board
members, parents and students have signed a letter to President
Clinton protesting the design of his proposed National Voluntary
Mathematics Test for eighth graders.

" * Competence in arithmetic and other basic skills is intentionally
not directly assessed.  

* Instead of computational competence, students will be able to have
calculators available at all times.  This is in direct contrast to
the desire of most parents that their children be able to calculate
for themselves.


* A number of the example problems are ambiguous, have potentially
significant cultural biases and are designed to be compatible with
mathematically weak strategies such as guessing.


* A major fraction of the score will come from subjectively graded
questions involving judgments about each student's individual
strategy and description of his work rather than on the correctness
of the actual answer


"
>For more information contact
>
>Michael McKeown of Mathematically Correct
>http://ourworld.compuserve.com/homepages/mathman/
>PO Box 22083
>San Diego, CA 92192-2083


http://207.1.177.17/mdb.html Kansas sample 4th grade test - geometry,
division, multiplication, perimeter, making change

posted for education use only
\clip\97\18\refmtest.txt F081897 (contains bird feeder example)

SEATTLE: NEW REFORM TEST DIFFICULT, GRADES ON EXPLANATION, NOT
CORRECT RESULT http://www.seattletimes.com/extra/
browse/html97/test_072197.html Copyright © 1997 The Seattle Times
Company Monday, July 21, 1997 Education reform's first test faces
examination by the public 
																					                      																					
"Students are measured not against other students but on how each one
measures up to a set of written standards." Uh oh. This is another
silly test like the CLAS that is so hard not even 99th percentile
students will meet the arbitrarily high set standard. It's just
stupid to give kids higher grades for their a good explanation than a
correct numerical result. All this effort to measure "higher order
thinking skills" only makes it more expensive to grade, and will make
the gap between the white/asian/rich and the black/hispanic/poor even
greater, and give the impression that everyone failed. Schools don't
teach how to make essays about solving math problems, they teach
solving math problems, and the old tests measured that just fine. The
problem isn't that the old tests weren't hard enough, the problem is
that students weren't doing that great with the old tests either.
The example test really requires taking shop in order to understand
how such an object is even built or how to read the drawings.

Gregory Cizek University of Toledo
responds that this would be difficult even for a 6th grader, if this
is an average question, it's a good thing they're giving it in
washington, it will make Ohio kids look better by comparison. Richard
Bray, home schooling advocate PHD also says the question is too
difficult.

\priv\96\08\newtest.txt Seattle Times April 21, 1996 " New test will
debut in schools this week It's linked to higher standards

www.sepschool.org separation of school and state

knows about 4th grade test expectations.  dr. gregory cizek prof educ
u toledo ohio 419-530-2611 gcizek@utnet.utoledo.edu


Issue Contents





 Performance Testing Issues

----------------------------------------------------------------------------

@@Ability

HIGH STAKES TESTING BUMPS AGAINST REALITY: KIDS DIFFER IN ABILITY
z60\clip\2002\10\phelp.txt
Mon, 28 Oct 2002
The Case Of The Truth-Telling (But Racially-Incorrect) Teacher
By Steve Sailer
http://www.vdare.com/sailer/truth_telling.htm
This brouhaha is an outcome of two contradictory trends in American society: 
1) The backlash against realistic thinking about intelligence that resulted 
from the The Bell Curve wars of 1994; 
2) The enormous expansion of "high stakes" testing in schools over the last 
decade.

@@ACT Work Keys

\clip\97\28\assess.txt http://www.edweek.org/ew/vol-11/08assess.h11
Education Week October 23, 1991 National Standards/Assessments:
Students [Comprehensive listing of major high school to work
assesmment efforts in 1991, including Bush tests, NAEP, Tucker, ACT
Work Keys, SCANS]

American College Testing Program: Developing Work Keys, a national
system for teaching and assessing employability skills that could be
used by employers and by students who plan to enter the workforce
directly after high school.


@@AERA

AERA BELIEVE NOT ONE STATE IS FAIR ABOUT HIGH STAKES TESTING
z45\clip\2000\09\aera.txt From Harvard Education Letter - Research
Online, September/October, 2000. See
http://www.edletter.org/current/aera.shtml
The full text of the AERA position statement is available
online at  http://www.aera.net/about/policy/stakes.htm . The AERA
website is at http://www.aera.net/ .
"These are all criteria that we believe have to be put in place in
order for testing to be fair," says AERA president Catherine Snow.
"Right now, there's not a state in the union that's abiding by all of
them." 


@@Against

People / Organizations against high stakes testing



LOS ANGELELES MINORITY GROUP CALLS FOR END TO EXIT TESTS
http://www.dailynews.com/
z63\clip\2003\02\panelex.txt Monday, February 10, 2003 - 9:09:20 PM
MST Panel urges suspension of exit exam Critics call failing students
victims of failing system
By Helen Gao Staff Writer

z47\clip\2000\12\edmis.txt
http://www.nea.org/issues/ataglance/high-stakes/popham.html National
Education Association NEA Teaching and Learning EDUCATIONAL
MISMEASUREMENT: HOW HIGH-STAKES TESTING CAN HARM OUR CHILDREN (And
What We Might Do About It) W. James Popham University of California,
Los Angeles November 2000 "SES-linked items do, indeed, spread out
students’ test scores.  But SES-linked items don’t measure what is
supposed to be taught in schools. " "If a state’s educators do not
understand why certain high-stakes tests not only yield invalid
estimates of instructional quality, but also are likely to lower
educational quality, those educators cannot inform parents or
policymakers about such problems."

wcala@@ROCHESTER.RR.COM writes:
Yesterday, about 30 of us (NY Superintendents) spent the day with
James Popham to determine our next move to counter the further
development of crap
tests (like the current 4, 8 and high school tests) for grades 3, 5,
7.

2002 - gave seminar to Washington WEA teachers union saying that
standardized tests are a fraud with respect to accountability and
high stakes, presented a video "test boosters" on how to raise scores
by, well, cheating. On TVW cable channel.


http://www.aasa.org/issues_and_insights/assessment/Building_Tests.pdf
Building Tests To Support
Instruction and Accountability
A Guide for Policymakers
Prepared By
The Commission on Instructionally Supportive Assessment
Convened By
American Association of School Administrators
National Association of Elementary School Principals
National Association of Secondary School Principals
National Education Association
National Middle School Association
October 2001
W. James Popham (Chair)
Professor Emeritus
Graduate School of Education and Information Studies
University of California, Los Angeles
Requirement 9: A state should secure evidence that supports the
ongoing improvement of its state assessments to ensure those
assessments are (a) appropriate for the accountability purposes for
which they are used, (b) appropriate for determining whether students
have attained state standards, (c) appropriate for enhancing
instruction, and (d) not the cause of negative consequences.

Fairtest critique - it still assumes the central over-emphasis
and importance of assessing and meeting state standards.
http://www.fairtest.org/k12/response%20to%20ISA%20reprt.html


@@Gap

WHEN REDUCING THE GAP IS ONLY AN ILLUSION

When tests change from a 80 / 40 pass rate for whites and blacks to
something like 99 / 96, it is said that the gap has been eliminated
but in fact the same ratios can be found on tests like the SAT which
have a 1 standard deviation gap. The POINT gaps appear to be smaller
when the test is made easy enough that most students pass, but the
bell curve is still there.

Test Gap Spectrum - Pass rates
43W  33B  Projected 203 TAKS 8th grade

CLOSING THE GAP OR SWEEPING IT UNDER THE RUG? 
96% VS 99% PASS RATE IS STILL A 4 TO 1 FAILURE RATE GAP!

z52\clip\2001\09\lessgap.txt
http://www.newsobserver.com/monday/news/Story/823149p-815448c.html
Wake tests teaching method 
z52\clip\2001\09\lessgap.wk1
Staff Photo By Alexis L. Richardson 
By T. KEUNG HUI, Staff Writer 
GARNER -- Eyes throughout Wake County are being focused on eight
schools that might have found a way to raise student test scores
dramatically and close the racial achievement gap.....

Pass rates:
            91-92        00-01
white       79           99
black       54           96
low income  55           98

"by 2000-01, across-the-board gains had all but eliminated the gap on
all the tests." For instance, 96 percent of Brazosport's black
students and 98 percent of low-income students passed the math test,
only slightly below the 99 percent passing rate among white students"

The truth is that the gap is still there, but it's hidden in a much
easier test when 99% of whites pass. The FAILURE ratio remained the
same - between 2 and 4 times that of whites. So the bell curve gap of
about 1 standard deviation is still there, while the point gap looks
smaller.

===================================================================

Group      Years
           91-92   00-01
Pass
white            79      99
black            54      96
low income       55      98

Fail
white            21       1
black            46       4
low income       45       2

Failiure Ratio
black           2.2     4.0 <- Ratio is actually worse
low income      2.1     2.0 <- 2.0 ratio is the same as before!

==============================================================

LA GRIFFE SHOWS THAT POINT GAP REDUCATION IS ONLY A MIRAGE
http://www.lagriffedulion.f2s.com/gap.htm
Z75\clipim\2004\01\07\achgap\achgap.htm
Volume 6  Number 1 
January 2004 
CLOSING THE RACIAL LEARNING GAP  

"Governor Easley cited data for students in Grades 3-8. In a single
year blacks and Native Americans had narrowed their achievement gaps
by 6 points each, Latinos by 4."  "This diagram shows how a point gap
can be manipulated by playing with the passing score. ..Setting the
passing score at level 1 produces a gap of 34 points. At level 2 the
gap shrinks to 14 points...Changing an exam's difficulty can achieve
the same result in a more insidious manner. A symmetry exists between
the minimum passing score and the test difficulty.In Texas and North
Carolina pass rates have been improving. Both states are in the high
pass-rate region. There, gains in pass rates by whites will be
accompanied by even greater gains by blacks. When pass rates are high,
incremental student gains guarantee a gap decrease..  In the past
decade no meaningful gap reduction occurred in either North Carolina
or Texas

How to become the most celebrated education administrator in the US:

1. Set the passing score on all state tests at the point where 80% of
white students pass.

2. Institute education reforms like those in North Carolina and Texas.
As performance improves, the racial learning gap will decrease,
following the curves of Figures 8 and 9.

3. Continue this policy until the gap is so narrow that further
reduction will be insignificant.

4. Use your newly acquired reputation as a gap-busting administrator
to land a similar job in another state. There, start again with step
1. Three such iterations should guarantee international celebrity

 @@A-Level see @@GCSE


An "A level" is the second level GCE or vocational NVQ level of of
British secondary-school examinations, taken usually two years after
the first GCSE. Usually used for college admissions, it is more like
the SAT or ACT than a state high school exit exam.

Boys do better than girls at the A-level, but not the GCSE
33% of boys, 39% of girls take the A-levels, comparable to US SAT/ACT
rates and US 4 year degree rates.

BOYS DO BETTER AT HIGH RISK FLASH A-LEVEL VS GCSE
JUDITH JUDD, Why flash boys shine at A-level., Independent, 07-10-1996, pp 1. 
www.elibrary.com
\clip\99\09\boyalev.txt
"Teachers questioned by the London and Leicester researchers talked
of boys' "risk-taking approach, their greater willingness to sound
stupid." Girls, they said, "write at length, lack the courage to
discard irrelevant detail and perform less well intraditional exams."
At GCSE, girls gain 8 per cent more A-to-C grades than boys in all
subjects. But at A-level, boys do better, given their GCSE results,
than do girls.  "

NOW 1/3 OF UK TAKES A-LEVELS, GIRLS HAVE EDGE AT HS GCSE, BOYS
AT COLLEGE ENTRANCE MATH SCIENCE
\clip\99\09\girlgc.txt
Judith Judd Education Editor, Macho attitudes key                               
to boys failing., Independent, 07-25-1998, pp 6.                                
In 1985 a fifth of girls stayed on for A-levels.                                
Now the figure is almost two-fifths (39 per cent)                               
compared with only 33 per cent for boys. However,                               
girls continue to take traditional "female                                      
subjects" at A-level. [comparable to SAT taking rates in US]

UK COLLEGE ADMISSIONS BASED ON GCE AS APTITUDE TEST A-AGGRE was the
grade point of the student's GCE 'A' level examinations. It was used
to capture the aptitude and intellectual ability of the students.
Admission into local universities is based solely upon this
aggregate--a prevalent criterion among colleges in the British
Commonwealth system (Lumsden and Scott 1987). "

"Students' performance in economics: Does the norm hold across
cultural and institutional settings? " ( Journal of Economic
Education/Electric Library ) 


@@Alabama

z43\clip\2000\07\stanal.txt Testing the system Local schools find out
scores, fate Thursday By CINDY FISHER BIRMINGHAM POST-HERALD On
Thursday, scores will be released for one of the most important tests
given in Alabama public schools: the Stanford Achievement Test.  If
scores on the national standardized test don't show students are
learning, schools could be taken over by state officials. 

1985 - 6th grade test for graduation
1993 - 8th grade test for graduation
1999 - pilot 11th grade test for graduation.

ALABAMA CALLS G6-8 TEACHER TEST TOO EASY WHEN GRAD TEST IS G11
\clip\99\10\altest.txt
Montgomery Advertiser Friday, July 16 "Schools chief calls test too
easy "It never occurred to me tht a test like this, which is taken
basically by college students at the end of their sophomore year as
they are about to go into their junior year would be at such a low
level.  ..The proposed agreement would provide remedial courses for
prospective teachers who failed the test ..While the Praxis I has a
middle school level of difficulty, Alabama is starting a graduation
exam for students that is based on questions that have an 11th grade
difficulty 

A+ is equivalent to Goals 2000.  They were the consensus building
part of David Hornbeck's plan (known in Alabama as Governor Folsom's
reform plan, the Teague Plan...).  They have not given up.  Now they
have co-opted the current move to "better teachers" so they can force
their agenda forward through teacher certification (using
"child-centered" "Experiential learning" "group learning", etc.).
Forgive me for not being clearer.  Sharman

ALABAMA ADOPTS EXIT EXAM FROM HELL Birmingham News: Students facing
tougher exit test February 27, 1999 Students facing tougher exit test
CHARLES J. DEAN News staff writer 02/26/99 Question: Ryan makes $3
per hour more than Scott. If 3 times Ryan's rate plus 4 times Scott's
rate is $65, what is Ryan's hourly wage? Answer: $11. (solve ax+bx=c
algebra) Question: A simple majority in Congress votes for a proposed
amendment to the Constitution. What happens next? Answer: It's dead.
Such a proposal requires a two thirds vote to pass. (no adult knows
this, even if they look it up) state school superintendent, has
predicted that possibly 50 percent of students in one third of the
state's high schools might flunk the exam the first time around. 

ALABAMA MAY LAY OFF STANFORD ACHIEVEMENT TEST Alabama: Montgomery
Advertiser, Monday, Nov. 30, 1998. Page 1.  "Educators want new
standards" Test case: The Alabama BOE may ask legislators for an
alternative way to measure student achievement by John F. Maclean
Morton said 97,000 of the 440,000 students tested on last year's
Stanford Achievement Test scored below the 23rd percentile, or in the
lowest quadrant of the test. (22% is actually better than national
average distribution!)


@@Alaska

In 1993, Alaskans embarked on a campaign to bring higher standards
and accountability to their public school system. The cornerstone of
this effort was the development of content standards in ten core
subject areas.  The standards represent what Alaskans want students
to know and be able to do as a result of their public schooling.

Alaska is (1998) developing an exit exam that would turn their  
diploma inti a 12th grade cim

Alaska test plan
test response

June 30, 2000 Anchorage Daily News
EXIT EXAM TEST SCORES INDICATE PUBLIC EDUCATION IS IN TROUBLE
By Mike Doogan Daily News Columnist
If the first high school exit exam is any indication, public education in
Alaska is a disaster. Two of three sophomores failed the math test. One of
two flunked writing. One of four failed reading. 

www.educationnews.org 2/2000 Anchorage Daily News Graders to
'bookmark' exit exams 2.4 JUNEAU - Lawmakers were reassured Thursday
that the method the state will use to set the cutoff point for
passing high school exit exams is fair and well-researched and avoids
pitfalls other states have encountered.

State will refine tests this year  2.4
What are the exit tests?
They are three tests - in reading, writing and math - that high school
students must pass to earn a high school diploma in Alaska beginning in
2002. The tests are also known as the High School Graduation Qualifying
Exam. This year they'll be given statewide March 7, 8 and 9.


@@Alliance

ALLIANCE OF CONSERVATIVES, LIBERALS, CORPORATE, EDUCRATS 4 STANDARDS
z68\doc\web\2003\07\standref.txt
http://www.louisville.edu/journal/workplace/issue5p2/rougeforum.html
.. has been hit with at least some degree of criticism—notably both
from the Right, which sees standards-based reform as imposing on local
school district autonomy, and from the Left, which sees it as racist,
sexist, and classist.. willingness to take such criticism seriously
yet still maintain that it can satisfactorily be accommodated by ..
the prevailing framework.  "the demand for standards-based reform
itself—the standardization imperative—goes unchallenged, at least
among the alliance of conservative and liberal politicians, corporate
elites, chief school officers, and teacher union leaders.  "


@@Alternative Assessment
@@Alternative Diploma

TORONTO ELMIMINATES NON-ELITE DIPLOMAS, OFFERS SKILL CERTIFICATE
\clip\2004\02\torskill.txt 

Fri, 20 Feb 2004 Toronto Star Certificate rewards job skills, not
marks Toronto board to issue 200 in June Seen as alternative to
dropping out KRISTIN RUSHOWY EDUCATION REPORTER
      Toronto high-school students doomed to fail because the new
curriculum is too tough will be able to earn a certificate showcasing
their job skills rather than their school smarts. 
In the past, many of the students who will be eligible for the skills
certificate would have studied at the "basic" level and earned a
diploma.  Now, the basic level has all but disappeared, leaving those
students struggling to learn material beyond their capability, or
taking courses developed locally by boards that often don't count for
credit. ...  Almost half of students who begin high school end up
going straight into the workforce - with or without a diploma. 

1990: NCEE GETS GRANT FOR ALTERNATIVE ASSESSMENT
z40\clipim\2000\03\19\edgrant.efx Education Grants for New Kinds of
Tests Dec 17, 1990 SF Chronicle University of Pittsburgh and NCEE
will use grants to develop tests based actual performance and
portfolios. From Pew Chartible Trust and John D. MacArthur foundation.

1st graders tested by 5th graders rated
by happy and sad faces
"Student Centered Classroom Assessment" by Richard Stiggins and
"Seven Kinds of Smart" by Thomas Armstrong.

ETS: ASSESSMENTS DON'T TEST FOR TEAMWORK, CONSTRUCTIVISM
zip36\clip\99\17\cheating.txt LA Times Sunday, Aug. 22 Cheating in
School Reflects Basic Confusion in Society By emphasizing the wrong
things in student testing, we end up inviting a culture of
compromise. By MARI PEARLMAN. Wants tests based on collaboration,
constructivism, not single correct answers, scary coming from makers 
of the best multiple choice tests we have as compared to awful 
performanc baesed tests.


@@Answer Disclosure

Mother wins right to see incorrect test answers on CIM test
http://www.eagleforum.org/educate/1999/sept99/ct-mother.html
Connecticut Mother Wins Battle With State Education Department
Five-year struggle brings victory for parents' rights [to see which
questions she was scored wrong and why, but not to see the entire
testj]


@@Apartments

BUY LAND TO KEEP APT KIDS OUT TO KEEP TEST SCORES HIGH
z48\clip\2001\03\1strike.txt From free St. Louis paper:
http://riverfronttimes.com/issues/2001-03-07/shortcuts.html First
Strike Hazelwood schools blocked apartments to protect test scores IN
Riverfronttimes.com, By D.J. Wilson. Buys land from developer to keep
out transient kids in apartments who lower test scores.


@@Arizona AIMS
@@AIMS

Arizona will tie graduation diploma to passing the AIMS test. The
math test was so difficult, one parent who graduated with an
engineering degree and law school could not pass the test. SPI Keegan
left town after AIMS technical problems forced a rewrite and the
public demanded delaying high stakes or killing it entirely.

"I was wrong," Ms. Keegan, a speech pathologist first elected
superintendent of public instruction in 1994, said in an interview.
"But if we're going to shoot wrong, I want to shoot too high and then
moderate. "

90% of whites, 97% of blacks fail the math section.

Arizona Republic AIMS test articles
http://www.azcentral.com/news/education/aims.shtml

%%Errors

AIMS SAMPLE MATH PROBLEMS ARE MATHEMATICALLY INCORRECT
Experts find numerous math errors. Substance June 2001 p. 46
Christian Richardson. "If the released items are representative of
the AIMS, many of the items are simply incorrect mathematically" said
James Middleton, Arizona State University professor of mathematics. "
Many have confusing wording or have multiple correct answers". The
state program administrator says "it is possible for errors to slip
through the publising company".

%%General

PARENTS SUE AIMS, SAYS STATE HAS NOT PREPARED STUDENTS TO PASS
clip\2006\04\aimslaw.txt
http://www.azcentral.com/arizonarepublic/local/articles/0420governor-aims0420.html
azcentral.com
Governor: AIMS hurdle still in place
Associated Press Apr. 20, 2006 12:00 AM
Gov. Janet Napolitano said high school students shouldn't read too
much into a class-action lawsuit that tries to block use of the AIMS
test as a graduation requirement.

KEEGAN VANISHES AFTER AIMS BOMBS WITH PUBLIC
z51\clip\2001\09\mumaims.txt
http://www.arizonarepublic.com/arizona/articles/0909b3update09.html
Ex-state schools chief mum on AIMS Sept. 09, 2001 technical problems
forced Keegan to keep moving back the effective date of the
graduation test. By early this year, state lawmakers threatened to
put the test to a statewide vote. Amid the turmoil, Keegan resigned
in May

"Student Failure Causes States To Retool Testing Programs" Jacques
Steinberg, New York Times Friday, December 22, 2000, FRONT PAGENot
only did Arizona education officials put material on the 10th-grade
math exam, including some calculus and trigonometry, that most
students would not be taught until 12th grade.. in the early 1990's
the state had introduced a precursor to AIMS, called ASAP (Arizona
Student Assessment Program), standardized tests that used a common
subject like the rain forest as grist for reading, math and writing
questions. Soon after her 1994 election, Ms. Keegan killed the
program, saying its value was amorphous.  ... to delete from the test
those concepts that they considered too advanced, including analysis
of matrices and writing of algorithms. .. after the second round of
failing test scores (16 percent passed the math, 33 percent passed
the writing) were released this fall, Ms. Keegan pulled back again.

z46\clip\2000\11\keegan.txt Nov 22
http://www.arizonarepublic.com/news/articles/1122aims22.html Arizona
Republic Keegan backs off AIMS requirements Faced with dismal scores
and relentless criticism, state education chief Lisa Graham Keegan
waved a white flag Monday, putting the state's first high school
graduation test on hold. 

z43\clip\2000\07\azcanc.txt
ARIZONA TO SCRAP FALL AIMS TO GET RID OF HIGHER ORDER MATH
Jul 1st, 2000 edition of The Arizona Republic online,
http://www.arizonarepublic.com
Curiouser and curiouser State may cancel fall AIMS test The
controversial AIMS graduation test needs so many changes that this
fall's exam may be scrapped.
http://www.arizonarepublic.com/cgi-bin/gp.php3?f=/news/articles/0701aims01.html   
The Arizona Department of Education is pushing for the delay because
it wants to remove short-answer math and reading questions, said
spokeswoman Patricia Likens, making the test easier to grade and
quicker to score.
The Education Board already agreed Monday to remove higher level math
from the exam because many students had not learned those concepts. 

February 24, 2000 Arizona Republic AIMS math test may get easier More
The math section of the AIMS graduation test may be softened if
students fail it again this spring, Arizona's schools chief said
Wednesday, departing from her usual defense of the exam.

\clip\99\20\aimsmath.txt http://www.azcentral.com/news/1123aims.shtml
AIMS dominates discussion at education board meeting Board to
refigure AIMS math test By Kelly Pearce The Arizona Republic Nov 23,
1999 

                         Reading Math  Writing
Exceeds                  14      0      1
Meets                    47     11     29
Near or below standard   38     88     70

Parents, teachers and administrators sent the state Board of
Education a message Monday: The math test it requires for high school
graduation is way too difficult. 

AIMS PROBLEMS WAY BEYOND 10TH GRADE MATH
HOW CAN YOU BE WORLD CLASS AND FLUNK 90%????
http://www.azcentral.com/news/1117math.shtml AIMS scores not so bad,
pros say Some items not covered in class Related Articles Both Garcia
and 17-year-old classmate Gordon Lafontaine were in Algebra 3 when
they took the AIMS test. But they said there were problems on the
test that they had no idea how to answer.  They could answer them
now, they said, after spending almost a semester in pre-calculus. 
"Nor do the scores mean that the students are stupid, especially as
other tests show Arizona kids doing about as well as kids elsewhere.
..State scores are at the national average in math on most standardized
nationally normed tests given in Arizona, Glass said. Globally,
Arizona scored at a level statistically equivalent to Germany,
Denmark and Israel. 
"

TEST REQUIRES ANOTHER $1200 PER STUDENT?? 
http://www.azcentral.com/news/1117keegan.shtml
$1,200 per pupil per year urged for AIMS test By Mike McCloy The
Arizona Republic Nov. 17, 1999 
State schools chief Lisa Graham Keegan believes she can get most
Arizona students to pass the AIMS test at an added cost of about
$1,200 a year per child....  ""We are going to have a train wreck,"
he said. "We are going to end up only giving 20 percent of our high
school seniors diplomas."  "



ARIZONA TEST FLUNKS 90% OF ALL, 99% OF MINORITIES
http://www.azcentral.com/news/education/1115aimsresults.shtml
\clip\99\20\aimsfail.txt AIMS scores fall short 9 of 10 test-takers
fail to meet minimum standards By Kelly Pearce and Jennifer Barrett
The Arizona Republic Nov. 16, 1999 Nearly 9 of 10 Arizona high school
sophomores failed to meet minimum mathematics requirements last year
on the AIMS test.  Approximately 99 percent of 124 Dysart High
students failed the math portion, but Koch said the poor scoring
extended beyond his district's boundaries.

AZ SUPERINTENDENT BARELY PASSES 10TH GRADE TEST AT 90TH PERCENTILE
http://www.azcentral.com/news/1116keegan.shtml Passed: Keegan
'relieved' at her AIMS scores By Kelly Pearce The Arizona Republic
Nov. 16, 1999 Keegan admitted that her math score - 520 - could have
been better. A score of 500 was "meets standards." 

ARIZONA HIGH STANDARDS A MODEL FOR "REAL" REFORM?
\clip\99\19\ariz.txt
http://www.manhattan-institute.org/html/cb_19.htm Center for Civic
Innovation Manhattan Institute Civic Bulletin No. 19 July 1999
Transforming American Education by Lisa Graham Keegan Lisa Graham
Keegan is Superintendent of Public Instruction for the State of
Arizona. [Arizona also has high standards gone amok with failure
rates so high they didn't even want to release the scores] What’s
going on in Arizona’s education system is really a backdrop for
what’s going on across the nation. In Arizona, academic achievement
and the motivation for children to be brilliant is increasing.

AZ TEST RESULTS SO BAD THEY DON'T EVEN WANT TO RELEASE THEM!
zip36\clip\99\16\aims.txt
http://www.azcentral.com/news/education/0824aims.shtml
AIMS release opposed
Education Board urged to be cautious										   
By Kelly Pearce The Arizona Republic Aug. 24, 1999 
																			   
The president of the state Board of Education isn't so sure she wants
the public to see the first round of scores for Arizona's Instrument
to Measure Standards.  Results of AIMS will be released in October to
today's high school juniors, 60,000 of whom took the reading, writing
and math exam last spring. But Mary Garcia voiced concerns during the
board's Monday meeting about announcing to everyone how many students
conquered the state's academic standards and how many didn't.  

HEY WE WERE ONLY KIDDING...  Arizona Republic Education board chief
relents One day after she spoke out against publicly releasing the
first round of scores of the AIMS test, the president of the Arizona
Board of Education backpedaled and said she favors making available
some general, statewide numbers. 

%%Diploma

ARIZONA TO MAKE THEIR GRAD TEST MANDATORY EVEN FOR OUT OF STATE XFER

Note - Arizona has a republican governor and republican
superintendent who got 98% of vote in 1998. Both appear to be
strongly committed to "standards based education". Tying graduation
to test is functionally equivalnet to Marc Tucker's CIM idea, and
consistent with OBE requiring demonstration of high performance
rather than passing seat time.

http://www.ade.state.az.us/programs/foundations/ssa/AIMS/7-27-99update.htm
\clip\99\15\aims.txt School and Student Accountability Division July
27, 1999 Test to be offered up to age 22, not decided on giving test
to private schools, to be required even if you move in from another
state.

\clip\98\17\edclip04.txt AIMS rescheduled Today's freshmen to be 1st
group to test for diplomas By Cathryn Creno The Arizona Republic Nov.
24, 1998 

ARIZONA TEACHERS UNION SUPPORTS 3 LEVEL -CIM+ DIPLOMAS
http://www.azcentral.com/news/education/1119aims.shtml
\clip\98\17\aims.txt Teachers urge 3 levels of high school degrees By
Cathryn Creno The Arizona Republic Nov. 19, 1998 ... the math on the
AIMS is way too hard.  "My 11th-grader took the pilot test and told
me he bombed it because the questions were so terrible and bizarre,"
Leih said. She intends to keep her 10th-grader home the day the AIMS
is given next spring. ..is astounded by parents who do not think
their kids should have to know algebra and geometry to graduate. 

Diploma - CIM, CIM, CIM+ based on passing AIMS

%%Difficulty

AIMS MATH USELESS AS HIGH SCHOOL EXIT OR EMPLOYER REQUIREMENTS.
"SOMETHING IS WRONG WITH THE TEST"
http://www.asu.edu/educ/epsl/EPRU/documents/EPSL-0210-122-EPRU.html
The AIMS Test and the Mathematics Actually Used by Arizona Employees
Gene V Glass Arizona State University Cheryl A. Edholm Longwood High
School Middle Island (N.Y.)  Executive Summary Arizona’s Instrument to
Measure Standards (AIMS) produces unusual failure rates at Grade 10 in
mathematics.  For the typical Grade 10 AIMS mathematics question, only
three of 43 managers reported that it tested mathematics actually used
by their employees.  " a hurried process of content selection in 1995
that was subject to manipulation by external consultants to the
Arizona Department of Education and designed to exclude teachers in
the very subjects of their expertise.....one-to-one identity of the
AIMS Mathematics standards with the curriculum recommendations of the
National Council of Teachers of Mathematics, a document whose authors
have asserted is not a valid framework for a high-school exit
examination" ..."extraordinary failure rates on the AIMS test,
particularly the Grade 10 Mathematics test, have reinforced the
suspicion that something is wrong with the test"

The Texas high school exit exam, which fails approximately 10 percent
has much easier questions.  The most difficult math question in one of
the earliest administrations of the TAAS tested whether students could
calculate the size of an envelope required to hold an 8.5” X 11” piece
of paper folded vertically in thirds

\clip\98\19\aims.txt
http://www.azcentral.com/news/education/1222aims.shtml
State refigures AIMS math Education board OKs condensing 5 tests into
1 By Julie Amparano The Arizona Republic Dec. 22, 1998
"One parent, Fred Gamble, said he couldn't pass the test even though
he had an engineering degree and graduated with honors from law
school.  "

%%Poll

http://www.azcentral.com/news/education/1030aims.shtml Poll: 82%
support test for graduation Officials call results endorsement of
AIMS By Cathryn Creno The Arizona Republic Less impressed was K.R.
Scott, a Phoenix Mountain Pointe High School government teacher
running as a write-in candidate for state superintendent of public
instruction.  I think people support the idea of a high school
graduation test, but many strongly oppose this specific tool.
"Parents of sophomores are worried that nothing their children have
done in the last 12 years will be measured on the test."... the
state superintendent of public instruction, were shouted down at a
forum attended by about 300 Mesa and Tempe parents, kids and
educators 


@@Arkansas

z39\clip\2000\02\akscore.txt
Sat, 26 Feb 2000 TEST SCORES DON'T TELL WHOLE STORY OF SCHOOL
PERFORMANCE JEFF PORTER ARKANSAS DEMOCRAT-GAZETTE This school year,
Arkansas fifth-graders improved their performance on the Stanford
Achievement Test. Or maybe they didn't. [test scores rise due
to exclusion of the worst students]



"Richard G. Innes" <70224.434@compuserve.com> comments It is indeed
telling that Arkansas excluded much of their student with
disabilities population from their own testing program.  Arkansas
kids with disabilities have had 4th grade reading scores reported
from the National Assessment of Educational Progress each time the
test was given.  Here is how the scores for Arkansas students with
learning disabilities look:

Year    NAEP Score
1992    177  1994    157 1998    152

Small wonder the state wants to bury this information in their own testing
program.  You can bet most folks in Arkansas never heard of the NAEP.


@@Arts

Boston Globe 5/15/2000
Stronger schooling in the arts
On paper, it sounds like the best of ancient Athens. Public school
students in Massachusetts are supposed to sing and play instruments;
choreograph dances; read, write, act in, and build sets for plays; learn
about art history; draw, paint, and sculpt. The blueprint for this is the
Massachusetts Arts Curriculum Framework, adopted last June. Two goals run
through this Board of Education document: Teach students skills and make
them flame-bearers of human achievement in theater, music, the visual
arts, and dance. 

ARTS COMPETENCY TESTS? ARGH!
z40\clipim\2000\04\12\test.efx
In New Jersey, Be Creative or Else New York Times Education Life
April 9, 2000 1997 Nick Chiles NAEP gave an arts test where only 5
percent were graded best on self-portrait, 16% made a design of how
to transport a fish on a bicycle, whites did best in music, but
blacks approached whites on rhythmic embellishment. Whites and Asian
did best in visual arts, girls outdid boys in everything.



@@Backlash
see Testing Backlash




@@Bias, Minority

It is thought that conventional norm-referenced multiple choice tests
are "biased" against minorities. Marc Tucker believes that they
under-estimate the abilities of minorities. But the bottom line is
for all the minority friendly features of performance-based tests,
the gap is even wider on the new tests. Since open-response is harder
than multiple choice, and testing higher order thinking is more
difficult than memorization, this should be expected.

DIFFERENTIAL PASS RATES:
Percent Passing / Above average
Ranked by pass ratio
        
 White Black Ratio SD                                    SD units  
 93%  75%    3:1        1994 Seattle Police Written Test   0.80
 65%  28%    2:1        2002 CA High School Exit       
 60%  40%   1.5:1  .5?  G4 Seattle 1993 CTBS norm          0.51 ???
 43%  15%    3:1   .9?  G4 WASL 1999 Wash perf             0.86
 35%  13%    3:1   .9?  G4 WASL 1998 Wash perf             0.74
 35%   9%   3.5:1       1995 Seattle Video Police Test     0.96
 29%   2%   15:1   ?    G11 Stanford 9 Boston perf         1.50 ???
 25%   5%    5:1   .9   G4 WASL 1997 Washington perf       0.97
 15%   3%    5:1   .9?  G8 CLAS 1994 Calf perf             0.84
 
Last column estimated SD difference from pass rates by 
La Griffe du Lion il_leone_uno@@yahoo.com

PERFORMANCE BASED TESTS FAVOR GIRLS BY 2:1
Published Sunday, January 17, 1999, in the Herald-Leader FRANK
ANDERSON/STAFF KIRIS test's format helps girls top boys By Linda B.
Blackford Lexington KY HERALD-LEADER EDUCATION WRITER
"In 1998, for example, nearly twice as many girls as boys scored in
the highest category in reading, 37 percent to 19 percent. The
disparity increased from 1997, when the difference was 15 percentage
points."
That same pattern follows in Maryland, one of the few other states
with a performance-based test.
Note - activists are still concerned only when girls are outdone
by boys. No concern that boys are being left behind.

"The department conducted one study on race and gender in 1996 and
1997, but has refused to release it, saying it is a draft and needs
work."


@@Bias, Political

Cunningham and Garner Phillips
(1995) provides tables that quantify the weight given to different
topics.  She found that on the 1993-94 KIRIS assessment in reading,
23% of content focused on the experience of historically
disadvantaged groups, 13% respect for animals, 35% health, 17%
environmentalism, 7% criticism of traditional education, and 3%
evolution.  This left 3% for work and job issues, the only
non-controversial content.  None of the reading passages used content
from art, history, science fiction, space, and computers although
these made up 12% of the content in 1991-92 and 6% in 1992-93.  A
statewide Sensitivity Review Committee concluded that: · Taken as a
whole, the test favors a liberal viewpoint.  · Many questions appear
to seek a politically correct response.

   
@@Bilingual Testing


BILINGUAL SF AND SAN JOSE SCORE BETTER THAN NATIVE - ASIANS?
\clip\98\10\biltest.txt From: ADP 
http://www.sfgate.com/cgi-bin/article.cgi?file=/chronicle/archive/1998/07/07
/MN10594.DTL [How many of these "bilingual" students are Asian, who
routinely get better math scores??]  Bilingual Surprise In State
Testing Many native-English speakers outscored in S.F.  San Jose
Nanette Asimov, Chronicle Staff Writer Tuesday, July 7, 1998 San
Francisco Chronicle. District-wide 42-50 in math, bilingual 74-85!

SAN FRANCISCO
..
Grade 2
Reading Math Lang.
Native English speakers 49 50 47
Bilingual students 58 65 60
Bilingual ed. graduates 75 87 80

Grade 4
Reading Math Lang.
Native English speakers 49 42 48
Bilingual students 56 56 58
Bilingual ed. graduates 66 79 73

Grade 6
Reading Math Lang.
Native English speakers 47 50 47
Bilingual tudents 53 63 57
Bilingual ed. graduates 58 79 64
..
Grade 9
Reading Math Lang. S.S.(x)
Sci.
Native English speakers 50 64 61 54 50

Bilingual students 56 72 66 58 56

Bilingual ed. graduates 44 74 63 52 53

..
Grade 11
Reading Math Lang. S.S.(x)
Sci.
Native English speakers 56 64 64 67 54

Bilingual students 61 73 68 72 62

Bilingual ed. graduates 46 74 59 66 54


@@Blog

http://www.homestead.com/swygert/files/no2pencil.html
Who am I?
My name is Kimberly Swygert, and I'm a Research Scientist at a non-profit
large-scale high-stakes standardized testing organization. I prefer not to
say which one, not because I don't like this place, but because I want to
ensure as much as possible that my opinions posted on this blog are not
confused with the official opinions of my employers.
       
         
@@Bookmark (Setting Standards)

A bookmark procedure appears to be common to most standards - based
tests - a committee establishes a "bookmark" in an ordered list of
tasks, but the panelists are not allowed to know what % of students
got the answer, or consider wheher the task is compliant with the
written standards. The result is very high failure rates when given
to actual students.


ALASKA - OUR STANDARDS WONT BE A TRAIN WRECK (YEAH RIGHT)
http://www.EducationNews.org for article links
February 4, 2000
Anchorage Daily News
Graders to 'bookmark' exit exams  2.4
JUNEAU - Lawmakers were reassured Thursday that the method the state
will use to set the cutoff point for passing high school exit exams is
fair and well-researched and avoids pitfalls other states have
encountered.


http://www.dpi.state.wi.us/oea/profsec2.html II. HOW WERE THE
PROFICIENCY CUT SCORES SET?  The standard setting procedure selected
for use in Wisconsin was designed and conducted by the testing
contractor, CTB/McGraw-Hill. The panelists placed "bookmarks" at the
item in their item ordered booklets that represented the breaking
point between the proficiency categories: minimal performance, basic,
proficient, and advanced. 

Colorado As I recall, the tests in
Colorado were developed by CTB/McGraw Hill using groups of teachers
(and community persons?) who developed the questions, then they were
field tested, then the groups had each member mark where they thought
the various breaks should fall (mastery, proficient, etc.).  Then
they showed the groups the results of those breaks.  The groups were
shocked, and did make some adjustments, but they did not cave
completely.  The resulting data was, I think, fairly comparable to
NAEP results. A similar phenomenon in testing occured in TN when our
writing tests were developed.  Groups of teachers developed them and
set the standards, then were shocked at how poorly their students
performed. Dave Shearon
Nashville, TN

Seattle and Rhode Island describe similar processes for determining
proficiency levels.


@@Boys

z47\clip\2001\01\boytest.txt The Electronic Telegraph ISSUE 2065
Friday 19 January 2001 Boys 'lose out' in feminised exams By Liz
Lightfoot, Education Correspondent SEPARATE exams for boys and girls
were suggested yesterday to overcome the bias of the present
"feminised" system.  Madsen Pirie, president of the Adam Smith
Institute, said that modular courses and continuous assessment
favoured the systematic approach of girls rather than the
"risk-taking" of boys.


@@Boycott

GazetteNet
ARHS sophomores boycott MCAS
Wednesday, May 24, 2000 -- (AMHERST) - Nearly 20 percent of the sophomore
class at Amherst Regional High School is boycotting the Massachusetts
Comprehensive Assessment System tests. 


@@Boys

Boys generally do worse on the new performance based tests than
girls, even on math.

http://washingtonpost.com/wp-dyn/articles/A4236-2000Dec14.html
z47\clip\2000\12\mspap.txt MSPAP Gauges Student Performance Far
Better Than Nationally Normed Tests By Karin Chenoweth
http://washingtonpost.com/ac2/wp-dyn/A4236-2000Dec14?language=printer
Thursday, December 14, 2000 ; Page M10 
BOYS SCORE WORSE
"Generally, you will see that boys are testing much less well than
girls. Consider that 48.2 percent of eighth-grade girls scored at a
satisfactory level in writing and 15.2 percent of those at an
excellent level. But only 28.5 percent of eighth-grade boys were
satisfactory and only 6.9 percent of those were excellent. In reading
statewide, there is a gap of more than 10 points between boys and
girls."



@@Bracey, Gerald


Critic of education fads finds new assessments are ridiculously hard
when everybody flunks.

Bracey says Third graders
across the nation are throwing up on test day and there have been
test-induced suicides and nervous breakdowns documented in New York,
California, and Arizona.  No wonder.  Children are being asked to do
the impossible.  I define World Class as the 65th percentile on a
standardized test. Why does a World Class district like Gwinnett need
further proof of its stature? 

@@Bush, GW

MANDATE FOR EQUAL TEST SCORES DOOMED TO FAIL
http://www.worldnetdaily.com/news/article.asp?ARTICLE_ID=25708
zip54\clip\2001\12\medscore.txt Dec 17 2001 
Fat kids,
racial disparities and education reform © 2001 WorldNetDaily.com By
...demanding equal or nearly equal scores among all ethnic groups,
the current effort at educational reform is sure to fail. Few if any
schools will succeed in erasing racial disparities that relate far
more to disadvantages in home environments than unequal treatment by
educators. 

BUSH RECOMMENDS STANDARDS BASED, NOT NORM REF TESTING
z48\clip\2001\02\edwktest.txt February 21, 2001 Usefulness of Annual
Testing Varies by State By Lynn Olson Education Week ideally, the
administration wants states to use standards-based exams, but that
they could use an off-the-shelf, norm-referenced test if they chose.


@@California

The 1992-93 CLAS test was widely acknowledged to be a disaster
because it was unreliable to grade, and was scored more for
feelings than correct answers

The newer SAT-9/STAR test is a valid test for most students, but it
has been slameed for being required of students who are not
proficient in english, and the math is expecting perhaps too much
"real math" instead of not enough. It is also being used as a OBE/
TQM baseline to drive performance quotas, which is still a bad idea
no matter what test you use.

%%CLAS

HOW WHOLE LANGUAGE AND OUTCOME BASED EDUCATION FAILED IN CA
http://www.claremont.org/gsp/gsp49.htm
#1996-49
March 15, 1996
THE DECLINE OF ACADEMIC STANDARDS IN CALIFORNIA EDUCATION
The Story Behind the Student Testing Fiasco
Just what was wrong with the CLAS test? It evaluated students more on
their feeling than their analysis of a subject. It permitted students
to take certain sections of the test in groups. It didn't require
students to write complete sentences, much less complete paragraphs.
In some cases, students could draw pictures instead of write their
answers. In short, it furthered the state's policy of eschewing basic
skills for so-called "problem solving" skills. 
z43\clipim\2000\07\06\clas\clas.htm

MATRIX TEST SON OF DISASTEROUS CLAS TEST
\clip\99\06\matrix.txt
Another Test DEBRA J. SAUNDERS S February 14, 1999

EARLIER this month, the word started to percolate.  A 1995 state law
called for two statewide school tests. It first prompted California
to buy the multiple-choice STAR test, and then mandated another more
intensive test with open-ended questions -- a matrix test. Wags said
that state school chief Delaine Eastin's Department of Education was
trying to turn the matrix test into a Son of CLAS test.

11/02/97 CA SELECTS TERRA NOVA TEST
caltest.txt

Terra Nova is just repackaged CTBS
and it's a scam.

%%C-TAP VOCATIONAL / TECHNICAL ASSESSMENT

CALIFORNIA DEVELOPING CAREER TECHNICAL STANDARDS ASSESSMENTS
z48\clip\2001\03\ctap.txt
The tests to be developed were part of the Career-Technical
Assessment Program (C-TAP) for the California Dept. of Ed. by
Sacramento County Office of Education, with a total (then) of
$4,321,562.00

C-TAP will use performance-based assessment strategies, including
projects and specific work products collected over a period of time,
as well as on-demand assessments conducted within a specific time
limit (written, timed tests).  These "career-prep" components will,
among other things, demonstrate knowledge and skills needed within a
particular industry, as well as those that are common to a set of
related occupations WITHIN an industry.  Thus, I expect the SCANS
standards relating to specific industries and occupational clusters
will be used.
'\

%%Delay

GOODBYE STEM AND LEAF BOX AND WHISKER PLOT, MEDIAN, QUARTILE
z75\priv\2003\11\catest.txt
Los Angeles Daily News
High School Exit Exam to be made easier
By Jennifer Coleman
Associated Press
state Board of
Education voted Wednesday to remove some of the more difficult math
and English questions on the California High School Exit Exam.
"missing will be the lower quartile, median and maximum of a data
set. The board also deleted questions that had students use
"stem-and-leaf" and "box-and-whiskers" plots for data displays.
Instead, students will face more questions about computing averages"
Wednesday, November 12, 2003

40% FAILURE RATE - DELAY EXIT TEST BY 3 YEARS OR FOREVER?
z67\clip\2003\05\gradexam.txt
http://www.latimes.com/news/local/la-me-exit2may02,1,5598089.story
State Education Official Seeks to Delay Exit Exam
Report is cited showing high failure rate on the prerequisite to diploma.
"Only about 60% of the students in the class of 2004 have passed the
math portion of the test so far" 
By Duke Helfand, Times Staff Writer

z76\clip\2003\05\testhard.txt
http://www.sfgate.com/cgi-bin/article.cgi?file=/c/a/2003/05/02/MN15878.DTL
Friday, May 2, 2003 (SF Chronicle)
Exit exam too hard -- reprieve likely/State may delay denial of diplomas to students who fail
Nanette Asimov, Chronicle Staff Writer
Human Resources Research Organization of Virginia, concludes
that the exit exam "meets all of the test standards for use as a
graduation requirement."

   California High School Exit Exam Passing Rates for Class of 2004(x)
                           2001-2002    English  
                              10th-     language 
Group                       Graders       arts    Math 
All students                 459,588       81%     62% 
Economically disadvantaged   125,139       79%     54% 
English learners              77,446       54%     37% 
Special education             47,169       40%     22% 
   (x) As of January, 2003
   Source: Human Resources Research Organization  

%%High School

LOS ANGELES WHITE, ASIAN SCORE ABOUT SAME AS STATE
NEARLY HALF OF BLACKS FAIL MATH
% statewide math, english
All       74 75
African   54 62
Asian     91 85
Latino    61 62
White     87 88
% Los Angeles Unified
All       58 62
African   49 60
Asian     91 82 (about = state)
Latino    53 57
White     85 88 (about = state)
% Manual Arts High School
All       39 53
African   29 48
Latino    42 54
August 27, 2004 
By Duke Helfand, Times Staff Writer
Los Angeles Times

z60\clip\2002\10\caexam.txt
http://www.signonsandiego.com/news/education/20020930-1012-ca-highschoolexam.html
More than 50 percent of students fail high-stakes graduation test 
By Jessica Brice
ASSOCIATED PRESS 
September 30, 2002 

SACRAMENTO – More than half of the high school students who took the
state's high school graduation test this spring failed its math and
English sections, meaning they will have to retake the test or they
won't receive a high school diploma. 
Rate Index  Race
28 -2.32    B
30 -2.16    H
65  1.00    W 
70  1.07    A

CA HIGH SCHOOL MATH TEST PROJECTED WITH 50% FAIL RATE
Z46\DOC\WEB\2000\10\CATEST.TXT Updated Information on the High School
Exit Exam: Students exempted from the exam will not receive diplomas.
Districts may issue "certificates of attendance" to such students.
The estimated failure rate on the recent field test was 40% in
English language arts and 50% in math.

Cunningham: HUMMRO has just completed a 100 page evaluation of the
Ca.  HSEE and it is available at
http://www.cde.ca.gov/statetests/hsee/hsee.html .  This is a
remarkable document for several reasons.  First, it is extremely long
and detailed.  For those of you in California, or those who are
interested in assessment or the HSEE in particular, you can learn
just about everything from this document.  It is remarkable that such
a lengthy and complete document would be published before the test is
even administered.  All that has occurred so far is pilot testing.
Second, unlike most evaluations, this is a very frank and sometimes
critical evaluation. 

"For over half of the mathematics items and more
than 90% of the English language arts items, the average ratings
across districts suggested that at least one-fourth of the 10 th
grade students had not received instruction that would allow them to
answer the test item correctly." results from the analyses of field
test data showed that the test items are relatively difficult for
today’s 10 th graders, particularly in mathematics. If these items
reflect what we believe students need to know and be able to do, and
several panels of reviewers believe that they do, then a significant
number of 10 th grade students are likely to fail this exam.

%%New STAR tests

STUDENTS BELOW 40TH PCTILE WON'T PASS, NEED TO TAKE SUMMER SCHOOL
z55\clip\2002\04\casumm.txt Summer school takes serious turn Teachers,
students focus on state tests Julian Guthrie, Chronicle Staff Writer
Monday, April 22, 2002 ©2002 San Francisco Chronicle
http://www.sfgate.com/cgi-bin/article.cgi?file=/c/a/2002/04/22/MN68516.DTL
San Francisco Unified students who score at or below the 40th
percentile on the state's standardized exam -- the 50th percentile is
considered average -- are being "strongly urged" to attend summer
school... looking at the California high school exit exam is that not
one student who was at or below the 40th percentile was able to pass,
" said Brooks.

CA vs Maryland hard vs. fuzzy
standards based tests.

z42\clip\2000\07\sat9bad.txt
http://www.fresnobee.com/print/storynews/0,1737,167139.html,00.html
"Passing or Failing? The Stanford Achievement Test is alternately
lauded and panned by Valley educators as a way to evaluate student
achievement" Expecting all students to score above a nationally set
norm is ridiculous, Kuehn said.  "We call it the 'Lake Woebegone
effect' -- where all the children are above average," she said. "With
a norm, someone always has to be at the bottom."

NOBODY WANTS TO BID ON CALIFORNIA GRADUATION TEST!
http://www.latimes.com/news/state/20000108/t000002429.html
\clip\99\2000\01\fasttest.txt Saturday, January 8, 2000 By MARTHA
GROVES, Times Education Writer Spurned last month by test publishers,
state education officials have issued another plea. It is in the form
of a memorandum inviting publishers to submit "an informal bid" to
develop and implement the high-stakes test 

%%Stanford 9

ED PROFESSOR CLAIMS NO "GREAT LEAP" FOR LEP STUDENTS
He says everybody gained LEP or no, questions claim of 150% improvement
http://www.stanford.edu/%7Ehakuta/SAT9/
What legitimate inferences can be made from the 1999 release of SAT-9
scores with respect to the impact of Proposition 227 on the
performance of LEP students?
Kenji Hakuta Professor of Education, Stanford University (650)
725-7454 hakuta@leland.stanford.edu
Lots of tables showing state test scores and improvement by state,
district.
                                          LEP Students
State of California,           State of California,            State of California, Language
Reading                        Math
Grade  19981999 Change         Grade  19981999  Change         Grade 1998 1999  Change
    2   19  23       4             2   27   34      7             2   19    23      4
    3   14  18       4             3   25   32      7             3   19    24      5
    4   15  17       2             4   21   25      4             4   23    26      3
    5   14  16       2             5   21   24      3             5   21    23      2
    6   16  18       2             6   24   28      4             6   22    24      2
    7   12  14       2             7   22   24      2             7   19    21      2
    8   15  17       2             8   23   25      2             8   19    21      2

                                         All Students
State of California,           State of California,            State of California,
Reading                        Math                            Language
Grade  19981999 Change         Grade  19981999  Change         Grade 1998 1999  Change
    2   39  43       4             2   43   50      7             2   40    45      5
    3   36  40       4             3   42   49      7             3   39    44      5
    4   40  42       2             4   39   44      5             4   44    46      2
    5   40  41       1             5   41   45      4             5   44    46      2
    6   43  45       2             6   48   52      4             6   47    49      2
    7   41  43       2             7   45   47      2             7   49    51      2
    8   44  46       2             8   45   48      3             8   47    49      2

STANFORD 9 TESTS KIDS 2 YEARS ABOVE OVER GRADE LEVEL IN CA
6/99
Riverside Press Stanford 9 exam daunts youngsters Math is tested at
two years over grade level.  By Kamrhan Farwell Millions of
California children took a new math test this spring that even the
brightest students are expected to bomb.  The problem is, the new
killer quiz landed on students' desks before the lessons did.


@@STar
%%Star

Tough new math test based on rigorous rather than fuzzy standards,
but either way both test flunk most students relative to expecations
which are still not based on the curve and what actual students can
do vs.  what ideal situation would be. Tests based on algebra /
geometry rather than the usual middle school level math on CIM tests.

STAR is a creature of the California Board of Education (tradition),
established to the dismay of the California Department of Education
(progressive).

Star Test details
To download this file go to
http://www.cde.ca.gov/cilbranch/sca/star/parent/STAR99%20Parent%20Pkt.pdf

Cunningham:
ESSAY TESTS DON'T WORK
THERE IS NO GOOD WAY TO SET PERFORMANCE STANDARDS EVEN WITM M-CHOICE
STANDARDIZED TESTS NO GOOD BEYOND K-5
At the present time there is no accepted, correct way of setting
performance standards.  There aren't even any good proposed ways of doing
this.  The most carefully researched and expensively implemented are the
performance standards used by NAEP.  These have been savaged by almost
everybody.

AZ AND CA BOTH DUMPED PERFORMANCE TESTS BUT HAVE SCORE INFLATION
\doc\web\99\09\testinf.txt
"Arizona's situation was quite similar to California, they discarded
a performance assessment system and followed it with the Stanford 9,
pretty much in the same years as California."

* Ron Dietel: Research conducted by our center shows that after a new
test is introduced, scores consistently go up the second year and
also in the years to follow. If a new test or a new form of the same
test is introduced, the scores return to earlier low levels.  (See
for example, the 1998 CRESST Policy Brief by Robert Linn.) Therefore,
research supports Barbara's comments and what you have heard.

http://mathematicallycorrect.com/starbkg.htm
i:\clip\99\13\starbk.txt
CA MATH TESTS ALGEBRA AT GRADE 8, EASTIN COMPLAINS MATH TOO HARD
\doc\web\99\09\alg8.txt
State's Kids Do Better Than Expected in Tests of New Curriculum
Nanette Asimov Chronicle Staff Writer Friday July 2, 1999 ©1999 San
Francisco Chronicle URL:
http://www.sfgate.com/cgi-bin/article.cgi?file=/chronicle/archive/1999/07/02/MN28429.DTL
* students test for what they have not been taught
* need curriculum first, then test, not test first
* teacher and Eastin say math is outrageous, most adults would 
not get test
* arthur  - are they testing all 8th graders on algebra 2?

4 Grade 8: Mathematics Algebra 1 
Which equation is equivalent to 3(x - 7) - 2(x-1) = 8 .  A x - 23 = 8
B x - 19 = 8 C x - 8 = 8 D x - 6 = 8 

CA STUDENTS FALL SHORT ON NEW "STANDARDS BASED" MATH TESTS
For background information see
http://mathematicallycorrect.com/startest.htm
http://mathematicallycorrect.com/starbkg.htm

Mathematics (based on 15 Foundational Skills from Sta