WSU ED PROF IN KAPPAN: 4TH GRADE WASL MATH IS TOO DARNED HARD zip38\clip\99\19\testhard.txt link Oct 10, 1999 State's 4th-grade math test too difficult for students' development level, critics say by Linda Shaw Seattle Times staff reporter See .gif of lemonade problem http://www.leconsulting.com/arthurhu/99/15/lemon.gif Arthur Hu Comments: It's finally about !@#$% time somebody else sat down and took a good hard look at the ridiculous sample problems they've posted since I've been trying to tell Olympia and the press since 1997 that the test is a test of 7th, 10th and beyond grade skills. WASL IS _THE_ SPI 2000 ISSUE If voters believe Terry Bergeson, she will be re-elected, and voters will continue to support "higher standards" based reform. If voters are outraged that WASL is an unexcusable multimillion dollar boondoggle and scam that plays God with our kids minds and future, she will be run out of office. Which will it be? This article does not mention: * Even this ed professor did not compare the skills with benchmarks which clearly state that indirect measurement with forumulas, comparing fractions with unlike denominators, and ratios are clearly grade 7 on the state's own definition of developmental benchmarks. Washington in fact is the only state with skill benchmarks which are specific enough to disqualify nearly half of the alleged 4th grade sample questions. Direct answers to many sample "4th grade" problems can be easily found in junior high math or high school algebra or statistics textbooks. It is not a matter of "higher order thinking" but being forced to figure out skills which will be directly taught in higher grade levels. It is not up to a "teacher committee" to decide what has been historically been taught at each grade level, and what actual kids are capable of demonstrating. Link: http://www.k12.wa.us/reform/ealr/standards/math.html MORE DIFFICULT THAN THE COLLEGE SAT * WASL versions of 4th grade problems found the College level SAT are actually far MORE difficult than the SAT versions of, for example, sorting and place value problems. FLUNKS OUT BLACKS AND HISPANICS WORSE THAN THE CTBS * The 1997 test flunked 95% of Blacks and Hispanics, with a race gap equal in size to the standard deviation found in "biased" IQ and SAT tests rather than the smaller gap found in standardized tests such as the CTBS. FLUNKS OUT ABOVE-AVERAGE ASIANS AND HISPANICS * The test penalizes some largely Asian and Hispanic schools in Beacon Hill and Grandview near Yakima that scored above the 60th percentile above average on the CTBS, but far below state average on the WASL. Predominantly black schools with merely below average 25th percentile CTBS scores such as Thurgood Marshall had 95-100% flunk rates on the WASL. The WASL is one of the few known tests where the Asian math average is BELOW the white average score. BLACKS EQUAL TO WHITES IN 5 YEARS, YET JUST AS FAR BEHIND! * If gains are to be believed, state Black students in 2001 will at the same rate as Whites as recently as 1997. Yet nearly one standard deviation continues to separate the races, and predominantly black schools such as African American Acacemy and Thurgood Marshall continue to populate the bottom of the list with 0 to 95 percent failure rates. RAND FINDS HUGE GAINS IN SIMILAR KIRIS "SUSPECT" * Rand Corporation has stated that gains of a standard deviation over the space of 5 years are "highly suspect" if not certain proof of the invalidity of the claim that rising scores are proof of rising skills. See link: http://www.rand.org/publications/MR/MR1014/ SCAM: SECRETLY LOWERING THE BAR Rather, the steady 10 point annual rise, whether by accident or deliberate manipulation, effectively lowers the bar by 10 points per year to meet the quota set of 80 percent passing in 10 years. At the end of 10 years, the test will no longer be as difficult as the top 20 percent, but reflect what all but the lowest 20 percent can do. The state has provided absolutely no documentation or other tests to confirm that the tests are equally hard and reflect substantial gains in skills. ONLY AN EASY TEST WILL RESULT IN 80% PASS RATE LIKE TAAS Despite the claim of having been set to "world class standards", no nation, state, city, or even school has been able to demonstrate 80% passing rates with a test this difficult. Texas is one of the few states with 80 percent pass rates, but their TAAS actually does conform largely to Washington's unambitious written EALR grade level standards, in contrast to the assessment which appears to have been written to G7, G10, or higher level benchmarks. 4TH GRADE TEST MORE DIFFICULT THAN G10, WHICH HAS NO HS MATH IN IT! * Looking at the actual test, the 4th grade is the most difficult. Some problems exceed high school level mathematics in complexity, such as explaining in detail how to use sampling to estimate numbers in an entire population. The 7th grade test is 8 to 9th grade level. The 10th grade test is mostly middle school level arithmetic easily solved with a few punches on a 4 function calculator, there is NO high school level math (algebra, triginometry, advanced statistics) on the 10th grade level test. PERFORMANCE BASED TESTS, OUTCOME BASED EDUCATION FLAWED FROM THE CORE * Performance based tests in general cost 10 times as much to score, measure fewer tasks, are inaccurate and unreliable to scores, tend to be invalid measure of curriculum, and suffer from a wider racial gap than standardized tests, in addition to deliberately omitting percentile score rankings which can be compared with other states or nations. Performance based tests are based on the troubled philosiphy of Outcome Based Education and Constructivism which has had a record of failure throughout the 1980s and 1990s across the nation. HOW CAN YOU SET PASSING STANDARD, CORRECT SOLUTIONS _AFTER_ THE TEST?? * Rubrics for determining what will be judged as a "correct" answer is not established until AFTER the test has been given and all 60,000 responses surveyed. The passing score is not established until after all tests have been scored by a committee that is not even permitted to compare problems with grade level benchmarks for compliance. The tests are "scientifically" constructed to be of equal difficulty across all years even though Terry Bergeson has stated that some writing tasks were substantially more difficult in some years than others. Hopefully this will mark the beginning of inevitable standards based train wreck that will bring down Terry Bergeson in the 2000 election, whether to me or anybody else. --- article starts here ------------------------------------------- {NOTE: In accordance with Title 17 U.S.C. section 107, this material is distributed without profit or payment for non-profit research and educational purposes only.) A longtime education professor at Washington State University recently made an item-by-item review of the state's new test for fourth-graders. Donald Orlich concluded the reading, writing and listening sections of the sample tests he studied were tough but reasonable. But the math test, he concluded, was unreasonable, and unfair. Sample One Two Three Lemonade power XX XXXX X -> Lemonade Water X XXX XX Which makes the strongest lemonade? You state that this is a ratio problem. Well, guess what? Ratio does not appear in most math textbooks, and on the WA state standards, NAEP standards, and draft specification until after the 6th or 7th grades. Yes, one could use "reasoning" to solve this with no knowledge of ratio, but any one who used straight forward math would look at the 3 pictures and construct the 3 ratios to compare: 2/1 4/3 1/2 To do a formal comparison requires converting to a common denominator, in this cases 1/6 12/6 8/6 3/6 which makes the first the largest ratio. Yes you could informally compare the fractions by drawing picture of 4/3, and reducing 2/1 to 2, but even that's beyond what covered in 4th grade which is division, and comparing fractions with common denominators. What's more, comparing fractions with unlike denominators is also a task which should not be assessed until well after the 4th grade. Yes, perhaps the top 10 percent of kids will have the IQ or the background from their educated affluent parents to figure out a ratio problem with comparing fractions, but this is a test of what ALL students should know and be able to do, and it is a test of 4th grade EALRs, not the upper grades. MikeHihn@WALiberty.org/CC >The article says more about a fairly dumb retired EDUCATION >professor than about 4th graders. >For him (the prof) it probably is too hard -- because he went about >solving the example question in a way that is barely literate. A >correct solution requires no math ability at all -- other than the >ability to count to .... two. >The question uses pictures, which curiously are not included with >the online edition, and are difficult to describe verbally. But >I'll try, for those who do not receive the Times. A GIF is >attached, but may not transfer to all the lists shown here. >There is a series of three pictures, each with a different version >of ingredients for making lemonade. The questions is -- Which >recipe will create the strongest lemonade? >The professor claims only 2% of kids nationwide can solve the >question, because it includes three variables, each of which must >be converted into a fraction. Is that true? >The first picture is one container of water, and two smaller >containers of lemonade. That one's easy, virtually self-evident, >and sets the standard for the rest -- one "pair" of mix containers >per container of water. Duh. >Plus, the sole water container is visually connected to one "pair" >of lemonade mixes. >The second picture is three glasses of water, but only two of the >waters are matched with a "pair" of lemonade containers, thus >clearly weaker on visual evidence alone (one of the waters has no >matching pair of mixes). >Third picture is a reverse ratio of the first -- two waters and >only one mix. >Thus, I submit that if the kid -- a fourth grader -- has ever made >lemonade (or Kool-Aid), the answer can be derived from the pictures >alone, with no math skills at all. >If you did not receive the attached GIF, then you cannot answer my >own question. Is this really too tough for even the SAT's -- as >claimed by Arthur Hu? And if so, does one need a high SAT score to >fathom recipes in (say) the Betty Crocker cookbook? >Oh yeah, the perfesser is a retired EDUCATION perfesser, not a math >perfesser. Nuff said?>-Mike Hihn http://WALiberty.org >WA Liberty PAC & Grassroots Liberty Foundation