\doc\web\99\07\waslnews.txt To: weissbach@weissbach.com, BriMaloney@aol.com, ruthe@komotv.com CC: arthurhu@halcyon.com, Tips@komonews4.com, news@king5.com, kironews7@kiro-tv.com, doug.margeson@eastsidejournal.com, mlin-new@seatimes.com, jhou-new@seatimes.com, dlil-new@seatimes.com Subject: Blowing cover on the "secret" WASL test Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Status: More information. According to the 1997 sample booklet, "these questions, unlike the actual questions which are secured materials, may be reproduced". Clearly, a serious error has been made by including a sample question on the actual test. It clearly shows that making the test "secret" really only makes it easy for the CSL to get away with serious errors with no one noticing. I hold Terry Bergson personally responsible for this error, and for believing the incompetent "experts" who claim that 4th graders can solve a sorting problem that isn't even taught in college computer science courses. This is what happens when we let people run the tests who aren't even literate enough to pass their own 4th grade tests, or be able to compare tests against their own benchmarks without relying on committees or experts. You shouldn't have to be a 99th percentile MIT grad to figure this out. Indeed, this kind of stuff is what "every public official" should "know and be able to do". This sort of incompetence should be a firing offence, especially when it is inflicted on 24,000 of our children who will be declared "not proficient" by this insane test and the entire misguided "reform" movement. I still don't know why I'm the only citizen out of 5 million residents who realizes the problem with this test and is willing to tell the OSPI and CSL that there is a problem ----------------------- Hi, I was given the opportunity to look over the actual "high security" WASL test. that is being given to tens of thousands of our students. I could not take notes, but I saw enough. The real problem with releasing the test to the public isn't cheating , it's so that the public can see how ridiculous this test really is. I am a citizen and parent who is the leading opposition candidate for Superintendent of Public Instruction in 2000 to stop the insane education reform movement, and restoring high traditional standards that nearly all students can at least pass, but only the few are the best. Turns out it has some questions from the original sample booklet, I had posted the solutions to some of the problems at http://www.leconsulting.com/arthurhu/index/washtest.htm (main page) http://www.leconsulting.com/arthurhu/99/04/sampwasl.txt (sample test) I have reposted this solution, plus hints to some other problems on the usenet groups seattle.politics and wash.politics. along with a $25 challenge to anybody who can produce a b average 4th grader who can solve the problem. Normally, when the secrecy of a state test has been compromised by posting solutions on the internet, it's a huge story. So far, few people have shown interest in these solutions. Before I had tried to show everyone that this alleged 4th grade test was actually testing skills listed at the states own definition of 7th and 10th grade material. By looking at the actual test, the current 4th grade test is significantly simpler than the sample test, and probably simpler than the past 2 tests. Cognitively speaking, most the 4th grade problems are much more difficult than the 10th grade problems. Most 10th grade problems can be solve with at most one or two simple arithmetic operations. Most 4th grade problems involve significant thinking and construction. For example one asks 4th graders to explain how to find how many students like what without asking every student. The answer is a textbook page explaining how sampling and scaling works at a high school level, and that's where sampling shows up in the EALRs. The 7th grade test expects 7th grader to either know or figure out the formula for the volume of a can, but it's a 10th grade ealr benchmark, and the 10th grade test has a table which tells you the formula v=pi*r^2*h. Similarly, there is a sample 4th grade question which requires that the student convert from in to feet, this is specified as grade 7, and the 7th grade problem tells you there are 12 inches in a foot. I'd be willing to go on air and "spill the beans" on this allegedly high security test. It's highly unfair to use sample problems that some parents and teachers have drilled their kids on, especially when I've posted solutions to most of these problems. It's also amazing that no one has noticed that the indirect sorting problem that appears on the 4th grade problem goes significantly beyond any sorting algorithm taught in computer science classes, certainly beyond what any 4th grade math lesson teaches. For most students and adults, they will not be able to do this even after being shown the method of solution. This is the reason that 80% of normal students, 95% of minority students and even 30% of the best schools "failed" to meet the "standard". The public also has to know that the "standard" is not set until AFTER the test is given and scored, and the scoring "rubric" is not created until the test is given and all responses reviewed. At no point in time does the "standard setting" committee have the opportunity to find that a problem is not compliant with grade level to start with. The public must also know that no state, city, school or even small test group of average students has EVER been shown to meet the state goal of 80% passing. Every state with similar tests, and the national NAEP test show similarly high failure rates. Daytime phone 425-487-8034 eve 425-814-2183 Other people who can speak to reform issue Marda Kirkwood CURE: 253-631-1163 (education opposition activist) (Kentucky testing expert) gkcunn01@ulkyvm.louisville.edu He's happy to take calls too at home at 502 452-9764. 2114 Lakeside Drive Loiusville Kentucky 40205 you can see the posting on www.dejanews.com seattle.politics: SAY NO TO HIGHER STANDARDS DEMAND HIGH TRADITIONAL STANDARDS It's testing time again. The WASL will be given to tens of thousands of our students next week. It will be used as the measuring stick of proficiency. If they don't pass the WASL, they will not get their "Certificate of Mastery". Anyone who does not have a CIM, ideally, will not be able to get a job or continue their education. Any fourth grader who is not "proficient" ideally should be held back until they can demonstrate what all students are "expected to know and be able to do" if education reform gets it way. Reformers are in charge of the "standards", not you. The first year 4th and 7th grade results were that 80% of students failed (fell below "meeting the standard") The MEDIAN where 50% fell below was a "1" meaning demonstrates essentially NO knowledge. 95% of blacks and Hispanics fell below the "standard" that proclaims that "all students can succeed". A majority of minorities demonstrated "little or no knowledge". All other standardized tests put WA students equal or above national average, and in the top 20% of international nations in grade 4. The test is designed so that even a student who demonstrates 99th percentile proficiency on a traditional test of addition, subtraction, muliplication, addition with decimals, fractions, and very simple statistics can fail the new test. The math EALRS do not require memorizing the multiplication at any grade. They DO require exposure to women and minority mathematicians. The Rand Corporation studied similar tests in the Kentucky KIRIS and concluded that gains such as last years WASL where EVERY high school in the state showed a 50% gain in students passing the standard is due to a built in inflation effect caused by easier questions and teachers teaching previous years questions which were repeated the next year. No state, city, or even small test group has demonstrated the goal of 80% passing with any representative cross section of students on any of these performance -based tests. Only 2% of high schools passed Virgia's goal of 70% passing. NONE of WA state schools, not even the top 1% met Washington's goal of 80% passing which Terry Bergeson assures us is a realistic goal in only 4 years. There is zero organized opposition to the WASL, and the press has refused to print any criticism of it, unlike the Boston Globe which has printed dozens of experts who claim the MA test is inapproprate for the grade level. The WASL and 1209-based education reform movement is backed by democrats, republicans, industry, education unions, school districts, and just about everyone else who makes a living in the education industry or controlling it. Millions of dollars are backing this reform effort. I have shown these problem to the CSL and OSPI, the response is that as long as the "process" has approved these problems, they are appropriate for the grade level, there is no appeal process. Other states use nearly identical methods to establish the "standard". The "standard" and even correct solutions and scoring guide are not created until AFTER the first test is given. All 30,000 repsponses are categorized to create a scoring "rubric" for each problem. After the tests are scored, a committee is picked to determine the number of problems that are needed to set the "standard". At no point in this process does the standard setting committee have the opportunity to find out that a problem may not be compliant with the benchmarks themselves. For example, the sample test includes ratio, proportionality, indirect measurement area=width x height, independent probability, unit conversion, all tasks which are specified at grades 7 or grade 10, these have all been approved by the standards commmittees. The standard is set with no knowledge of what percentage actual students got correct. While standardized tests are carefully calibrated so that only a few miss the easiest, and only a few get the hardest problems, if a performance based test results in a 80% flunk rate, it is considered to be a valid pilot. If the test does not match content, it is the responsibility of schools to "re-align" their curriculum to the test. Previously, a test which tested content which was not taught was considered to have no "instructional validity". This test is designed to show improvement from year to year, and from grade level to grade level as the 7th and 10th grade tests are progressively easier. Cognitive tasks in the 4th grade test are much more difficult that simple arithmetic which dominates 10th grade questions. The scale problem below would stump most CS grads who interview for microsoft. Is it possible to stop this movement? Can the ant move a rubber tree plant? Can a bug stop a truck by splatting on the windshield?? Probably not, but I'm sure as heck going to try. This is a repost taken from the sample booklet anyone can get from the Commission on Student Learning in Olympia, with my solution. Sample problem may be used on future tests. CONTEST RULES 1. $25 to first 4th grader with equivalent of B or worse math grade average who hand writes a correct solution to the scale problem and sends it to me. Students with "A" or top math grades are NOT elibible. A random drawing if they arrive the same day, contact me by email and mail or fax solution to me. Solution must be with no assistance by parent. My contention is that NO average 4th grader is capable of solving this problem. 2. $25 to first person who codes a program in C, Visual Basic or Java or any other language that can solve this general balance problem. No age or ability limit. eve phone 425-814-2183 day phone 425-487-8034 Arthur Hu ARE YOU UP TO THE CHALLENGE? Now the review problems... WASL 4TH GRADE SORTING PROBLEMS FROM HELL \doc\web\99\02\block.txt How many of your 4th graders can solve this sorting problem? Do you want your 4th grader held back because he doesn't meet the "standard" of what "he should know and be able to do"? Do you want your kid's diploma denied because you couldn't pass an equally hard 10th grade test? This is exactly what states like Texas and New York are threatening to do, we could be next. Can you solve it? How long did it take? Is this the sort of problem you want on a test where you have to finish 30 other problems in one day? Do you want your kids wasting an entire week trying to solve problems like this when a standardized test only takes a couple of hours on one day, and costs only $2 not $40 to score? See the math benchmarks for Washington State's EALR at http://cisl.ospi.wednet.edu/ComSL/MATHBMK.html BTW, Terry Bergeson claims somebody HAS done a 2nd review of problems and pronounce them fully compliant and reasonable within the goal of 80% passing once everybody starts teaching to the new EALRs. Do YOU agree? Sorting problems: problem 17 page 70 "Scales and Blocks" (sorting) The pictures below show the results when some boxes are placed on a balance: 1---- T J 2---- L X 3---- R W 4---- J W 5---- R L 6---- X T 7---- T W --- Name two boxes that are heavier than Box X. Write the number of the picture or pictures you used to find each of your answers _______________________________ Tell whether box R is lighter or heavier than Box X. Write the number of the picture or pictures you used to find your answer _________________________________ List the 6 boxes in order from heaviest to lightest: 1. 2. 3. 4. 5. 6. Comment - there is nothing in the benchmarks that specifies being able to sort 6 boxes based on 7 inequalities, except "problem solving". There are likely no textbooks that give a solution to this problem. Most adults can't do this problem, I doubt more than a few percent of 4th graders can do this, even after being shown a solution algorithm. This problem is not even appropriate for most 12th graders unless that have taken an algorithms course that includes format sorting algorithms. 4th Benchmark: identify, compare, and order whole numbers and simple fractions. Nowhere is sorting of a list of numbers or items mentioned, even at grades 7 or 10 This problem is not just ordering two numbers, it is sorting a list of unknown values, and asking for a solution as to their order. A benchmark to match this problem would be to "sort a list of six items given 7 inequalities between different pairs of items". Grade level appropriate problem Block A 100 lb Block B 75 lb Block C 200 lb Block D 115 lb Arrange these blocks in order of increasing weight. Appropriate Grade level benchmark: "Sort a list of 4 3 digit integers". One Solution, based on insertion sorting algorithm 1. T, J. 6 shows T is heavier than X, 1 shows J is heavier than T, which is also heavier than X. 2. based on following sort order, R is lighter than X 3. from heavy to light: W J T X L R reverse this result: RLXTJW Letters to sort JTLWRX-6 blocks total # is sorting list * write heavier from left to right * add a letter to left or right if it is definitely to the left or right of what we already have. * try out each letter in sequence * stop when we have all 6 letters Look for J TJ sequence #TJ JW, put W at right end sequence #TJW done J!T!LW!RX Look for T XT put X at left end sequence #XTJW done J!T!LW!RX! Look for W RW? R would be in the middle, wait Look for R RW RL Still in the middle Look for L, lighter than lightest LX Put L at left sequence #LXTJW done J!T!L!W!RX! Next L RL Put R at left again sequence #RLXTJW done J!T!L!W!R!X! All letters complete, stop Problem 35 p.90 In 1991, the five states below had the greatest number of people visit their state parks and recreation areas. The states are listed in alphabetical order Visitors (in thousands) ------------------------ California 70,444 New York 60,744 Ohio 67,222 Oregon 39,479 Washington 46,813 How does washington rank in terms of total number of visitors? A. third largest B. fourth largest C. fifth largest Solution 1. Count how many states are greater (3), then add one for washington's position, answer is fourth. 2. Sort the list, find washington on the list (4 from top) This involves rewriting the list, or writing numbers beside each line without getting confused. Note that sorting a list is NOT a 4th grade required skill. This problem can figured out by many fourth graders, but it is still much more complex than simply "ordering whole numbers". The answer is not directly taught, but must be "figured out", which means that you cannot expect all children to master this problem unless they are either "smart" or have been coached on these particular kinds of problems. Whenever you test for the ability to solve a problem for which there is no known or taught solution, you are testing for cognitive ability, or IQ, which cannot be taught or assessed directly, and the advantage goes to kids with highest IQ and out-of-school resources and exposure. --------------------------------------------------------- Final comment - please note that any compliant question should be simple to answer by any adult with a 6th grade education. That is NOT true of the current 4th grade test, which would be difficult for any 4th grade teacher, let alone student or parent. This should not be. Standards should be based on what actual kids and teachers have been shown to be capable of, not what some consensus committee believes it should be set to 10 years from now.