Monday, January 26, 2009

The Perfect WASL Test Score Scam,

Review of Washington State WASL and Other Assessment Problems 2009

Assessment From Hell

I first became concerned about tests like the WASL when I saw the CLAS test being rolled out in California. It is my opinion that when you make a test more difficult, or "raise standards", you will tend to aggravate rather than close gaps between groups, and my perception (among others) is that the WASL was engineered to flunk most students on its introduction but pass them 10 or 15 years later to "prove" the efficacy of the prescribed reform program. It only appears that Bergeson's people were not able to successfully engineer the math test to be scored in this manner because of underlying massive fundamental math problems. I listened to WA legislative testimony of the woman who was in charge of CLAS who was informed that NO students were allowed to get the highest "4" math score, which explained why I could not find ANY students who had scored this high, even in the best examination schools.

When I first saw the 1997 sample math problems, I was alarmed upon comparing the problems to the published grade level EALRs. Judging which bag of marbles was more likely to have a white marble requires jr to high school level proportions, comparing fractions, and probability. There was a series of scales which was to used to order blocks, which would be a complex logic problem that I was not even taught how to solve up to my master's degree at MIT. On another problem, the 10th grade problem stated that similar triangles would be used, while a 7th grade problem asked about "similar slides", but there was a 4th grade problem using a flagpole and fire hydrant with shadows which not only made no mention of the similar triangles method, but the INCORRECT solution published in the Seattle PI did not even use similar triangles. By contrast, most of the problems on the 10th grade test I reviewed could be solved with 5th grade level decimal arithmetic. Almost none of the 4th grade problems can be solved with 4th grade arithemetic alone. Even OSPI studies concluded that many problems were beyond their stated grade level. Committees which set the "standard" never evaluated whether the problems were developmentally appropriate or compliant with grade-level learning standards in the first place.
I also perceived a problem with the state goal of 80% passing when in the initial year of 1997, only the very best school in Bellevue achieved such high pass rates. It makes no sense that it is easily achievable that every child can be taught by the system to achieve the same level of performance as the average of the top 5% of schools - if an when 90% of students "reach standard", in truth, the rank-ordering of students by income, race, parent education etc. is exactly the same as the start of the process.

I have also participated in scoring tests for one state writing project by the same company that scores the WASL. I observed officials who appeared to be moving the definition of scoring points to meet desired pass rates. We were instructed that NO paper was to be scored a "4", the equivalent of an A grade without special permission. Published results were far less than 1%, which means that students are more likely to attend Harvard or Stanford with a top 1% SAT score than achieve top writing score. I also observed that the worst high school papers in the bottom 10-15% were worse than what my 3rd grader (who wrote at a 5th grade level) was bringing home.

I've also done detailed demograhic analyses of Asians and other subgroups for the SAT and California STAR test showing such details as Asians being over-represented at both the top AND worst test score, and Filipinos scoring above the average overall, even if they lag the top Asian groups, and African American in Washington scoring relatively higher than most other states, even if they still lag other groups. I do believe there are deep structural differences between ethnic communities that cannot simply be legislated away, or made to disappear merely by redesigning an assessment system as one of many promises of WASL and ed reform. Neither can the "bell curve", which exists within as between all subgroupings of humanity can be easily dismissed as a political fiction. WASL scores also have a bell curve distribution, with fewer at the right and left tails, and most near the middle of performance levels.

Perfect Score Scam

I would like to call your attention to both the recent news articles released by OSPI of students in elementary school who received pefect WASL scores. Contrast that to my son's recent 10th grade writing test which was a PERFECT score. His PSAT scores put in somewhere in the top 10 percent of high school students. According to the OSPI, fewer than 1% of students receive a perfect score in any one test, let alone all 3. This would be consistent with scoring I saw for NC where fewer than 1% were allowed a top score of "4". Yet I asked OSPI how many 10th graders got a perfect score in writing, and the reply was over 18 percent. That is one in five. Nothing on the OSPI website provides enough information to establish the value of a perfect score such as a distribution.

This means that any "A" student probably got a perfect score as well as many "B" students. Scored sample papers show that anything higher than a "2" must have been a passing score, yet the worst papers were scarcely better and had similar errors to the best 4th grade papers. Bergeson claimed that 90% of students passed a high WASL standard. How could ANY standard passed by 90% of students be called high? What would happen to "high" standards if the University of Washington guaranteed admission to all but the lowest 10% of the class? If Washington students are anything like the papers I graded, the lowest 10% write at a 3rd or 4th grade level. That would mean that the WASL, which was justified because too many student were passing at a 4th grade level, has set a 10th grade "standard" that has been deliberately and deceptively at a 4th grade elementary school level. It would also mean that it would be equivalent or perhaps even harder for a person to pass the 4th grade WASL standard than the 10th grade standard.

Have you ever heard of a test with a curve scaled such 20% of students bump up against the right wall at the highest score while only 10 percent fail? And how can this be the case when at the elementary level, it has been scaled so that 20-40% fail and fewer than 1% get a perfect score? These are parts of the same system with identical grade-appropriate definitions of achievement? This is a scam that must be exposed so that Washington citizens know how badly the system was run.


While I had an opportunity to provide input on the new math standards, there has been essentially no response to my criticism of the problem solving standards which clearly ask students to solve problems that are many grade levels ahead. For example, kindergarteners are asked to compute how many fingers there are if 3 children who need gloves. Kindergarteners are not even taught formal addition, yet this would be a good introduction to multiplication which is not introduced until 2nd grade even at a conceptual level. It gets much worse at higher grade levels, and it is exactly the same problem as with the original 4th grade WASL test. There is a tangram problem that would probably be difficult for a college graduate like you to solve at the 4th or 5th grade level.

Once something is called "problem solving" then it becomes an excuse to expect mastery of a problem far before the grade level at which the formal solution method is introduced. Thus kindergarteners have to figure out how to solve a problem they will be explicitly taught in 3rd grade, and 4th graders figure out statistics problems that textbooks will cover in 8th grade, etc. This is problem which is a unique feature of "standards-based" fuzzy mathematics.

Higher Standards Fraud

"Higher standards" often merely takes problems that would be outrageously difficult for most adults and makes them requirements for lower grade level students. Another example from Marc Tucker's "New Standards" was proposing that 4th graders design a bike trailer with 3-view drawing and parts list down to the cotter pin. The high school example was a student building a hybrid car from scratch with a "motor donated by the power company" and "some assistance in welding the suspension frame" from his teacher in six months, in addition to a normal high school homework and activity load. Right. Would you like to guess how many man-months it would take for Honda to produce such a prototype from the drawing board to completion? Bellevue's Riley superindent made national fame for proposing requiring that all high school students take and pass AP level courses. Yet how can it make any sense to REQUIRE a COLLEGE level course just to get a high school diploma, and how does that help students who can barely muster elementary school level reading and math skills? We can have all students succeed only if we emphasize the grade-level basic skills that nearly all students CAN learn if it is taught and tested that they have learned it. Some students can and will achieve beyond grade level, but we cannot achieve parity merely by requiring that all student perform beyond was used to be a reasonable grade level. The purpose of a public education to simply provide AN education, a seat in a schoolhouse for a standard number of hours. The end goal of education should not be to require nor promise anybody's notion of "high" or "world class" standards.

The ESHB 1209 education deform act of 1993 that was passed more or less identically across many, many states including Oregon, Texas, and Massachusetts _IS_ the problem. It's not too many kids scoring too low. There is NO failing or pasing score on a true standardized test. 50% of students will always score below the median, which is grade level. The real problem is government acting like Asian Parents Who Go Crazy at B+ demanding everybody meet higher standards until they puke. What we do need is to get rid of the notion that everybody must get an A and anybody who is below average is a failure. We need to give out MORE not fewer honest D grades. We should teach our children (and parents) it is OK to get an honest D if it only means you are as far below average as an A is above, and C means average or lower.

Liberals say it's materialistic to pursue more stuff, yet if group A gets better scores than group B, then it's a crisis that can only be solved with a new math standards (which have been abandoned as stupid) a new certificate of mastery (which was abandoned as stupid), and a new WASL assesssment (which is about to be abandoned as stupid) If those are the three legs of the Outcome Based Education stool ( which was called "higher standards because OBE was abandoned as stupid) doesn't that mean the EDUCATION REFORM HAS FAILED BECAUSE IT WAS STUPID??

My kids don't need more class hours, more class days, more PE homework (When in the blazes did they need to assign PE homework??), more parent volunteer time, more service and graduation requirements, more math beyond calculus, more assessments. They don't have to score better than students in any other country or any other class or race. THEY DON'T NEED MORE ANYTHING. We just need a BASIC FREAKING EDUCATION. ARGH.

If you read this please post a comment if you think education has gone out of control.


Charlie said...

Thanks for your useful and insightful analysis. I'm a school teacher in California trying to get some reforms together and this is good ammunition.

BlArthurHu said...

I didn't mention that in my 4th grader school had a tremendous increase in WASL math test scores, but he showed me some practice tests that looked almost exactly like the real test that was later released. I complained, they said it was impossible, but that teacher was gone the next year.

Charlie said...

Here's my salvo, which links to you. To strengthen my argument, I'd really appreciate being able to post a little bit more info about you in my post so you're not just an anonymous source. Please take a look and let me know how I can better introduce your evidence--it's a work in progress:

Thanks again, and looking forward to hearing from you. Looking over your post a second time I'm finding more to look into.

BlArthurHu said...

I ran for WA state supt of instruction in 2000, just about tied for 2nd / 3rd place and they finally voted the incumbent Terry Bergeson out office in 2008 to kill the WASL. Columnist for Asian Week, and keeper of of diversity and education statistics.

Anonymous said...

I'm glad an ADULT is really paying attention here. If a student tries to complain or use a well-thought out argument, teachers merely roll their eyes and (metaphorically) cover their ears and scream, "LALALA I CAN'T HEAR YOU!!!"


Oh, by the way, I am a Liberal (see your third-to-last paragraph). Don't generalize, m'kay?

BlArthurHu said...

to "liberal", thanks for the comment. I just want to point out the disconnect when people say you should not desire to have lots of stuff, and then promise or require everybody to have lots of stuff. Hope we're in agreement here. It should not be the job of government to promise or require "success" of anybody.

Anonymous said...

Then we are in agreement.

Thank you for the clarification. The way people will often use a party name as an insult just really gets to me.