E:\DOC\WEB\99\01\neapset.txt The entire basis of "Standards Based Education" is that you can set objective "standards" that are indepdent of percentile scores of what actual kids can do. Gerald Bracey cites evidence that it does not work for NAEP, it certainly didn't make any sense on the Virgina test, and some day soon, hopefully, all of us people who know about tests will enlighten the rest of the population so that such a monstrosity is never, ever inflicted on the population again. YOU CANNOT HAVE A COMMITEE, ESPECIALLY A POLITICALLY MOTIVATED COMMITTEE SET AN OBJECTIVE PASSING STANDARD ON CONTENT FOR WHICH THERE IS NO OBJECTIVE DEFINITION OF MASTERY, LIKE ACADEMIC ACHIEVEMENT. PERIOD. ------- Forwarded Message Follows ------- Send reply to: "Gerald W. Bracey" From: "Gerald W. Bracey" To: Subject: Re: [wa-ed-deform] Bracey - VA "failure" is higher than Iowa or most nations Date sent: Tue, 26 Jan 1999 14:06:15 -0800 When you say that NAEP's claim of only 20% proficient means that you have to be at the 80th percentile, you misinterpret NAEP proficiency levels. These levels, basic, proficient and advanced, were established when Checker Finn was head of hte National Assessment Governing Board. Checker wanted to sustain the sense of crisis instilled by " A Nation At Risk," so he got the board to set them very high. In fact, they were set so high that even the ideologues realized they had to back off. NAGB also hired a team of expert evaluators to evaluate the standard setting process. The evaluation team said it didn't work. NAGB fired them. At least, they tried to, but something in the contract prohibited it. In any case, the proficiency levels are criterion-referenced, not norm-referenced. That is, referenced to what the judges thought kids should be able to do, not referenced to the performance of other people. They are still ridiculously high. This was pointed out this spring by Lye Jones, a psychometrician at North Carolina, Chapel Hill. He noted that in the 1996 NAEP science assessment, only 18% of 4th graders were proficient and only 2% were advanced. But, said Lyle, these are the same 4th graders who finished third in the world among 26 nations in TIMSS. Does it make sense to say so few are doing well? He didn't answer his question. Didn't have to. -----Original Message----- From: arthur hu To: gbracey@erols.com Cc: wa-ed-deform@egroups.com ; fair-diversity@egroups.com ; education-consumers@ripple.dundee.net Date: Monday, January 25, 1999 3:24 PM Subject: [wa-ed-deform] Bracey - VA "failure" is higher than Iowa or most nations (Gerry, you obviously don't get it. They're setting HIGHER standards, they don't care if their goal happens to be higher than Korea or whereever, by gosh, if their committee wants to shoot for the sky, why the heck not? After all, they've got 8 years to get there, right?) (Oh yeah, the individual student average were actually much higher than state like WA which flunked out 80% of all, and 95% of blacks, not 45% of 4th graders in math) I did not get the attachment, but thanks for pointing this out. WA students at 4th grade are even with Iowa, yet had 80% flunk rate. The kicker is that the standards based movement says these tests are all benchmarked to international standards, yet they clearly flunk out students who are performing way above average levels of the best nations on the world. That's what happens when you have a committe set the pass rate, and they don't care since they have eight years to meet the goals. According to the RAND study of the KIRIS test, the goal required the average school to improve by 2 standard deviations, when the difference between the US and Japan was not even a standard deviation. You really need to expose the lie that these tests are benchmarked to international anything. Asian and European math courses are fully traditional, and they do _not_ include the typically above grade level skills found on most reform-based tests. If anything, you saw in the Korean document that foreign nations desire to copy AMERICAN reforms, they are not expecting the US to adopt their standards! It's wild to see the US justifying "world class standards" as the reason to reform to compete with Asia and Europe when they're all saying exactly the same thing "we must copy the Americans if we are to compete with them". Since the NAEP tests a cross-section of students, when they say only 20% are proficient, they mean you must be 80th percentile to pass. This is all based on the naiive notion that all we have to do is toss the bell curve and set a standard that all can meet. Except the new standards are even tougher than than the old ones that poor and minority students couldn't meet either. It's all wishful thinking, just like the "goals and timetables" of affirmative action. Wouldn't it be more practical to forget entirely about setting goals independent of actual student ability, and just go back to a curve with 50th percentile defining "expected grade level performance"? Send reply to: "Gerald W. Bracey" From: "Gerald W. Bracey" To: Subject: Re: [wa-ed-deform] Massive fail rate makes some question credibility of VA tests Date sent: Mon, 25 Jan 1999 17:38:54 -0800 > OK guys, here's the deal on Virginia's "massive failure." > > 1. In TIMSS, Iowa scored higher than 35 of 41 countries in math, 40 of 41 countries in science. > > 2. Iowa students score between the 62nd and 68th percentile on domestic standardized tests (depending on age and subject). > > 3. Students in Northern Virginia schools that FAILED the new tests score between the 75th and 80th percentile. > > I attach a letter that was sent to the Washington Post the day the data were released. The Post might yet publish it. One version appeared in Northern Virginia newspapers. > > > -----Original Message----- > From: arthur hu > To: education-consumers@ripple.dundee.net ; wa-ed-deform@egroups.com > Date: Monday, January 25, 1999 1:51 PM > Subject: [wa-ed-deform] Massive fail rate makes some question credibility of VA tests > > > I know a lot of you out there still think a test that's so hard that > everybody flunks is a good thing, but look at this. These tests might > be a tool for the moderate left or moderate right, but it's no good > for the freedom right, or the take it easy on the poor left. > > > http://www.edweek.org/ew/current/19va.h18 > January 20, 1999 Education Week > Massive Failure Rates on New Tests Daze Va. By Jessica Portner > > arthur hu kirkland WA arthurhu@halcyon.com "fairness in diversity" http://www.leconsulting.com/arthurhu -------------------------------------------------------------------------------- eGroup home: http://www.eGroups.com/list/wa-ed-deform Free Web-based e-mail groups by www.eGroups.com