Edublather extraordinaire from Hal Alford at the Verde Valley Online new site:
There is considerable controversy surrounding the federal "No Child Left Behind Act" because the law is based almost solely on competency testing...
While a single-test assessment system may be well intentioned, the practice seems to violate many fundamental psychometric, pedagogical, and ethical principles and standards of the education profession and may be harming rather than helping many children. The testing practice can be especially damaging to children of poverty who lack the supports and assets that are available to children elsewhere. It should be no surprise that most of the 276 schools identified as "underperforming" for the 2002 school year, were from schools located in economically distressed areas.
It is no surprise, because many of those schools are in fact "underperforming," and are shortchanging their students of an education. The whole point of NCLB was to refuse to allow loopholes for schools that don't use the money they have to educate children as well and as efficiently as possible. If the tests were removed, these schools would still be underperforming; we just wouldn't have objective evidence of it. These results are a validation of the testing program, not a criticism of it.
Characterizing schools as "underperforming," as the Arizona State Department of Education has done, publicly embarrasses and stigmatizes children and entire communities. Should a school under perform according to test results for a second consecutive year; they are classified as "failing." The law requires that "failing" schools take corrective steps, such as replacing the principal and allowing parents to send their children to other schools. This becomes costly and creates chaos for already beleaguered schools.
Oh, how sad. Much better that an inefficent principal remain in place. Much better that parents be forced to keep their children in bad schools, with no input other than their tax money. How dare they want the same choice that more financially-gifted parents have? If schools are so already-beleaguered, why should parents be forced to stay and support them?
Due to external pressures on schools to perform well on the tests, educators are reluctantly but invariably gearing their curricula to "teach to the test," a practice that undermines the teaching and learning process and negatively affects teacher and student morale.
I've said enough about "teaching to the test" before, so I won't go into all that again, but let me just point out that the test in question is AIMS, Arizona's Instrument for Measuring Success, and it measures only reading, writing, and arithmetic. Try as I might to go the distance with these testing critics (and I admit I don't try too hard), I just can't get worked up about teachers being forced to teach reading, writing, and arithmetic in an timely manner.
Are they doing so in Arizona? Doesn't yet seem like it. Here's a sample high-school level reading test. The readability of items on this form can range anywhere from fifth-grade on up, even though it's 10-, 11th-, and 12th-graders who take this exam. Note that the two sample items are a technical manual and a weather map. Then go to this document, scroll to page 7, and note that only 62% of 10th-graders, 42% of 11th-graders, and 32% of 12th-graders met expectations on this reading exam.
Granted, the 11th- and 12th-graders who took the exam were either new students or previous flunkers, but those results are still wretched. Some Arizona teachers have a much worse problem than morale; they're also incompetent.
...The tests are almost exclusively focused on math, reading and writing and ignore other important areas of learning such as art, science and social studies.
Which are all just oh-so-useful to kids who haven't mastered reading yet. It's admirable that Hal wants schools to focus on science, but when kids haven't mastered reading weather maps, I'd say they're not quite ready to start learning meteorology yet.
The tests fail to account for individual learning styles and developmental capacities, and they ignore state-of-the art research on multiple intelligences, including emotional intelligence. The emphasis on testing is certain to leave many children behind feeling discouraged and less confident about their ability to learn.
Gak, choke, urk.... *chugs Maalox* Ahhh. That's better. I really should be more careful about reading such piteously weeny edublather, at least while I still have these ulcers. My gag reflex is more sensitive than it used to be.
But seriously, while I hate to beat a dead horse, almost 40% of the 10th-graders flunked the basic reading comprehension part of the AIMS, and the math results are worse - MUCH worse. And Hal is worried about their emotional intelligence? Can these kids spell "emotional?" Can they count the number of letters in the word? Hal's emotional senses sure seems to be working overtime here in his concern for these kids. While some kids might indeed be "harmed" by tests, my guess is that those who haven't bought into the educrat theory that EI is more important than reading or math skills will come away relatively unscathed.
It worries me that Hal deliberately chooses to say that kids, rather than being discouraged about what they've learned, are instead going to discouraged about "their ability to learn." Please tell me he's not going where he's going with this...
The single-test assessment system discriminates against visually and linguistically limited learners who might otherwise be uniquely talented. History is full of examples of famous individuals who were said to have learning disabilities such as Winston Churchill, Walt Disney, and Albert Einstein. One can only wonder how they would have performed on standardized tests?
Yep, that's where he was going - the sacred space of learning disabilities.
All of you LD advocates who insist that Albert Einstein was learning disabled can just. Stop. It. Now. This claim, invented by LD advocates and repeated by gullible educrats, has been disproven by Einstein's biographers. Einstein was a quiet youngster, to be sure, but was a wonderful student from day one and was reading physics books by age 12. The only tests he ever flunked were ones he didn't bother to study for. He was not dyslexic. He would have been bored out of his gourd by the AIMS, I'm sure, but if he'd shown up for it he would have passed it.
What is it with LD advocates and these great history figures? All three of the great men mentioned above have been "diagnosed" with learning disabilities well after the fact, by those who have a political axe to grind. Was Winston Churchill learning disabled? Even the sketchiest biography mentions that he was shy and had a speech impediment, but notes that as soon as he discovered his passion (the military), he was a whiz-bang student. Does this mean we should conclude that all LD students are just bored, and goad them to work harder? Or should we follow Winston's example and send them all to military school?
The practice of wholesale versus individualized testing is also psychometrically unsound and it violates the ethical codes of professions whose members have been principally responsible for the evolution of the testing movement, such as the American Psychological Association, the American Counseling Association, and the National Council on Measurement in Education. The standards of these professions remind practitioners that assessment should be highly individualized and that educational and counseling decisions should never be based on single test scores alone.
Buried within this mound of edublather is this, the one valid criticism of testing in this article. I've written about this before. Note that the AERA qualifies its own claim, though, by saying that if only one test score is used for a high-stakes decision, there should be ample validity evidence, and the students should be allowed multiple chances to pass, and given remedial education if they fail. For example, all schools that use exit exams follow these rules, and they often provide alternatives (or loopholes) for students who don't pass the exit exam. What Hal wants you to believe is that the AERA's own statements forbid the use of standardized exams in high-stakes situations, which is not true.
Most importantly, the results of any test must be interpreted in light of many other factors including gender, age, race, ethnicity, disability, language acquisition, and socioeconomic status.
The document that Hal is allegedly getting this from is the Standards for Educational and Psychological Testing, and I've got my copy right here. The standards state quite clearly that:
* Overall passing rates need not be comparable across groups for a test to be considered fair
* Validity evidence should be presented for all subgroups
* Differential item functioning should be examined for all subgroups
* The test items should be examined for biased or offensive content
* Mean group differences should be investigated to ensure that the differences are not due to contruct-irrelevant skills
* Separate prediction lines should be used if the test is found to have differing predictive validity for different subgroups.
This is all from the section on "Fairness in Testing and Test Use." Note what Hal didn't mention in his "summary" of these standards? He doesn't mention that group mean differences are not in and of themselves unfair, and that if all validity and correlational evidence suggests it, it's perfectly okay to hold both Group A and Group B to the same passing score, even if the result is that 20% of Group A flunks while 80% of Group B does. No demographic information about the two groups is a priori of importance.
It's simply not true that the test results must be interpreted in light of gender, race, etc. Not if the test has been found to be unbiased. It shouldn't be assumed that the test is biased without empirical evidence, either.
[Devoted Reader Laura points out that students with organic disabilities who are unable to progress beyond a certain point, such as those with Down's Syndrome, should not be held to the same testing standards as other students. I agree, but I understand why NCLB demands that schools test these students. It's to discourage the schools from fudging their averages by placing non-disabled students into special education. There's a solution here that takes care of both problems, but I don't know what it is.
The Guidelines do note that test developers should make exceptions for disabled students, and this is the only group that is singled out for special treatment. But my comments above still hold for when Group A and Group B differ in a way that is not being directly measured by the educational assessment, such as race, sex, income level, etc. There's no reason to label a test as automatically unfair if fewer female students pass it, or fewer Hispanic students pass it, or fewer poor students pass it.]
Many empirical studies have shown that there are strong correlations between test scores and such factors as family income, education of parents, single-parent families, and school enrollment. Simply put, children who have two educated parents who are economically secure are much more likely to do better on tests than children who do not have the same assets.
And they're likely to do better in school; they're less likely to have run-ins with the law; they're more likely to have better health, and to go on to college, and to do well in life. This is a statement that supports marriage, not one that invalidates testing. Every measure of academic achievement correlates positively with family income and parental education; that's not a reason to believe those measures aren't assessing a valid construct. Grades correlate with income, too; does Hal believe we should abolish those?
This is the same old tired argument that, "It's just not fair that some kids have more advantages than others." Ironically, it's being used here as an argument against an act that is trying to ensure that kids from poor homes still get a good education in school.
Or, put another way, one can predict with astoundingly high probability that children from neighborhoods of poverty, many of whom are children of ethnic minority, will achieve poorer results on the tests than their affluent counterparts, irrespective of the quality of teaching.
But it's not irrespective of teaching, you see. These kids are more likely to go to bad schools, but if they come from poor homes or uneducated parents, they're more dependent on those schools for everything they learn. And those schools are less likely to hire good teachers and less likely to hang on to the ones they have. The good schools that don't buy this argument - the schools with outstanding achievement despite high levels of minority or lower-income enrollment - do so, in part, by relying heavily on testing. They don't tolerate poor test scores any more than they tolerate bad teaching.
It is for this reason that the National Association for Multicultural Education called for an end to the misuse of standardized and state mandated testing because it only creates more barriers to equal opportunity for large numbers of Americans.
Not true. The barriers are already there. A kid who can't read a weather map in 11th grade is at a disadvantage whether we test that or not. The tests are not the barriers. They're showing us which schools are the barriers. And when this many people are steamed about them, I know the tests are showing us something we should be looking at.
Assessment results should be interpreted only in relation to other behavioral data and the socioeconomic and cultural background of learners. They should take into consideration the diverse learning style and varying developmental capacities of all children. The best kind of assessment is ongoing, within each classroom, and geared toward teaching each child better in terms of his/her genuine needs.
Not true. This kind of claptrap is what allows educrats to believe that it's okay if a child of a certain "cultural background" can't read at grade level, or that "diverse learning styles" are what prevent well over half of Arizona's 8th-graders from passing the math portion of the AIMS. Good teachers instinctively tailor their approach for individual students, and to use tailored teaching as an excuse for letting kids flunk is an insult to their intelligence, and to ours.
You know, Hal is a retired principal. His school, Mingus Union High, seemed to do okay, despite a nauseatingly PC mission statement. This report card even brags about the high number of Flinn Foundation Scholars they've produced (which are chosen, in part, on SAT scores). But the school recieved only a "Maintaining Performance" rating this past year. Could that be the ultimate cause of this outpouring of seemingly-well-intentioned yet sadly-misinformed testing criticism?
Posted by kswygert at September 29, 2003 03:33 PM