Toward the end of last semester, a colleague of mine who teaches a course on special education had her students participate in a poster session, and I was invited to attend. The sessions covered the usual topics you would expect to see covered in a special education class, focusing on various disorders and (dis)-abilities, and the students, as they always do, represented themselves well.
One student, in particular, made a point that resonated with me. She was talking about ADHD and the medication of students who had been diagnosed with it. Her point was a simple one: doctors had some idea of the effects of medicating for ADHD with drugs like Adderal as soon as they started prescribing it, but some things can only be seen over time. And when you add in the variability of context (in other words: when you consider ADHD being diagnosed and medicated in the age of No Child Left Behind and the endless battery of tests it has authorized), it’s really hard to predict the effects of medication. Not even medication can be prescribed in a vacuum; its effects are often magnified, or diminished, by other factors.
This got me to thinking. I don’t want to get too carried away with this, but the thought I had was that there’s a pretty strong parallel between the point that student was making and a point I have tried to make more than once about testing and tracking in schools. If you live or work in a school district that places kids in classes primarily on the basis of test scores—or solely based on test scores, which appears to be the case where I live—then you have a drug problem on your hands. And the drug is “data.”
I’m not opposed to using information to make important instructional decisions. Teachers and school administrators should consider a range of factors when they try to connect kids to the curriculum. If tests can be developed that are valid and reliable, that connect in meaningful and clear ways to what was taught (a really hard thing to do, I might add, if the tests are standardized and are created far away from the places where the teaching actually happens), then the data they produce can actually be very helpful to teachers. Imagine a world where we give tests that accurately assess whether students have met established benchmarks and where, when the tests show that students haven’t, we know the exact “interventions” to use to make sure they learn. Sounds great, right?
That’s the theory of No Child Left Behind, in a nutshell: we need to be sure that the money we spend on schools is well spent, and the best way to do that is to make sure kids are learning. But how can we know that? Well, we’ll give a test. (Or a hundred tests, as it turns out.) This way we will have true accountability: students will be held accountable for learning by the high stakes nature of the tests (what’s the point of giving tests if there are no consequences for failing to pass them?), and so will teachers—we’ll even be able to tell who the good ones and bad ones are by seeing how their students do on the tests. We’ll also use the data generated by the tests to target the schools and students who need the most help. And, to prove our toughness, we’ll drum the worst teachers out of the profession and shut down the worst schools. Because we want our kids to be successful, damn it!
The results have been predictable—and not in a good way. Students are stressed, as are teachers; the high stakes we tied to the tests have caused teachers to teach to the tests, only amplifying that pressure. No one seems to be happy with what we have now. Even Arne Duncan, once as ardent a supporter of high stakes testing as you would find anywhere, has said that “testing issues...are sucking the oxygen out of the room in a lot of schools,” which may be the most lukewarm way of announcing a moratorium on testing possible. See what he did there? It’s not testing that’s sucking the oxygen out of the room; it’s testing issues—constant bickering about testing‐that we can blame for that. His hands were tied. And it’s only happening in a lot of schools—certainly not all, or even a majority, of them.
Baloney. Duncan’s own statement belies his addiction: like so many others who don’t do the actual work of teaching kids in this environment, Duncan clearly still clings to the idea that standardized tests can tell us what we want to know. I’m not buying it. I also wasn’t buying it when the former principal at my local middle school informed me, in measured tones, that data held the promise of helping teachers deliver “appropriate interventions” to students who needed them most. This was the same guy who came to the conclusion, solely on the basis of test scores, that enrolling my son in “academic” classes would “do him harm” because “he would be the lowest ability student in the class, and by a large margin.” (At our insistence, our son is in those classes now and doing just fine with a B average—not to mention the profound positive impact the change has had on his interest in school and his self esteem.) Again: it might be reasonable to come to that conclusion if the tests he took actually told us that. But they don’t. Obviously, they don’t. You wouldn’t believe how hard we had to fight to overcome the assumption that they do.
I think Arne Duncan, and a whole lot of other policymakers, may have a substance abuse problem—and that substance they are abusing is test-generated data about student performance in school, data that then is also used to try to determine how well students were taught even when the tests from which the data was derived were not designed to assess such a thing in the first place. Here’s how the American Psychological Association described substance abuse in the fourth edition of its Diagnostic & Statistical Manual of Mental Disorders, or DSM-IV:
A pattern of substance use leading to significant impairment or distress, as manifested by one or more of the following in the past 12 month period: Failure to fulfill major role obligations at work, school, home such as repeated absences or poor work performance related to substance use; substance-related absences, suspensions, or expulsions from school; neglect of children or household Frequent use of substances in situation in which it is physically hazardous (e.g., driving an automobile or operating a machine when impaired by substance use) Frequent legal problems (e.g. arrests, disorderly conduct) for substance abuse Continued use despite having persistent or recurrent social or interpersonal problems (e.g., arguments with spouse about consequences of intoxication, physical fights)
If the major role obligation of policymakers and school administrators is to educate children—not test their intellectual curiosity right out of them, a sign of poor work performance—then it seems safe to say many have a substance abuse problem. If the overuse of tests causes so much psychological damage that it manifests in physical harm—I’d be lying if I said my wife and I did not feel physically ill as we tangled with our son’s school district over several years about test scores—then we may have a substance abuse problem in our community. If teachers and administrators are feeling the need to cheat in order to help students pass tests—and this, of course, has been well documented—then they may have a substance abuse problem. And we all live in a world now where tests are being administered over the protests of so many and in spite of the negative consequences they engender, another sure sign of substance abuse, and one, quite frankly, that’s harming the relationship of many schools to the communities they serve. As another colleague of mine used to like to say, “The beatings will continue until morale improves.” That should be the motto of data-driven school reformers in the age of No Child Left Behind.
Tests—or, maybe more accurately, “data"—are the drug, and it’s high time for an intervention. Maybe I’m wrong, but the evidence would seem to suggest otherwise. You be the judge.