For a moment, forget the names. Let’s use standardized test, since that is a phrase the world is reasonably familiar with. Standardized tests are tests built in partnership with very talented statisticians as comparative instruments. That allows the scores they produce to be used to compare different students to each other as well as compare a student’s score this year to last year. Or in the instance of this latest idea about testing, a score at the beginning of the year to a score at the middle of the year to a score at the end of the year.
Make no mistake, those advocating for multiple tests across a single year are advocating for multiple standardized testing sessions, no matter what words their marketing departments come up with to convince us otherwise.
Standardized tests are a form of research that enables comparisons, but that’s it. That is the limit of what they do. They cannot comment upon the far more important questions, which have to do with the causes behind that score. Why did a student score where they did relative to their peers? Why did a student’s position change over time? That sort of thing.
We can’t know the causes unless we go look, which is oddly enough a step we rarely take in today’s world. One thing we do know about those causes, however, is that they occurred in at least two places: schools and everywhere else. We also know that what happens outside school has a much larger effect on where a student scores than most people realize, which is why judging schools from such an instrument is just plain dumb. Test scores can’t separate out where something was learned, so assuming it was all learned in school is always going to be wrong. That’s why judgments of a school based on a test score are invalid.
But what about a student? If you accept that using test scores to judge schools is a dumb thing to do—and you should—should we, as those hyping the new myth advocate, instead use them to inform learning, adjust the curriculum, and individualize instruction? After all, those are good things and we should do them, and if these more frequent tests can help do that, shouldn’t we take them up on their offer of trying to help?
The answer there is a resounding no. No way. No how. Forget it.
The reason? For statisticians to create tests that can be used for comparative purposes the only items that can be allowed on such tests are selected via a very narrow set of statistical criteria. Otherwise, comparisons aren’t possible, or at the very least, wouldn’t mean anything, like apples to basketballs. But what that does is eliminate the opportunity for assumptions to be made about the specifics of what was learned, or what should change in the curriculum, or how to individualize instruction. It means that the content of an individual items is no longer available for interpretation because the content was not the final factor that caused it to be included in the test. It was included because that piece of content in that form produced a highly predictable result when administered to large numbers of students.
The fact that such items cannot inform the specifics of what was taught and learned isn’t an accident or an oversight. You can have one of these statistically based standardized tests that enables comparisons, or you can have a test that is all about teaching and learning. But you’ll have to pick. Those are two very different designs with very different purposes. If you try to use one in the place of the other, you might as well be pounding a nail with a fish: no matter how hard you try you are never going to accomplish your purpose and it will always make a mess.
Those advocating for the use of such tests to inform teaching and learning do so without this understanding of what a standardized test is, how it works, what it was designed to do, and furthermore, what it can’t do. That would be like me advocating for one MRI machine over another even thought I know nothing about medicine. Or siding with one side of a legal theory because I heard the words in a movie even though I don’t know what they mean. Or offering up my interpretation of a set of hieroglyphics even though I have no idea what any symbols reference. I wouldn’t do any of those things because I am not qualified as a professional in those areas, which describes most of those who advocate for the misuse of standardized tests to be used as they so often are.
One of the things I find so frustrating about the current arguments regarding testing is that they are so often made under the guise of equity, which is a profoundly important goal everyone in education should embrace and make a priority. But if you care about equity, standardized testing is the wrong tool to get us there. I’ve made the technical argument as to why that is the case dozens of times and in both my books, but here let me make a simpler one: learning is a right I believe students have, and schools must be accountable for constantly shaping themselves to better deliver on that promise of learning going forward. That means they should do so against information that informs them about what was and was not learned, which is not in these standardized tests. Not accidentally. By design.
That means that anyone using these tests to inform learning is committing an egregious and messy error—go back to the visual of the nail and the fish. But where is that error most likely to occur? In the schools that have historically had high test scores or those with historically low test scores? If history offers any lessons, those with historically high-test scores have long been able to focus on learning and are likely to continue to do so, but those with low test scores have been directed to focus on higher test scores. The testing/equity advocacy argument is for those schools with historically low test scores to use the scores from these more frequent standardized tests to inform instruction which they was never in their design, which will encourage historically marginalized schools to act on misinformation. That will leave the students that can least afford it even worse off and help solidify them further into their historically marginalized position. The harder schools try within that very bad system, the worse it will get for the students.
Furthermore, it will solidify in the minds of those who harbor the belief that some students are simply beyond help that they are right in their belief. That some students really aren’t capable of learning. That sort of bias is one of the very things the equity argument is attempting to overcome, and must overcome, which means whatever we do in the name of equity must not perpetuate the bias. Trying to use equity as an argument for testing helps preserve the misuses of standardized testing which will backfire on those the argument was meant to support. Doing it two or three times a year will just compound the negative consequences.
Let’s finally be done with the myth that standardized testing can have whatever meanings we might want to assign to it. We’ve been hammering on that myth for three decades and the misunderstandings in it bear a huge responsibility for the fact that so many of our goals in public education—including the very critical goal of equity—have gone unmet. If equity is the goal, if learning for all is the goal, standardized testing is not the answer, and more frequent standardized testing is just more nonsense, no matter what you name it.
The answer there is a resounding no. No way. No how. Forget it.
The reason? For statisticians to create tests that can be used for comparative purposes the only items that can be allowed on such tests are selected via a very narrow set of statistical criteria. Otherwise, comparisons aren’t possible, or at the very least, wouldn’t mean anything, like apples to basketballs. But what that does is eliminate the opportunity for assumptions to be made about the specifics of what was learned, or what should change in the curriculum, or how to individualize instruction. It means that the content of an individual items is no longer available for interpretation because the content was not the final factor that caused it to be included in the test. It was included because that piece of content in that form produced a highly predictable result when administered to large numbers of students.
The fact that such items cannot inform the specifics of what was taught and learned isn’t an accident or an oversight. You can have one of these statistically based standardized tests that enables comparisons, or you can have a test that is all about teaching and learning. But you’ll have to pick. Those are two very different designs with very different purposes. If you try to use one in the place of the other, you might as well be pounding a nail with a fish: no matter how hard you try you are never going to accomplish your purpose and it will always make a mess.
Those advocating for the use of such tests to inform teaching and learning do so without this understanding of what a standardized test is, how it works, what it was designed to do, and furthermore, what it can’t do. That would be like me advocating for one MRI machine over another even thought I know nothing about medicine. Or siding with one side of a legal theory because I heard the words in a movie even though I don’t know what they mean. Or offering up my interpretation of a set of hieroglyphics even though I have no idea what any symbols reference. I wouldn’t do any of those things because I am not qualified as a professional in those areas, which describes most of those who advocate for the misuse of standardized tests to be used as they so often are.
One of the things I find so frustrating about the current arguments regarding testing is that they are so often made under the guise of equity, which is a profoundly important goal everyone in education should embrace and make a priority. But if you care about equity, standardized testing is the wrong tool to get us there. I’ve made the technical argument as to why that is the case dozens of times and in both my books, but here let me make a simpler one: learning is a right I believe students have, and schools must be accountable for constantly shaping themselves to better deliver on that promise of learning going forward. That means they should do so against information that informs them about what was and was not learned, which is not in these standardized tests. Not accidentally. By design.
That means that anyone using these tests to inform learning is committing an egregious and messy error—go back to the visual of the nail and the fish. But where is that error most likely to occur? In the schools that have historically had high test scores or those with historically low test scores? If history offers any lessons, those with historically high-test scores have long been able to focus on learning and are likely to continue to do so, but those with low test scores have been directed to focus on higher test scores. The testing/equity advocacy argument is for those schools with historically low test scores to use the scores from these more frequent standardized tests to inform instruction which they was never in their design, which will encourage historically marginalized schools to act on misinformation. That will leave the students that can least afford it even worse off and help solidify them further into their historically marginalized position. The harder schools try within that very bad system, the worse it will get for the students.
Furthermore, it will solidify in the minds of those who harbor the belief that some students are simply beyond help that they are right in their belief. That some students really aren’t capable of learning. That sort of bias is one of the very things the equity argument is attempting to overcome, and must overcome, which means whatever we do in the name of equity must not perpetuate the bias. Trying to use equity as an argument for testing helps preserve the misuses of standardized testing which will backfire on those the argument was meant to support. Doing it two or three times a year will just compound the negative consequences.
Let’s finally be done with the myth that standardized testing can have whatever meanings we might want to assign to it. We’ve been hammering on that myth for three decades and the misunderstandings in it bear a huge responsibility for the fact that so many of our goals in public education—including the very critical goal of equity—have gone unmet. If equity is the goal, if learning for all is the goal, standardized testing is not the answer, and more frequent standardized testing is just more nonsense, no matter what you name it.
No comments:
Post a Comment