It’s safe to say that no one likes to test more than Americans do. Throughout the 19th and 20th centuries, so-called testing experts have come up with scientific ways to sort people by intelligence, moral fiber, criminal leanings and personality traits. Testing experts have claimed to be able to predict things about your aptitude and appetites based on the lumps on your head, the shape of your face, the shape of your brain, the weight of your brain, your interpretation of some random-seeming drawings and your answers to some ambiguous open-ended question.
The so-called IQ test, adapted from the humanitarian ideas of French psychologists Alfred Binet, were for a time, used as a scientific tool for sorting superior citizens from defective ones. That much-vaunted test has been used and abused from 1914 to this very day.
Our obsession with testing has long been reflected in our schools. With the introduction of the high-stakes tests, connected with a bit of bipartisan legislation known as No Child Left Behind (rolled out in 2001), a certain kind of standardized test has become, in many places, the tail that wags the dog.
A lot of people are worried about that. And here’s why: Standardized tests are not like the tests you took in middle school social studies class. You remember back in seventh grade? Your learned about the Greeks and then you took a test on those toga- and democracy-loving people.
There was an easy question that you pretty much could answer if you looked at the book jacket of the textbook. There was a middlingly difficult question and then a question with an answer that demanded you master the material. Using a test like that, your teacher could figure out if you were daydreaming and throwing spitballs during the entire unit or hanging on her every word.
The standardized tests that kids take these days are a different beast. For starters, the subject matter they evaluate is too broad for there to be a full complement of easy, middling and hard questions about all aspects of what was taught. In an effort to boil down tests to an efficient, manageable and cheap-to-grade size, each question functions as a proxy for a large body of material.
In a standardized test given at the end of the seventh grade, for instance, there might be a single question about the Greeks, since that was only one unit of study among many. And that question will, in the view of the people who make up these tests (known as psychometricians), determine whether kids in this day and age have absorbed material from that unit.
Which, of course, means you have to tell a lot from a single question. So the question is, what question should they ask?
Principals often breezily equate high test scores with “we are doing a good job.” And now you can see why that is not exactly true.
Well, the experts have determined that on average, between 40 and 60 percent of kids will answer legitimate test question correctly. So in order to come up with that crucial question, our psychometricians come up with a single question and put it on a pilot test. Then it is given to kids, but it doesn’t really count. If between 40 and 60 percent of kids get it right, voila! It is considered a statistically sound question. And they’ll probably use it as a question on the tests given to next year’s batch of seventh graders.
Got all that? Good. So what makes people so crazy about standardized tests? The problem is not the tests—which are, after all, a kind of bland instrument. The problem is that schools start teaching to the test. Now remember, 40 to 60 percent of kids have to get it right. So the test questions by their very nature have to be pulled from the bottom third of the curriculum—the easy and easy to middling area—so that 40 to 60 percent of kids can get the answer correct. So if schools are teaching to the test, those kids are only getting taught what amounts to the bottom third of what they are supposed to be getting taught.
And that’s a problem.
Because real learning, the kind that builds on itself year after year, the kind that will prepare your child to thrive in high school and beyond is about complexity, not simplicity. Kids need to learn a whole lot more than what is going to be on those tests if they want to do well in school and in life.
If you want to know more, I recommend you read Harvard Professor Daniel Koretz’s book, Measuring Up: What Educational Testing Really Tells Us. He’s a very smart guy who will give you the good, the bad, and the ugly on standardized tests.
But even without doing a deep dive into the subject, you can look at test scores in a more sophisticated way. Is your child enrolled in a school that has high tests scores? That’s good, but there’s more to know.
Principals often breezily equate high test scores with “we are doing a good job.” And now you can see why that is not exactly true. To find out how the school is really doing, you can use the test scores to ask a few more questions.
See if you can get a few minutes alone with the principal. Now, here are the silver bullet questions to ask: How long have those test score been high? (A one-year bump in test scores doesn’t mean much.) When you break kids by ability levels, what are the trend lines for the lowest-performing kids? This tells you a lot because if the lowest performers—the hard to teach kids—are learning more, chances are good the teachers are engaged in solid teaching.
And more importantly, what is the principal doing to create those high tests scores—for strong learners and struggling learners alike? Is he/she hiring a librarian to get kids to read more history and a very lively art teacher who is turning them into little Pollocks and Rothkos? Great. Did he/she fire the librarian and hire one of the “how to teach to the test” firms that forces teachers to dumb down their lessons? Not so good.
Remember, both those schools might have solid test scores, but how they got there makes all the difference.