Tennessee officials are cheerfully announcing the advent of awesome new assessments that will be "not a test you can game." The tests will be delivered at the end of the year by winged unicorns pooping rainbows while playing a Brahms lullabye on the spoons.
“We’re moving into a better test that will provide us better information
about how well our students are prepared for post-secondary,” Tennessee
Education Commissioner Candice McQueen told reporters recently during
sneak peek at some of the questions.
Fans of the new standardized tests continue to be impressed at how these tests solve the problems of education in the 1970's. Assistant state commissioner of data and research Nokia Townes says that the tests will require "more than rote memorization," which puts them on a par with the tests I started giving my students thirty-five years ago. But that statement also indicates that officials still don't understand what "gaming the test" means.
Standardized tests are, by their nature, games. Where you have multiple choice questions, you have one correct answer and several other answers designed to trick and trap students who might make a particular mistake, so by their nature, they are not simply trying to capture a particular correct behavior, but are also testing for several incorrect ones.
But Tennessee officials are distracted themselves, focusing on unimportant test features. The questions have drop and drag! Which is, of course, simply a bubble test question with dragging instead of clicking or bubbling. The questions require multiple correct answers! Which is just a bubble test with more options and two bubbles to hit.
Because standardized tests are a game, gaming them will always be possible. The tests involve plenty of tricks and traps, and so we teach students how to identify those and spot when testers are trying to sucker them in a particular way. We learn specialized testing vocabulary (this is what "mood" means to test manufacturers). We learn what sorts of things to scan for in response to certain types of questions. In short, we teach students a variety of skills that have no application other than taking the Big Standardized Test.
Nothing has changed, really. The advantage of a standardized multiple choice test is that it can be scored quickly and cheaply. The problem is that it can't measure much of any depth. At its worst in the bad old days when Hector and I were pups, it measured recall. In the new and improved days since, it measures whether or not the student will fall for particular tricks and traps-- which may or may not have anything to do with how well the student understands and applies the understanding. And since our focus under the Core is almost entirely on performing certain operations and not at all on content, we're not really testing a student on what she knows, but on whether she can perform the required trick (of course, we're also testing whether or not she's willing to perform the trick, but we never, ever have that discussion).
The article says that the "oldest and most potent criticism of tests" is that "they force teachers to 'teach to the test' and focus unduly on
memorizing facts and testing tricks that students promptly forget after
completing the test."
Memorizing facts? Maybe. Testing tricks? Those will always and forever be part of test prep for standardized tests because those tests must be speedy and cheap, and the only way to do testing speedy and cheap is by building it out of testing tricks. And any testing tricks that manufacturers can use to create a test, students can learn to take it.