Rip Van Winkle's One Familiar Thing: The Absurdity of Assessing 21st Century Learning With 20th Century Tests
Imagine yourself as a World War I-era Rip Van Winkle. One sunny day in 1914, you drive your Model T to a pleasant meadow, stretch out under a tree, and lie down for an afternoon nap. You wake up a century later in a Trader Joe's parking lot.
Several aspects of the year 2014 come as a shock to you.
People hold devices the size of sardine tins in their hands, talking into them like lunatics in an asylum. They keep pecking at the shiny side with one finger, like chickens pecking at corn.
Though black coffee is familiar, people are paying the price of a ham for a single cup.
Military tanks called "Hummers," driven by civilians and painted bright yellow, rumble past at the speed of trains.
Exhilarated and terrified, you stumble into a nearby school. Chalkboards have been replaced by big glowing rectangles, and every classroom has a paperless typewriter hooked up to a box with its own smaller glowing rectangle.
Suddenly your eyes alight on a familiar object. A bubble test, invented by Frederick Kelly the year you stretched out in the grass for your century-long nap. Reassuring bubbles inked on paper, one right answer to be marked with a pencil.
You breathe a sigh of relief at this one familiar thing. At last, a little piece of 1914 preserved through the century's turn.
New Tools for New Times
Over the past century, our tools have changed dramatically. A powerful telescope recently found a planet not only the same size as Earth, but roughly the same temperature--a bit cooler, similar to San Francisco on a cloudy day. Teenagers carry around iPhones more powerful than computers that filled entire rooms a few decades ago.
If we wanted to, we could send astronauts to Mars; cost is a barrier, but the technology exists. Why, given these remarkable advances, are we still using paper-and-pencil bubble tests? (This is not a rhetorical question. I'm genuinely bewildered.)
Using century-old technologies in other domains would be absurd. Kodak's autographic system for roll film has no place in the iPhone era. In a nuclear age, German torpedoes that once sank British ships are obsolete.
The flaws in multiple choice tests are obvious. There's one right answer, while the complexity of actual workplace tasks and life decisions requires systems-thinking. All kids get the same questions and answer choices, though every child will be at a different level of understanding in the subject being tested.
Added to these limitations, you have delays in scoring so long that 5th graders who took a test last week might not get their scores until August--and when they do, they'll get no qualitative feedback, just a number.
Given the quantum leaps in technology over the past century, how could tests reflect new possibilities? Here are two innovations I'd like to see; I'd love to hear your own ideas for what's possible.
1. Tests could involve simulations of actual performance tasks.
Students could perform virtual science experiments, or engage in engineering applications that required them to design structures like buildings and bridges. Even with conventional academic content, students could manipulate a 3-D object, digitally dissect a frog, or use online research tools to create a report and presentation on a given topic.
I realize these kinds of tests already exist. Some of them are being piloted by PARCC and Smarter Balanced for the Common Core assessments. But this isn't cutting-edge stuff--it should have happened by now. When it comes to the technology involved, the average video game is several decades ahead of the average multiple-choice test.
I've heard of a few examples of 21st century technology used at the university level.
Pre-service teachers can learn classroom management in a simulated classroom, complete with student avatars in baggy jeans and One Direction T-shirts. Medical students can put on gloves to perform digital surgery without threatening a living patient or wasting a perfectly good cadaver. But even at the university level, multiple-choice and true-false quizzes are disturbingly prevalent.
2. Tests could measure individual growth.
Again, these tests already exist, but they're almost never used for high-stakes testing. I'm not talking about the "from Below Basic to Basic" degree of precision with most state Benchmark tests, as blunt an instrument as a stone club.
The computerized MAP test is multiple choice, but it's similar to the GRE in that it differentiates the level of questions based on whether a child gets the initial questions right or wrong. A 3rd grader might end up with questions on a 2nd grade level or a 5th grade level, depending on her reading and math abilities. You also get the results back in about five seconds instead of five months.
Most importantly, the MAP tests measure individual growth over the three points in the year the students take it. When my students took the test, it was easy to emphasize growth over the score itself. A struggling reader whose reading score increased by 12 points was celebrated just as much as an advanced reader whose score was two years above grade-level.
I'm bewildered by the fact that these kinds of tests haven't replaced paper-and-pencil, one-test-fits-all relics. I'm not asking for a hovercraft to replace my turn-of-the-century Model T. I'd be satisfied with trading up to a '98 Honda Civic.
What Will It Take?
We've come a long way from the world of 1914. Kids know it. Parents know it. Teachers know it. Principals, most superintendents, and the majority of commissioners know it.
What's more, we all despise bubble tests and the shifting, bizarre consequences attached. A principal at a high-poverty, high-achieving public school posted this query on her Facebook page: "The state test is like The Hunger Games because..." She got brilliant responses from teachers and administrators alike, including, "...just when you think you've won, they change the rules."
Nobody pines for kerosene lanterns, mimeograph machines, or 8-track tapes. What will it take to break our dependence on the 1914 bubble test?