|A teacher questions how well a standardized test can measure his kids.|
Dear Washington State Science Test Scorer:
I've been thinking about you. I know that one of these days, you'll be opening a box of test booklets that my students worked on not so long ago. For three hours, approximately one hour per day, my 5th graders read and answered multiple-choice, short-answer, and longer-response questions designed to measure their knowledge of Washington state's Essential Academic Learning Requirements for science.
Now it's your turn. You've been hired and trained to score my students' papers with a checklist or rubric. When you are finished, there will be 26 numbers—one for each pupil—that indicate how proficient my students are in science. This year, the test is being piloted. Once it's finalized, however, annual science scores for my school will be published every autumn in our local newspaper, along with scores in reading, math, writing, and listening. People in both Washington state and in Washington, D.C., will make judgments about our school based on those scores and any increases or decreases they show in the next few years.
The funny thing is that I could have told you who was going to do well and who was going to struggle before my class had even sharpened their No. 2 pencils. The test, like many assessments we're being asked to give elementary students these days, contained many open-ended items requiring students to think, analyze, and write at cognitive levels at the upper limit of their developmental abilities. I love posing such questions to students. They're at the heart of any good lesson, and we use them in classroom discussions to learn about science and to learn from each other. But there's a different dynamic at work when such items are given to 10- and 11-year-olds in test booklets. Under those conditions—when students work in isolation and aren't allowed to ask questions or clarify their thoughts through discussion—student performance is highly correlated with factors such as IQ, general background knowledge, attention span, and writing ability.
There are stories behind the names on the test booklets I'm sending you—things I think you should know. If I could sit at your side while you evaluate my students' work, here are some things I'd say:
I realize you'll have to give Vitaly a failing score. Except for the name on the cover, his test book is completely blank. But I hope you don't think he doesn't care or didn't try. Vitaly's family recently immigrated to the United States. He didn't attend school much before he came to this country. Vitaly is excited about learning to read, and he's making good progress. However, he's just now entering what English-as-a-second-language teachers call the "speech immersion stage" of language development—meaning that he's beginning to try out short phrases in English conversation. A couple of weeks ago, when I was checking with Vitaly about his lunch, I asked whether he knew what a peanut butter and jelly sandwich was. He smiled and shook his head no. According to the rules of the federal No Child Left Behind Act of 2001, I had to give Vitaly the science test along with the rest of my class. After he wrote his name on the cover, I kneeled by his desk and we looked at the first page together. I asked him if he understood any of the questions. He wrinkled his brow, stared at the page for a long moment, and then shook his head no. At that point, I patted his shoulder and smiled to let him know it was OK. Then I closed the test booklet and put it away.
Vitaly now knows what a peanut butter and jelly sandwich is; we haven't been able to communicate much yet about the scientific method. But I wish you could see the smile on Vitaly's face when he runs across the field behind our school with an insect net in his hands. And last spring, when we were doing a lab using baking soda to compare the acid content of various liquids, you should have seen how carefully he measured fruit juices with his syringe. Maybe Vitaly will grow up to become a veterinarian or a medical researcher.
Lindsey left quite a few of the essay-type questions unanswered. Reading has never been easy for her, and she sometimes goes off on a tangent when she has to follow written instructions on her own. But you should see Lindsey during class discussions—when I'm explaining how metamorphic rocks are formed or how diabetes affects people. She's like a sponge, soaking up information—and she remembers what she hears. That's the way she learns. Lindsey doesn't say much, but when she raises her hand, I frequently hear a question or a comment that's full of insight. Last fall, our class built rubber-band-powered go-carts. When other students became exasperated by the technical problems of finding ways to increase wheel diameter or rubber band engine power, Lindsey was the one with the ideas and persistence who got things going again. She's also an excellent facilitator, a person who can help three other squabbling students get back on task. I believe Lindsey has a future as a mechanical engineer.
Tests can turn out to be more like sieves than measuring cups. Some of the most important things we do in the classroom end up leaking through the holes.
Logan's test booklet has some excellent answers—and some questions you may think that he skipped. Actually, it's not Logan's style to skip anything. You see, whether I give Logan an art project or a writing assignment, he needs lots of time to process and think things through before he can even make the first mark on his paper. Before I understood this, I sometimes tried to keep him moving along at the pace of his classmates. I quickly found that made both of us unhappy. I eventually discovered that if I let Logan work at his own pace, he would often complete a piece of artwork or writing that was so creative, detailed, and thoughtful that it would give me goose bumps. Even though the instructions in the science booklet said that Logan could have taken as much time as he needed, he was overwhelmed by the number of items on the test. Remember, he's only been an 11- year-old for a month. Although he left many items blank, I'll bet some of his answers are as thoughtful as any you'll see. I'm guessing that Logan will grow up to become a biological illustrator, or perhaps a children's science book author or artist.
I know that the science test I gave was just the pilot. I also know that a committee has yet to evaluate the results and establish a cutoff point that will identify which students pass and which don't. But this emphasis on testing has a downside. I keep thinking about recent inservice training I've attended on teaching reading, writing, and math. The people in charge were all good presenters. They were enthusiastic and had the interests of students at heart. But as I left the sessions, I was saddened by the fact that in each one, I'd heard about how the strategies I was learning would improve my students' scores on state-mandated tests. I didn't hear, in any of them, how the activities would help my students develop a greater love of reading, writing, or math.
I could tell you stories about Michael, Jennifer, Richard, and every other child in my class—but I hope you understand my point. Evaluating students is a complex task. The science tests I just gave—and others like them—can turn out to be more like sieves than measuring cups. Some of the most important things we do in the classroom end up leaking through the holes. A well-constructed test can give teachers important information about what students know and where to focus instructional time. It's just that I'd hate for anyone to think that a single number represents what the 5th graders in our school know about science. It's a lot more complicated than that.
Vol. 15, Issue 3, Pages 43-44Published in Print: November 1, 2003, as Letter Imperfect