‘I’ ON CULTURE
Surprise! Everyone agrees that the FCAT is a lousy system for testing children. It took only 10 years for general agreement. After all, we all want bright, successful kids. If the tests don’t show some real progress by our children, the obvious next step is to get rid of the test.
Florida is looking for ways. Some teachers have already been told that a new test might replace the Florida Comprehensive Assessment Test, testing “higher-order thinking skills.” In other words, it will actually be harder. Students wouldn’t have to answer dumb questions based on reading paragraphs; they’d have to work out complex issues and make inferences. You might ask, “But won’t that mean fewer students will pass?” At that point, educators will hurriedly assure you that the new tests will just show what a great job they are doing.
Logically, a new, more challenging test should make it harder for borderline students to pass. But, as in most things these days, there is a cute catch. One of the real problems of the FCAT is the use of those pesky multiple-choice questions. If a student facing a math problem asking for the total of two plus two and he selects five, well, he is wrong. If he misses enough of those questions, he fails. If he needs to get 40 out of 60 right and gets only 39 right, he fails. Either you are correct or you are wrong. Nothing in between.
On the higher-order thinking tests, things get a lot fuzzier. When I was a school administrator back in New York, I was convinced that turning the three-hour English Regents Exam that had a lot of multiple-choice, fill-in and spelling questions into a six-hour, essentially all-essay, exam would be a disaster. Instead, a higher percentage of students passed that test, which measured mastery, than the test designed to ensure simple literacy.
The reason was that in essay questions, there are no simple right or wrong answers. Teachers were carefully trained to mark on a scale of one to six. One meant the kid really didn’t write anything much, and six was for only the very best. Very, very weak answers got a two, and strong essays got fives. But threes and fours were rather fluid. Four was the passing score; three meant it was a good try but not quite good enough. And markers knew that the bosses wanted a higher percentage of kids to pass. And a one-point change really seemed not that large. Using a bell curve, most answers should have been three or four and probably around the same numbers for each. But there were always a lot more fours.
The markers were well aware that a minor change in enough students’ grades would be welcomed by all, challenged by almost no one. And if enough students did not pass, markers might just lose jobs. Machines that mark multiple-choice questions don’t have those worries. And everyone celebrated; administrators breathed a sigh of relief; no one tries to get rid of a principal when the kids are doing fine. And teachers could relax, knowing that there would be less pressure. Parents would smile since their children would graduate from high school.
Of course, the college teachers and employers of the students were not nearly as pleased a couple of years later when they discovered that the literacy problems were now on their shoulders. So our kids might be going to a new world of testing where, if two plus two doesn’t equal five, they can get credit for staying within single digits with a bit more because, after all, five is not much more than four. So they will be almost right, and that will make things all right.
Formalized tests can then join horseshoes and hand grenades as events in which a close miss still counts. School boards will be happy because there will be fewer complaints, and parents in a lot of countries around the world will probably be pleased that their students who work their behinds off for an average of about a month’s more schooling each year and then go to special “drill classes” to make sure they retain the information will not have to worry about competition from American kids.
The FCAT is very flawed, but let’s be careful about what replaces it.
