Reading exam flunks the accuracy test
By A TIMES EDITORIAL
Published April 30, 2007
If a standardized test is to measure how much a student learns each year, then it must be calibrated with precision. But the FCAT, which Florida uses to assess students, teachers and schools, takes an inexplicable detour in high school. Its reading exam makes 10th-graders look as though they suffer amnesia, and Gov. Charlie Crist needs to ask why. Why have state education officials refused to fix their own mistake?
This testing irregularity is unmistakable. Only 32 percent of last year's 10th-grade class passed the reading FCAT, but three years previous the rate was 52 percent. More telling, these same 10th-graders scored nationally in the 67th percentile on the Norm-Referenced Test. By comparison, 75 percent of third-graders passed the reading FCAT while scoring in only the 61st percentile nationally.
Across the board, from third to eighth grades, FCAT reading scores correlate directly with the national test. In ninth and 10th grades, that link is shattered.
"What the (FCAT) data tells people is that their students are performing badly all of a sudden, " says Barry Farley, assessment coordinator for Lake County schools. "The truth of the matter may be, for thousands of these kids, they are performing at the same level, but the scale has changed."
This disparity traces to the Department of Education and decisions it made, in 1998 and again in 2001, to ignore the recommendations of an education panel and put a thumb on the 10th-grade reading scale. But the costs of this error have grown enormously.
Teachers and schools are now being judged on how much their students learn each year, yet the reading scale distorts the picture at high school. That's one reason elementary and middle schools are awarded A's at nearly four times the rate of high schools, why high schools last year received only $23-million of the $151-million in school bonus awards. It also may explain why 11th-grader Kristen Jackson of Tampa is in danger of not receiving a diploma, even though she has scored in the 96th percentile on the national reading test.
Since the FCAT was first created, it has come to be used in ways for which it was never envisioned: to determine whether some students can be promoted and others can graduate; to decide whether schools get financial bonuses or put on probation and possibly closed; even to help compute a teacher's annual pay.
Whether or not that is asking too much of one standardized test, this much is clear. With the stakes so high, the test itself has to deliver accurate results. That's obviously not the case in high school, and state education officials cannot keep pretending otherwise.
Crist, who met recently with Hillsborough superintendent MaryEllen Elia to discuss this issue, told reporters: "If there are always ways that we can make it, you know, better, why wouldn't we?"
But Crist, who appoints the board that sets these test scales, will find that the bureaucrats are not eager to admit their own mistakes. They accept no excuses from teachers, and Crist should accept none from DOE. A flawed test formula produces flawed results, and this time DOE must be held to account.