Study praises FCAT as indicator of learning
February 11, 2003
By Leslie Postal | Sentinel Staff Writer
As thousands of Florida students start taking FCAT today, a new study finds the test a reliable gauge of academic performance and refutes the complaint that it encourages teaching to the test over real learning.
The Florida Comprehensive Assessment Test, the linchpin in the state's education reform efforts, comes with high-stakes -- the FCAT is used to grade public schools, help determine third-grade promotions and decide whether high school seniors earn a diploma.
This year's FCAT season begins today, as students sit for writing exams. Reading and math tests and a new science exam will be given in the coming weeks. Students in third through 10th grades take the tests.
FCAT critics argue that high-stakes testing creates a "narrow" focus that doesn't lead to "real learning." But the improved scores Florida students have shown in recent years appear to reflect genuine learning, according to a study by the Manhattan Institute of Policy Research.
That's because Florida students have also improved on the Stanford 9, a national test given at the same time as FCAT. The national test comes with no consequences, so "it's like an audit" of FCAT, said Jay Greene, a senior fellow with the institute and the study's lead author.
Similar results on the two tests, Greene said, suggest improvement on FCAT has come from real learning and has not been distorted by the pressure and consequences tied up with the exams. If teachers are teaching to the FCAT, he added, they are teaching what most everyone wants students to learn.
"If the test is well designed, and it's administered properly, the teaching to the test might be a good thing," he added. "It means students are learning to read and to do math."
Greene of Davie said his study should put to rest that one major complaint about FCAT. But it also raised some questions about testing programs elsewhere.
The institute's study backed up what Florida education officials have said about FCAT -- but contradicted two recent studies by Arizona State University researchers.
Audrey Amrein, one of the Arizona researchers, contends high-stakes tests have become an "easy way to control what's going on in the schools," though they haven't led to real progress.
Her studies compared results on high-stakes state tests with results on national tests, such as the SAT, Advanced Placement exams and the National Assessment of Education Progress. They found states with high-stakes tests had made little gains, or had posted declines, on these national exams. Their results for Florida, however, were not clear-cut.
Greene criticized those studies, saying it didn't make sense to compare scores on tests only college-bound students take with scores on yearly state tests almost everyone takes.
His study looked at high-stakes and low-stakes tests given to the same groups of students at the same time. The study looked at Florida and Virginia and at seven school districts including Boston and Chicago.
In both states and all the districts, the standardized tests were an accurate measure of student performance. But only in Florida did the high-stakes tests also accurately measure a school's role in student achievement from year to year. Leslie Postal can be reached at email@example.com or 407-772-8046.
©2003 Orlando Sentinel