TALLAHASSEE, Fla. (AP) -- Florida's interim education commissioner told worried lawmakers on Thursday that glitches in the first release of scores under the state's new teacher evaluation system were being fixed and that there's no reason to delay its implementation.
The Department of Education took down a website showing nearly 97 percent of Florida teachers were rated "effective" or "highly effective" in the last school year within hours of putting it up on Wednesday.
"This is a painful year," Stewart told the Senate Education Appropriations Subcommittee. "Any time you implement something this large for the first time there are growing pains."
The website was restored later Thursday after being corrected. The statewide results hardly changed. They now show 21.9 percent of teachers were highly effective instead of 22.2 percent and 74.6 percent were effective compared to 74.5 percent. The differences were similarly small for teachers evaluated as needing improvement, developing and unsatisfactory.
The problem was that some districts submitted duplicate information because separate salary reports are made for teachers who are paid from more than one source.
"Obviously, for teacher evaluation they should be reported once," Stewart said.
The adjustments did not affect individual teachers' evaluations. The report is preliminary as some school districts failed to meet a Nov. 30 deadline for submitting their data. A final report including all districts is scheduled to be released in January.
The evaluations are based heavily on tests that measure student progress. They eventually will be used to make decisions on which teachers should be retained and how much they should be paid.
The Florida Education Association, the statewide teachers union, and some local school officials have urged the state to delay full implementation of the evaluation system that's set for the 2014-15 school year.
Stewart said the biggest problem still looming is that not all students are tested. The most common exam, the Florida Comprehensive Assessment Test, includes reading, writing and math but not for all grades. Students don't begin taking the FCAT until the third grade.
As a result, some school districts are using school-wide averages to assess teachers whose own students aren't tested. Those teachers, including Stewart's daughter, who teaches kindergarten, are being evaluated based on tests taken by some students they've never taught.
"That's one of the things that we are facing that is problematic," Stewart said. "That is one of the biggest concerns that we hear and will be corrected each year as we move forward."
The solution depends on the ability of districts, in some cases working with the state or other districts, to develop local tests for grades and subjects not covered by the FCAT or other standardized exams. Florida, meanwhile, is planning to replace the FCAT with end of course tests and other exams as it phases in new common core standards over the next two years.
Sen. Dorothy Hukill, R-Port Orange, noted wide discrepancies in evaluations from one school district to another and questioned how they can be compared.
They really can't, Stewart replied, because the districts have the flexibility to set up their own measuring criteria within state guidelines. For instance, she said, some factor in whether a teacher's students meet preset achievement goals and others do not.
Stewart received a vote of confidence from Sen. Bill Montford, a Tallahassee Democrat who's also CEO of the Florida Association of District School Superintendents.
"We know that was the first shot across the bow and is not anything to get too upset about," Montford said. "We'll have to work and change it and make it better."