Suddenly, 10th-graders are FCAT flops
A moving bar makes failures of students who tested well once and still outrank U.S. peers.
By LETITIA STEIN and THOMAS C. TOBIN
Published April 15, 2007
Many educators aren't sure what test scores say about the reading ability of Florida's high school students. The state's ninth- and tenth-graders do poorly on the FCAT, which sets a higher bar in high school than in lower grades. But the same students perform well on the Norm-Referenced Test, or NRT, which measures them against their peers nationally.
[Source: Florida Department of Education | Times Chart: Ron Bracket]
Florida's 10th-graders look like terrible readers. Their FCAT scores are the worst in the state.
Yet those same students are among the best readers on a test that compares Florida students with their peers across the United States. They also score well on the FCAT math test.
Why the confusing results?
Blame an FCAT system that holds students in different grades to very different standards.
Nowhere is that more apparent than in high school, where the bar is highest. Only one-third of Florida's 10th-graders met FCAT reading standards last year. By contrast, nearly two-thirds of seventh-graders passed.
The disparities have consequences:
More than half of Florida's elementary schools earned A's last year, compared with fewer than 20 percent of high schools. Elementaries received $81-million in FCAT reward money. That compares with $23-million for high schools.
"We do not make kids dumber when they come to high school," said Jeff Boldt, the principal of Chamberlain High School in Tampa, which has earned straight C's since school grades debuted.
State officials acknowledge the standards are far more rigorous for high school students, but say they need to be to prepare them for college and work.
But some testing experts say large inconsistencies between grades and subjects can undermine confidence in the system.
Kristen Jackson, an 11th-grader at Tampa's Alonso High, has narrowly failed the FCAT graduation requirement in reading twice. But by another reading test, she can read as well or better than 96 percent of her peers nationally.
"I was actually crying when I failed," said Kristen, who earns A's and B's. "It tortured me. It was a horrific experience."
- - -
From the first day of school, principal Bob Heilmann starts telling ninth-graders at Hillsborough's Riverview High to brace themselves for a harder FCAT.
He asks them: How many students ran a mile in middle school? How fast?
"Eight minutes," they say. "Nine minutes."
"In high school, you have to do it in six minutes or under," Heilmann warns students, comparing their mile times to the high-stakes test. "The FCAT in high school is a different animal."
The warning rings true across Florida.
"I don't think there's any question that the standard is higher at high school than it is at elementary school," says Lee Baldwin, director of testing for Orange County schools.
"What the 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."
Florida's FCAT greets ninth- and 10th-graders with a one-two punch:
- Ambitious "cut scores" require students to correctly answer a greater number of difficult questions to be considered proficient.
- The test itself becomes harder and more tedious. In 10th grade, passages average 900 words and feature mostly informational text, which many students find less engaging than the literary readings that are stressed on lower-grade FCATs.
"It's like sitting down and reading your tax manual if you've never done taxes," said Lynn Dougherty-Underwood, a reading supervisor for Hillsborough schools.
Last year's class of 10th-graders felt the impact. When they were in seventh grade, 52 percent scored at their grade level in reading. Three years later, only 32 percent did as well.
Did skill levels really drop 20 percentage points in three years? Or is something else going on?
"It's almost an unrealistic expectation," said Cathy Fleeger, an assistant superintendent in Pinellas.
To better understand what was happening, Hillsborough compared the state's two required tests - the FCAT, which measures how kids are grasping state standards, and the Norm-Referenced Test, or NRT, which compares Florida students with the nation's.
The district found that 10th-graders who just met Florida's reading standards ranked in the top 20 percent of their peers nationally last year, while third-graders at the same level placed just above the bottom third. At most other grades, students in the same situation were near the national average.
In Pasco County, school officials found similar results: A typical 10th-grader just making the FCAT graduation requirement scored as well or better than 64 percent of peers nationally on the NRT.
Anyone would expect an above-average student to graduate, said David Scanga, Pasco's director for research and evaluation. "It's very difficult for the parents to understand."
- - -
When Florida officials set the FCAT bar high, they did so with their eyes wide open.
In September 1998, nearly 80 teachers spent four days at a Palm Harbor resort helping the state set initial cut scores for fifth-grade math and the reading test in grades four, eight and 10.
A movement to increase standards was in full swing. Reports had surfaced that American students lagged behind their peers in Asia and Eastern Europe. Some Florida educators mocked the "minimum competency" tests of the 1970s and '80s.
"The politicians were going around waving their arms calling for world-class standards, and that tendency was driving standards up," said Howard Mitzel, a well-known researcher in the field of standard-setting.
Teachers at the conference answered the call. Their sobering task: Recommend scores that would label kids as passing or failing.
Make the scores "challenging but reasonable," Tom Fisher, then the state's testing director, told them. Use your best judgment, he said. "That's all you can do."
Still, many teachers were stunned when the results of their work appeared on a projector screen. Using scores from the first FCAT in 1998, computers predicted how students would do in later years.
The numbers weren't pretty.
Tenth-graders would be among the worst in reading, with 61 percent falling below "satisfactory."
A little-noticed adjustment took place after the teachers adjourned. State education officials raised the 10th-grade reading bar even higher, and lowered the math standards.
Then-Gov. Lawton Chiles defended the high standards, especially in reading. Kids needed to be qualified for the work world, he said, and "that hasn't been happening in Florida."
His successor, Jeb Bush, soon raised the stakes, latching on to the FCAT as a way to hold schools accountable.
In 2001, the state again called on teachers to help set cut scores as Florida expanded the FCAT assessment to all students in grades three through 10.
As in 1998, the state took the teachers' work and went them one better in high school, upping the recommended cut score in ninth-grade reading by 40 points while reducing it by 19 points in math.
Cornelia Orr, the state's top testing official, said the 40-point increase was necessary so the ninth-grade cut score would fall between the eighth- and 10th-grade scores established three years earlier under Chiles.
As for the lower bar in math, she explained that many high school students weren't getting the math courses they needed in time to do well on the FCAT. The state has started to address that problem by requiring those courses earlier in high school, she said.
Nine years later, the cut score decisions on reading and math are one reason Florida high school students perform so differently in the two subjects - unlike students in most of the lower grades.
- - -
Meranda Maley, a senior at Tampa's Alonso High, said she did well on the FCAT in middle school. But when she moved to high school, her performance suddenly dropped off.
"I'm like, how come I understand something in eighth grade, 'cause my scores were really high? And then I get to 10th grade and I'm like, wow ...," she said, her voice trailing off as she tried to figure out what happened.
She's not alone.
Last year, 32 percent of 10th-graders tested at their grade level in reading on the FCAT. Across the state, 126,000 students failed to meet the high standards. Ninth-graders scored poorly as well, while students in most of the lower grades did reasonably well.
Assessment experts say states should avoid large disparities in scores between grades and subjects.
Teachers, they say, won't know how to prepare their students. Parents, students and the public will question the system's validity when results appear to be based more on cut scores than effort.
"It should be the same level, or hoop, to jump at every grade," said John Hilderbrand, director of assessment and accountability in Hillsborough.
The inconsistencies suggest Florida should reevaluate where it placed the FCAT bar, said Joseph Torgesen, director of the Florida Center of Reading Research, who remains confident in the test itself.
In 2003, Florida officials decided not to raise FCAT cut scores even more, as originally planned. Neither the system nor the public was ready, they concluded.
Raising third-grade scores, for example, would have expanded the controversial practice of holding back low performers from fourth grade.
"The ultimate solution probably will involve some combination of raising the standards in the lower grades and lowering the standards in the higher grades," said Torgesen, a professor of psychology and education at Florida State University.
That could be difficult in an era when educators speak of raising middle and high school standards. Then again, Florida's bar for high school students may be so high the public might see the need to bring it down, Torgesen said. "People would generally understand."
Letitia Stein can be reached at firstname.lastname@example.org or (813) 226-3400. Thomas C. Tobin can be reached at email@example.com or (727) 893-8923.
[Last modified April 14, 2007, 22:36:20]
[an error occurred while processing this directive]