By Joshua Benton
Staff Writer
Page 1A
Texas officials have tried to artificially boost test scores by eliminating 10 percent of the state’s students from the No Child Left Behind accountability system – including many of the state’s most disadvantaged children.
But federal authorities quietly blocked the attempt last month – along with three other proposed changes that would have improved the appearance, if not the reality, of Texas schools’ performance.
It’s the latest step in the continuing dance between the U.S. Department of Education and states seeking to make life easier for their schools.
“We have this race-to-the-bottom problem,” said Kevin Carey, a researcher at the Education Sector think tank who has studied how states negotiate with the federal government. “One state comes up with a particular wrinkle that has the effect of reducing pressure on schools to achieve. Other states notice it and say, ‘Oh, yeah, can we do that too?’ ”
Under the No Child Left Behind law, American schools are evaluated each year on whether they have made “adequate yearly progress,” or AYP. Schools that don’t make adequate progress for several years – usually because of low test scores – can lose federal funds or be shut down.
But states are allowed some flexibility in how their schools are judged.
States regularly propose new statistical methods for increasing the number of schools making adequate yearly progress, and it’s up to federal officials to approve or reject them.
Calendar rule
The changes Texas proposed are complex, but together they would have allowed hundreds of low-scoring campuses to get off of the federal list of failing schools.
“We tried to look comprehensively at what had been approved for other states to see if Texas could take advantage of those as well,” said Criss Cloudt, the associate commissioner for accountability at the Texas Education Agency.
TEA requested the changes in an 11-page letter to the Department of Education sent Feb. 1. The most sweeping was a change in the way a school would determine whether a student’s scores counted in its passing rates.
Currently, students’ scores on the spring Texas Assessment of Knowledge and Skills tests count only if they were enrolled at the school on the last Friday of the previous October. That’s the day Texas has long compiled its official enrollment count.
The idea is to avoid penalizing schools for the performance of students who arrive in a school only a few weeks before test day.
But TEA wanted to expand that protection and proposed counting students scores only if they had been enrolled in the same school district a year earlier.
In other words, a student who moved to Dallas in June and attended a local school the entire school year would still not be counted in that school’s scores the following spring.
The change would have boosted schools’ showings because students whose families move around a lot tend to have lower academic performance than those with more stable home lives.
Katrina evacuees
The state estimated that, had the rule been in place last year, 90 Texas schools that failed to make adequate yearly progress would have been saved from that fate. Because of the way adequate yearly progress is calculated, the rule would also have allowed hundreds more schools to avoid being judged on the scores of minority subgroups.
TEA leaders argued that other states have been allowed similar calendar arrangements. But federal officials said Texas’ request went too far because it would eliminate one out of every 10 Texas students from the school-rating process and disproportionately affect the disadvantaged.
“What they submitted to us would result in the exclusion of more students,” said Catherine Freeman, a senior policy adviser at the Department of Education. “We obviously don’t look favorably upon that.”
Even though some other states have had a similar calendar approved, each state is evaluated individually for the impact a proposed rule would have on test results, she said.
Dr. Cloudt said the impetus behind the proposal was to find a way to exclude Hurricane Katrina evacuees from the testing rolls of Texas schools. But the state’s proposal makes no mention of the hurricane and would have excluded many more Texas natives than hurricane evacuees.
The federal government has since approved a separate plan to exclude all Katrina evacuees from AYP performance ratings this year.
‘Safe harbor’ changes
Along with the calendar proposal, Texas also requested several changes in what’s called the “safe harbor” process. That’s a provision in No Child Left Behind meant to protect schools that have low scores but are improving rapidly. TEA wanted to give schools more ways to argue that their scores are improving.
One, called a “matched profile analysis,” would have let schools boost their test scores if the racial composition of their students had changed in the previous year.
Another change would have allowed schools with low scores to qualify if those scores fell within a “confidence interval” – in other words, if there was a decent chance the school could have scored better if its students were tested again.
Federal officials rejected them all.
In a letter to state Education Commissioner Shirley Neeley, Assistant Education Secretary Henry Johnson wrote that the federal department “questions the validity and methodology” of the racial-composition proposal and that the use of confidence intervals was unnecessary.
He also questioned the state’s proposed use of a particular formula in calculating scores because “the formula design is more likely than other possible formulas to accept that a school or district had made AYP.”
Special ed rule
Texas used several of those proposed safe-harbor provisions last year – including racial matching and confidence intervals. Dr. Cloudt said they will now stop using them.
Finally, the Department of Education also rejected Texas’ request to count students as “special education” for up to two years after they stop receiving special-education services. That would boost special-education scores substantially in many schools.
Each of the four major rejected changes would have saved between 66 and 90 Texas schools from missing AYP in 2005. In all, they could have flipped the results of 303 of the 1,152 failing schools in Texas last year – although it’s likely that some of those schools would have been helped by more than one of the changes. Last year, 15 percent of Texas schools failed to make adequate yearly progress.
Dr. Freeman said the Department of Education is seeing a small boom in the number of states seeking changes like the ones Texas proposed.
“They’re definitely looking for ways to reduce the number of schools to be identified as missing AYP,” she said.