More than 40 states have notified the U.S. Department of Education that they want to make changes in their accountability plans approved last year under the No Child Left Behind Act.
Federal officials had encouraged states that wanted to make changes that would take effect with this spring’s testing cycle to let the Education Department know by April 1.
“We said, if you get the requests to us by April 1, we will guarantee that we’ll get a response to you in time to make that work,” Raymond J. Simon, the assistant secretary for elementary and secondary education, said in an interview last week.
Most states plan to take advantage of the added flexibility the Bush administration has offered in recent months on the testing of students with disabilities or limited English skills, as well as on the methods states can use to calculate test-participation rates. (“States Given More Leeway on Test Rule,” April 7, 2004.)
But the department last week turned down a request from congressional Democrats to have the additional flexibility apply retroactively to schools that had failed to meet annual achievement targets based on last year’s test results.
States also are modifying their plans after seeing what won approval in other states and learning from their own experiences with the federal law that President Bush signed in January 2002. A few are resubmitting proposals that got shot down the first time around. Some are requesting changes that would require amending the law.
Connecticut, for example, wants to continue testing in reading and mathematics only in grades 4, 6, 8, and 10, rather than in grades 3-8 and once in high school, as the No Child Left Behind law requires. The proportion of Connecticut students who score at or above the “proficient” level on the National Assessment of Educational Progress is typically among the highest in the nation, noted Betty J. Sternberg, the state commissioner of education.
“Given the fact that we had this accountability system, which we believed created that situation, we should be allowed to continue under that system,” she argued. The state would rather devote its time and resources to closing achievement gaps between students of different backgrounds or launching high-quality preschool programs, she said, “than toward creating more tests, which frankly will tell us what we already know.”
Drawing the Line
Connecticut also would like to change how the law’s penalties are applied to schools. It wants consequences, such as the provision of tutoring or permission for children to transfer to other schools, to apply only to subgroups of students who failed to make annual targets, not to the entire school population.
State officials also want to reverse the order of those consequences, so Connecticut districts could offer students free tutoring before the option of switching schools. And the state would like to apply different consequences to schools that fail to meet their achievement targets than to schools that fail to test 95 percent of youngsters, as the law requires.
“There’s an issue about the consequences fitting the crime,” said Ms. Sternberg. She said Connecticut would like to focus its help on the schools with the greatest academic need.
Mr. Simon cautioned that Education Department officials would not approve any changes that they believe require altering the legislation. “Our goal is to wring every ounce of flexibility out of the law, but not to change the law,” he said. “That’s where we draw the line.”
Many states also are asking for more time—and flexibility—to meet the law’s requirements for testing special education students.
At least seven states have requested exemptions to a 1 percent cap of tested students who can score “proficient” on an alternate assessment that does not measure grade-level standards and have it count as proficient when calculating “adequate yearly progress,” a key measure of schools’ success under the law.
In 2002-03, about 9 percent of special education students in Texas took an alternate assessment in reading based on their instructional levels rather than their grade levels. Eight percent took an alternate assessment in mathematics. The state has proposed reducing those figures to 7 percent for AYP purposes in 2004-05 and to 6 percent in 2005-06.
“We are trying to make a major commitment in our proposal to lower our percentages of students taking alternative assessments,” said Criss Cloudt, Texas’ associate commissioner for accountability and data quality. But the next regular session of the legislature doesn’t convene until next January, and most of the rules regarding testing students with disabilities require changes in state law. Schools and districts then will need time to make their adjustments, Ms. Cloudt said.
Montana also has asked for an exception to the 1 percent cap. “It really stems from our rural nature,” said Bob Runkel, the state director of special education.
More than eight in 10 districts in Montana test fewer than 100 students. So under the 1 percent rule, they could not count even a single special education student who took a test that was not measuring grade-level standards as proficient when calculating adequate progress. The state has asked that, in determining such progress, each district be able to count at least two students who did well on alternate assessments as proficient.
Still, Ms. Sternberg of Connecticut says states need even more flexibility to measure the performance of special education students appropriately. The 1 percent rule applies only to students with significant cognitive disabilities. In the past, Connecticut also has given some students with mild or moderate disabilities out-of-level tests, and wants to continue to do so.
Another set of changes states are seeking relates to the statistics involved in calculating adequate yearly progress.
To make adequate progress, a school must meet annual performance targets for its total student population and for subgroups of students who are poor, speak limited English, have a disability, or come from a racial- or ethnic-minority background. States get to decide how big those subgroups must be before considering them for AYP. Many states are proposing changes in those rules, particularly after seeing what other state have been allowed to do.
Tennessee, for example, wants to increase the minimum group size to 55 for students with disabilities or limited English skills, up from 45 under its current plan. It wants a minimum group size of 200 students at the district level. It also wants to reduce the measurement uncertainty involved in making subgroup decisions through the use of a “confidence interval,” a statistical technique that accounts for measurement and sampling error.
“Other states have done that, and we want to do that,” said Connie Smith, the director of school accountability for Tennessee.
Count Just Once
Maryland wants to maintain a minimum group size of five, along with the use of a confidence interval, a decision already approved in its initial plan.
Now it’s requesting that if a student falls into more than one subgroup—free school meals, special education, limited English—the state count the student in only one subgroup.
“It’s extraordinarily logical,” argued Ronald A. Peiffer, a deputy superintendent in the Maryland education department. With a larger minimum group size, certain subgroups disappear from the accountability calculations for almost every school, he said. At the same time, 52 percent of Maryland’s special education students are also eligible for free or reduced- price meals, as are 62 percent of students who speak limited English. “So very frequently,” Mr. Peiffer said, “a school would have a student show up in all three groups.”
The proposed change would do a better job of identifying schools that are really having difficulty, he said, while maintaining the focus on breaking down statistics by student subgroup.
In most states, a school is identified as “needing improvement” if it misses a target in the same subject for two years in a row, even if it’s not the same subgroup that misses the target each time. For example, if African-American students missed the target in 5th grade reading one year and Hispanic students missed it in 5th grade reading the next, the school would still be subject to consequences under the federal law.
At least some states, such as Minnesota, want to identify schools only if they’ve missed the annual target in the same subject for the same subgroup for two years running.
“I’m not sure where that’s going to go,” conceded Jessie R. Montano, the director of No Child Left Behind consolidated programs for the Minnesota education department, noting that the U.S. Department of Education had rejected such requests in the past. The argument for the proposed change is that schools and districts could then focus on consistent areas of weakness.
In California, parents have the right to opt out of statewide testing for their children. Last year, federal officials denied the state’s request not to count such “opt outs” in determining whether a school tested 95 percent of its students. “We’re asking for that same flexibility again,” said Camille E. Maben, a senior adviser to state schools Superintendent Jack O’Connell. “We’ve heard back a couple of times for clarifying information, but no word on whether or not they’ll take the amendments,” she said last week, referring to federal officials.
Although 14 states signed a letter earlier this spring asking the Education Department to consider a more growth-oriented model to measure progress, few states have taken on that issue in their revisions. One exception is Tennessee, which for the past 12 years has been using what is called a value-added system to measure students’ academic growth.
Under the federal law, schools that initially fail to make adequate progress may still be counted as making AYP under a “safe harbor” provision that requires them to show significant improvement over past performance. Tennessee wants to require such schools also to demonstrate growth on its value-added measures. The federal department has already approved the use of such measures by veteran teachers in Tennessee to show that they are “highly qualified,” another requirement of the law.
“If it’s good enough to use as a measure of teacher effectiveness, then we ought to be able to use it for school effectiveness,” Ms. Smith argued."We’re just delighted with the new flexibility that we did not have initially,” she added about the signals coming from the federal agency. “This is a more common-sense approach.”
Assistant Secretary Simon said each state has been assigned a person within the Education Department to work on its requests."Then we all meet weekly to talk about what different states are asking for, and what the issues are, so that we all can deliver a consistent message,” he said. “Keeping in mind that states are unique and what might be approved in one state wouldn’t necessarily be approved in another.”
Once the department is satisfied that it has the necessary information, it will make a decision. After the state incorporates the approved changes in its plan and officially submits a revised version, the department will sign off on it. States are free to continue amending their plans at any time.
“We’re proud of the states for taking another look at their plans,” said Mr. Simon, a former Arkansas state schools chief who joined the federal department last year. “I think it’s a positive sign that states and districts are trying to make this work.”