Following months of pressure from states and education groups, U.S. Secretary of Education Margaret Spellings has announced a pilot program that will let some states use what are known as growth models to evaluate the success of schools and districts under the federal No Child Left Behind Act.
Growth models track the progress of youngsters over time and could give schools credit for moving students toward proficiency.
Proponents argue that such models are a fairer way to judge schools than comparing the performance of different cohorts of students.
To make adequate yearly progress, the law currently requires schools and districts to meet annual targets for the percentages of students who score at the proficient level or above on state reading and mathematics tests in any given year.
Watering Down Demands?
“I think it’s probably the most important thing the department could do at this point,” Allan Olsen, the president of the Northwest Evaluation Association, a nonprofit test publisher in Portland, Ore., said of the Department of Education initiative, which Ms. Spellings announced here on Nov. 18. “Ultimately, the measure of individual student growth is the best indicator of whether a school is effective or not.”
“We see this as an opportunity to raise the bar and build a fairer system for our school districts and states,” said Valerie A. Woodruff, Delaware’s education secretary and the president of the Washington-based Council of Chief State School Officers.
But some civil rights groups expressed serious concerns about the proposal, which they fear could lower expectations for the nation’s most disadvantaged young people.
“We are particularly worried about how the children who are most behind and most often left behind will fare when states change their accountability systems,” the Citizens’ Commission on Civil Rights, a watchdog organization based in Washington, said in a statement. “We fear that too many school board members and educators will view the growth-model approach as an invitation to water down expectations for student achievement in order to reduce the numbers of schools identified for improvement.”
‘Bright Line’ Principles
Secretary Spellings announced the pilot program, for up to 10 states, during the annual meeting of the CCSSO, held in Richmond, Va. States must apply for the program by Feb. 17 if they want to use the new evaluation methods for the current school year.
The U.S. Department of Education says it will let up to 10 states use growth models this school year to measure whether schools and districts made adequate yearly progress under the No Child Left Behind Act, as long as those states’ models:
• Expect all students to be proficient by 2014 and require that academic-achievement gaps be closing for all groups of students;
• Set expectations for annual achievement based on grade-level proficiency, not on students’ background or schools’ characteristics;
• Hold schools accountable for student achievement in reading/language arts and mathematics;
• Ensure that all students in tested grades are included in the assessment and accountability system, hold schools and districts accountable for the performance of each student subgroup, and include all schools and districts;
• Test students in grades 3-8 and once in high school in both reading/language arts and math with assessments that have been given for more than one year, have been approved by the Education Department for the 2005-06 school year, and produce comparable results between grades and school years;
• Track student progress as part of the state data system; and
• Set criteria for making AYP that include participation rates on state tests and at least one other academic indicator, such as attendance rates for elementary schools and graduation rates for high schools.
SOURCE: U.S. Department of Education
“We’re open to new ideas,” the secretary said Nov. 18 in announcing the program. But she stressed that to be eligible states must meet certain “bright-line principles” of the law.
In particular, any model still must ensure that all students will score at the proficient level or higher in reading and mathematics by 2013-14, as the nearly 4-year-old law requires. The model must continue to hold schools and districts accountable for the performance of subgroups of youngsters, including those who are poor, speak limited English, have disabilities, or come from racial or ethnic minorities.
“We’re not just looking for any level of improvement,” Ms. Spellings said. “A successful growth model under No Child Left Behind must put all students on track to be on grade level by 2014. That means when a student is behind, one year of progress for every year of instruction is not enough to close the gap. We will expect more.”
State superintendents in states such as California and Oregon have already expressed interest in applying for the pilot. But some wondered how many states would be able to take advantage of the new flexibility, given the requirements for participation and the short timeline for submitting plans. For example, states must have at least two years of testing data in each of grades 3-8 and high school, a prerequisite that disqualifies the 23 states that are adding new tests in some of those grades for the first time this school year, as required under the federal law. (“State-Test Programs Mushroom as NCLB Mandate Kicks In,” this issue.)
A state’s testing system also must have received approval under the federal government’s peer-review process for the 2005-06 school year. While about half the states have undergone peer reviews so far, only six have received even partial approval for their testing systems, and none has received final approval. (“Federal Review Puts State Tests Under Scrutiny,” this issue.)
“Very few states are going to be capable of complying with all the requirements that the department is going to impose on them,” predicted Jack Jennings, the president of the Center on Education Policy, a Washington-based organization that tracks implementation of the federal law.
More Rigor Urged
Ms. Spellings noted that while many states may not yet have the testing or data systems in place to meet the requirements for the pilot, her department currently allows states to use a so-called index to calculate AYP that gives schools credit for student-performance gains at levels below proficiency. Nine states now use index models under the federal law, she said, and she encouraged other states to consider doing so. Kentucky, for example, has asked to amend its accountability plan to use an indexing system.
Even so, some groups expressed worries that even 10 states would be too many for a pilot program. The Citizens’ Commission recommended limiting the experiment to five states and, in some cases, encouraging or requiring a state to carry out its growth-model experiment only in one or a limited number of districts “to ensure quality control.”
“The commission believes 10 states is far too many states for an accountability experiment that is not authorized in the statute and that carries a number of risks of failure,” the group said in its Nov. 18 statement. Among other steps, it recommended that the Education Department contract for a third-party evaluation of the experiment, with interim and final reports available to Congress as it debates the next reauthorization of the federal law, scheduled for 2007.
One of the major debates expected to come up during the reauthorization is whether to rewrite the law to explicitly permit the use of growth models.
The Education Trust, a Washington-based research and advocacy group, similarly called on the department to limit the pilot program to only a few states, and to make all state programs open to public scrutiny and third-party researchers “to objectively analyze whether these growth models work.”
Others questioned how much relief the pilot program would offer to schools and districts if the 2013-14 deadline for getting all students to proficiency remains intact. “That will render most growth targets unreachable,” said Scott Marion, the associate director of the Dover, N.H.-based Center for Assessment, which works with states and districts to improve their testing-and-accountability systems.
Brian Gong, the director of the center, said that many of the fundamental design decisions that states must make in crafting growth models still await clarification from the Education Department. Those questions include what level of expected growth will be considered good enough for schools to make their annual performance targets, and how measures of growth can be included as a component of making AYP, along with the current mechanisms spelled out in the law.
“I think it’s a significant step forward,” Mr. Gong said of the pilot program, “and the details, we’ll have to see how those play out.”
“I am glad to see the federal government’s recognition of the validity of a growth model to measure improved student achievement,” said California schools Superintendent Jack O’Connell, one of more than a dozen state school chiefs who signed a letter in March 2004 urging the federal government to permit the use of such models under the No Child Left Behind law.
Secretary Spellings told the chiefs meeting in Richmond that the pilot would provide information about whether growth models are useful policy tools.
“If it doesn’t work,” she said, “we’ll go back to the old way of doing business.”
Associate Editor David J. Hoff reported from Richmond, Va.