Corrected: A previous version of this story incorrectly identified Alabama as one of the states that is a candidate for the growth-model pilot for the 2006-07 school year.
The Department of 91制片厂视频 is preparing to take another small step in its experiment in evaluating schools based on individual students鈥 academic growth, while state officials say they are working toward the day when that approach is commonplace.
The department plans to announce, possibly as early as next week, that one or more states will join its pilot program allowing states to monitor the gains of individual students鈥攁s opposed to the performance of different cohorts of students鈥攚hen making accountability decisions under the No Child Left Behind Act.
Secretary of 91制片厂视频 Margaret Spellings will choose from among five states that reapplied for the growth-model pilot program after being rejected earlier this year. At the same time, the 91制片厂视频 Department will begin reviewing the applications of seven other states that arrived for the second round of the program.
Experts say that states are improving their ability to track individual growth, and they predict that a total of 10 states will be approved to use growth models for accountability decisions under the federal school law for the current school year.
鈥淭he capacity is growing quite rapidly,鈥 said Eric A. Hanushek, a senior fellow at the Hoover Institution, a think tank based at Stanford University, and the chairman of the federal panel that reviewed the first round of applications as well as the revised ones. 鈥淚f you ask how many states can do it today, it鈥檚 more than half.鈥
But the other states could develop that capacity 鈥渇airly quickly,鈥 he added.
And state officials believe that all 50 states will be able to use growth models when Congress reauthorizes the almost 5-year-old NCLB law, a process scheduled to begin early next year. In its recent recommendations on changes to the law, the Council of Chief State School Officers said it should be revised so that its accountability system is based on individual student growth.
鈥淚 believe that will be a reality for most, if not all states, in the near term,鈥 said Valerie A. Woodruff, Delaware鈥檚 secretary of education and the president of the Washington-based CCSSO.
Status vs. Growth
Secretary Spellings announced last year that she would approve as many as 10 states to participate in the growth-model pilot. (鈥淯.S. to Pilot New Gauge of 鈥楪rowth鈥,鈥 Nov. 30, 2005.)
Under growth models, states track each student鈥檚 academic progress from one year to the next and reward or sanction schools and districts based on their ability to improve students鈥 achievement.
The NCLB law鈥檚 current accountability system determines whether a school or district made adequate yearly progress, or AYP, based on the aggregate test scores of students in grades 3-8 and once in high school. To make AYP, the enrollment of every grade in a school or district must must meet performance targets, and so must certain demographic and racial subgroups. That system compares the scores of different cohorts of students, meaning AYP decisions are based on whether one group of 3rd graders at a school performed better than the cohort of 3rd graders from the previous year.
School officials say they prefer growth models because they believe such a approach accurately reflects their success in improving individuals鈥 achievement. Results under the so-called status model, the current NCLB system, can be skewed by sudden changes in demographics and other factors out of educators鈥 control.
The Department of 91制片厂视频 has said it will approve as many as 10 states to experiment with accountability models measuring individual students鈥 academic growth under the No Child Left Behind Act.
States approved starting with 2005-06 school year:
鈥 North Carolina
鈥 Tennessee
Candidates for approval starting with 2006-07 school year:
鈥 Alaska
鈥 Arkansas
鈥 Delaware
鈥 Florida
鈥 Oregon
SOURCE: U.S. Department of 91制片厂视频
In May, the 91制片厂视频 Department approved only North Carolina and Tennessee to participate in the growth-model pilot, and both states released AYP results for the 2005-06 school year based on that model.
Also in May, the department rejected the applications of six other states. Five of them revised their applications to address criticisms offered by the review panel led by Mr. Hanushek.
In its revised application, for example, Delaware increased the size of a test-score increase required for a student to count as making adequate academic progress.
The state鈥檚 original proposal gave reviewers 鈥渁 little heartburn,鈥 said Ms. Woodruff, the Delaware schools chief, because it would have credited schools and districts for ensuring that a student stayed at the same performance level from one grade to the next, even if the student stayed at the level labeled 鈥渂elow the standard鈥 over that period.
Ms. Woodruff said Delaware argued that students demonstrate that they have progressed a year academically when they score at the same achievement level for two consecutive years. But the federal department鈥檚 peer-review panel said that students scoring in the 鈥渂elow the standard鈥 level on the state鈥檚 tests must move up one category to a level of proficiency for them to be considered as making adequate progress.
鈥淲e conceded that and moved forward,鈥 Ms. Woodruff said.
Ms. Woodruff declined to say whether she had heard the status of Delaware鈥檚 application for the growth-model program. But, she added, 鈥渨e look forward to a formal announcement.鈥
The 91制片厂视频 Department is likely to announce the additional states next week, said Chad Colby, a department spokesman.
The other states under consideration to be added to the growth-model pilot are Alaska, Arkansas, Florida, and Oregon.
States Getting Ready
Although the 91制片厂视频 Department has set a high standard for approving growth-model plans under the No Child Left Behind law, Ms. Woodruff and Mr. Hanushek suggested that states will be ready to do more with growth models when Congress renews the law. Congress is scheduled to reauthorize the law next year, but the final changes could be delayed until 2008 or even later.
States are quickly adding student-identification codes to their data systems鈥攁n analysis tool that can track a student鈥檚 educational progress much the same way an adult鈥檚 Social Security number tallies lifetime earnings.
Also, all states鈥 assessment systems have been in place for at least two consecutive years to meet the federal law鈥檚 testing requirements and are unlikely to be changed in ways that would disrupt the annual data needed to follow a student鈥檚 career over time, Mr. Hanushek said. Without such continuity, states wouldn鈥檛 have the ability to assure an individual student鈥檚 achievement was progressing.
鈥淥n average, the states improved quite a bit in terms of how their plan would interact with the [growth-model] program,鈥 Mr. Hanushek said. 鈥淪tates are developing a fair amount of capacity to deal with these things.鈥