Advertisement
YOU ARE HERE: LAT HomeCollections

California and the West

State Rewards for Schools Are Delayed

Education: Flawed data on participation rates on Stanford 9 tests proves a hurdle for achievement bonus payouts.

September 13, 2000|MARTHA GROVES | TIMES EDUCATION WRITER

Flawed and inconsistent data are creating headaches for state education officials as they attempt to sort out which schools will qualify for big payouts based on improvements in Stanford 9 test scores.

The state had hoped to issue the eagerly awaited results on Sept. 26 but has pushed the date to Oct. 4.

The problems raise questions about the reliability of the state's fledgling accountability program, which determines how the state will distribute hundreds of millions of dollars in rewards.

To qualify for the bounty under the Academic Performance Index, schools must show gains for students overall and in various subgroups, such as students from low-income homes. Schools also must demonstrate that they tested a certain percentage of their student body.

State data crunchers have spotted widespread problems in the way some of those numbers have been reported. A top education official said the situation reveals that many schools have failed to develop adequate methods to track their students.

"The governor believes people need to get better at helping kids achieve. Sloppy systems don't help that," said John Mockler, Gov. Gray Davis' interim education secretary.

Davis and lawmakers recently authorized more than $900 million in bonuses, scholarships and other rewards as part of the state's ambitious testing and accountability program.

So far, the system is based strictly on the Stanford 9, a basic skills test that has been given the last three years to public school students in grades two through 11. The money is intended to reward schools that show sizable improvements in student performance on the exam.

William L. Padia, director of the office of policy and evaluation for the California Department of Education, said that nearly 1,000 of 7,000 schools still have at least one troublesome figure on their reports. Most are small to medium-size schools.

Padia added that the department had decided to give districts one more opportunity to fix the data with test publisher Harcourt Educational Measurement in San Antonio, Texas, although "they've had the chance to correct this since January 2000."

Errors or inconsistencies, Padia said, cropped up in statistics on ethnic subgroups and family income. There also were problems with reports indicating which students were new to a district, and therefore should not be counted in the Academic Performance Index.

The state also flagged many discrepancies in numbers that reported what percentage of a school's students in grades two through 11 took the test. To qualify for rewards, elementary and middle schools must have tested 95% of eligible students; for high schools the number sinks to 90%.

State officials have indicated that some schools will be disappointed to learn that, although they raised students' scores adequately, they will not qualify for rewards because they tested too few pupils.

Attempting to determine individual schools' "participation rates" is proving arduous.

For example, the "official" school summary reports, prepared by Harcourt and posted on the state Department of Education's test-program Web site (http://star.cde.ca.gov), contain misleading figures that make it impossible in many cases to reliably compute the percentage of students who took the test.

Consider Beverly Hills High School. Data on the Web site indicate that 58% of students at the school took the Stanford 9. Yet Principal Ben Bushman said that more than 1,000 unneeded test copies were ordered accidentally from Harcourt. Those extras got lumped into the "total enrollment reported" figure and skewed the results. The actual participation rate was about 96%, Bushman said.

Further complicating the matter has been confusion at some schools over how to account for students who were granted waivers and did not take the Stanford 9.

Those students will not count against schools in computing participation rates. However, if a school qualifies for rewards by raising scores significantly, it would not receive a $150 per-student bonus for any students who were exempted.

Padia said the state plans to release Academic Performance Index results in stages to make them easier to grasp for schools, parents, students and teachers.

The initial round, issued in January, provided the state's first ranking of public schools and established growth targets that schools would have to meet to qualify for rewards.

Schools that fall short of those targets or slip backward could be subject to state interventions that in extreme cases could include reassignment of top personnel or state takeover. Others could reap rewards ranging from $150 per student to bonuses of as much as $25,000 for individual teachers at schools with extraordinary gains.

The Oct. 4 results will indicate which schools met their growth targets. Come January, the state will unveil new growth targets based on an Academic Performance Index with a new component: scores from the reading and language arts portions of the Stanford 9 that are geared to state standards.

Reward money won't begin to flow until January at the earliest, Padia said.

Advertisement
Los Angeles Times Articles
|
|
|