California education officials are revising statewide rankings that compare the academic performance of schools that share similar demographics, after learning that more than half the state’s 8,000 schools supplied incorrect data on student poverty.
The “similar schools” rankings, which were designed to weigh schools’ results on state tests against those of schools with comparable numbers of low-income students, were recently removed from the education department’s site on the World Wide Web. Department officials say they expect to post revised rankings in April.
A separate state rating system that ranks schools based strictly on how their students fared on the 1999 Stanford Achievement Test-9th Edition— regardless of the schools’ demographic make-ups—remains unchanged.
After nearly 400 schools complained that the similar-schools rankings did not reflect their true demographics, the education department asked all schools in February to either confirm or recalculate the number of students they served who received free or reduced-price lunches. Some 4,300 schools, more than 10 times the initial number, responded that their actual number of students receiving subsidized lunches differed from the figure they had reported to the state in March of last year.
As a result, the reported percentage of such students surged from 33.6 percent to 47.6 percent statewide. Nationally, slightly less than a third of students received federally subsidized lunches in 1999, federal officials say.
“Nobody realized at the time that this data was initially reported that this information would be used in any kind of a high-stakes accountability system,” said William Padia, the director of the California education department’s office of policy and evaluation. “As a result, they took a very casual attitude in filling this in.”
Broader Issues Seen
Some observers say the mix-up points to a larger problem associated with basing high-stakes rankings on the limited school-level data at the state’s fingertips.
“Any attempt to compare schools with other schools is only as good as the indicators you have,” said Gerald C. Hayward, a co-director of Policy Analysis for California Education, a think tank based at Stanford University and the University of California, Berkeley. “This just underscores how poor our data-reporting system is.”
State officials have long said the academic-performance index—which includes both types of ranking systems—will eventually incorporate such factors as attendance and graduation rates. But they say it will take some time, up to three to five years, before they can accurately gather and report such figures statewide. (“Calif. Schools Get Rankings Based on Tests,” Feb. 2, 2000.)
The state is giving schools a greater incentive to report accurate information when they administer state tests this spring. While schools will receive a data-reporting allocation of $2.52 per student, they could be required to refund up to $1.25 per student if the state discovers that any of the information provided is inaccurate.
A version of this article appeared in the March 22, 2000 edition of Education Week as Calif. Will Redo School Rankings To Correct Errors in Poverty Data