Schools May Get Help on API Quirk

A state lawmaker said Tuesday that she will ask for an investigation into a statistical anomaly that allows school scores on the Academic Performance Index to go down even if all the groups of students within the school improved their performance.

Most of these schools--about 70 statewide, including seven in Orange County--had significant demographic changes from one year to the next that explained their declining overall scores. But the schools were nevertheless told they would be ineligible for the massive pot of financial rewards handed out to improving schools.

Testing experts were taken aback by the problem Tuesday.

“I’ll be darned. I don’t think anyone thought about that one,” said Jerry Hayward, co-director of policy analysis for California Education, a think tank run by UC Berkeley and Stanford University. “This is clearly something the state ought to take a hard look at.”

Officials at Ball Junior High School in Anaheim agree. Teachers and administrators there were convinced they would be eligible for rewards, which could include $25,000 checks to teachers. Both Latino and white students, the only two statistically significant groups at the school, exceeded their state-set benchmarks for improvement.

But two weeks ago, Ben Carpenter, principal at Ball, learned that his school was not eligible because the school’s overall score had gone down. Worse, Carpenter found out that Ball had been placed on a list of underperforming campuses that faced state takeover if scores did not come up within three years.

Changing demographics were the cause. The population of white students went down by 24 students, while that of Latino students went up by 103. Latino students--who might not speak English and often come from poorer and less-educated families--typically score much lower on standardized tests. In the case of Ball, Latino students scored about 200 points lower than white students.

State officials were unmoved by Ball’s protests. They said that the school officials had had the option of exempting their campus from this year’s API and noted that some schools with demographic changes still managed to boost their scores enough to qualify for rewards.

State Sen. Dede Alpert (D-Coronado), who sponsored the legislation that created the API, said Tuesday that she had been unaware that the potential for such a problem existed and that it merited further investigation.

“When people have made efforts and they have improvement, how sad that this would be the result,” Alpert said. “These are the kinds of glitches we have to work out. . . . Not only is there prestige, there’s a lot of money on the line [for schools that improve].”

Alpert said she would ask a committee that advises the State Board of Education on API-related matters to investigate the problem and might propose legislation to allow schools to appeal their ranking.

The API, a complicated new state program now in its second year, is designed to improve education in California’s struggling schools, making them more accountable by giving them specific growth targets to reach each year. Schools that achieve their growth targets are rewarded with more money; schools that fail to meet their targets could face sanctions.

For now, API scores are based solely on the Stanford 9 test, but eventually they will include graduation rates, attendance figures and other measures.

Each fall, schools are graded both on how well their students’ score overall on the Stanford 9 standardized test and how much they improve from year to year. In order to make sure teachers concentrate on helping all students equally, schools are also assigned a separate score for each subgroup within the school that makes up at least 20% of the student body.

Six other schools in Orange County, including two other junior high schools in the Anaheim Union High School District--Sycamore and South--were the victims of similar but less dramatic statistical anomalies. The other schools were: College Park Elementary in the Newport-Mesa district, Pyles Elementary in the Magnolia District, Utt Middle School in the Tustin district and Lincoln Elementary in the Santa Ana district.

Officials across Anaheim said they believed the situation was extremely unfair, not to mention confusing.

“It’s tough to understand how something like this could happen,” said David Steinle, assistant superintendent for the Anaheim Union district. He said he still did not understand how it was mathematically possible for the school’s overall API to go down when all its subgroups went up. “The question our teachers had was ‘What more can I do?’ ”

Educators also were perplexed in the Newport-Mesa Unified School District, where scores at College Park Elementary went up for every group while overall scores fell. Eventually, they realized it was because the school had eliminated the fourth grade last year, which led the white population to decrease from 105 to 51 students and the Latino population to drop from 155 to 122.

They had not wanted to complain, said Peggy Anatol, the district’s head of testing and assessment, “because it sounds like an excuse.”

“They probably never anticipated this problem,” Anatol said.

Educational experts agreed that officials had not expected the situation, but said that now that it’s come up, it must be addressed--especially because the number of Latino students is predicted to grow across the state, making such demographic changes inevitable at some schools.

“We’re not trying to open up a box of excuses,” said David Marsh, a testing expert and education professor at USC. “But you’re losing money. You’re not getting credit for the hard work you’re doing. I would think there are some ways the state could make adjustments in the calculations.”

Like many experts, Marsh noted that in a program as complicated and far-reaching as the API, glitches in the first few years are to be expected.

Carpenter, the principal at Ball, said he is just happy that there’s a chance someone will take another look at his campus.

“I had some parents call me this morning and say they know the school is still a good school,” he said. “In our district, no one could logically figure out why this happened. I hope someone will take a good look at this.”