Inaccurate info, misclassifications skew test scores

By The Associated Press

The results were released last week and are a central piece in the state's plan to turn schools around. The index will be used to gauge students' improvement and determine which schools receive cash rewards.

The state Education Department rated each of the state's nearly 7,000 public schools on their students' 1999 achievement test scores, broke down performance by various demographic factors and even noted the average education levels of parents.

The reporting glitches from some schools did not affect the overall statewide rankings, but comparisons among schools with similar characteristics were flawed because of it, many educators said.

A handful of other schools were simply left out of the ranking because they were misclassified, said Paul Warren, deputy superintendent for accountability.

Saul Martinez Elementary School in Mecca, which was ranked worst in the state, showed large percentages of parents of students with college diplomas. In fact, most of the parents at Saul Martinez never even earned a high school diploma, said Martha Tureen, assistant superintendent of Coachella Valley Unified School District.

School district said the errors stemmed from relying on children to provide accurate information about whether they came from poor families, were still learning English and if their parents went to college.

Beaumont Superintendent John Wood said that asking students, especially those as young as eight, to fill out family education and family financial information was a big mistake.

"Children that young don't have a clear idea about things like degrees and family income," he said.

Pat McCabe, senior demographer at the California Department of Education, said that despite the errors, the rankings on a statewide basis are good.

Other schools were left out entirely because of confusion over the law that created the Academic Performance Index.

It provided several categories of schools that would not be given scores, including schools with fewer than 100 students, those run by a county board of education and "alternative schools, including continuation high schools and independent study schools." Such schools, the law says, will have a separate accountability system.

Seven schools in Los Angeles Unified School District were left out of the index because they identified themselves as "alternative" in annual reports filed with the state.

Several of them have the word "alternative" in their titles, a vestige of their origin in the movement of the 1960s and 1970s to design instructional programs.

Over the years, however, they have moved back into the mainstream.

Warren said before deciding whether to include the schools in the system, the Department of Education must determine what the designation "alternative" means and whether the schools derived any financial benefits from having that status.

It may be two weeks before a decision is made about whether to include the schools in the accountability system, Warren said.

In the meantime, the state will re-compare schools using more accurate district and state data for at least 20 school districts around the state that misreported the numbers of poor students.

To make sure the similar schools rankings on the 2000 index are more accurate, districts will be charged $1.25 for each Stanford 9 test that has incomplete personal and family information. The money will be taken out of the $2.20 districts receive per test to help pay for costs, such as mailing results to parents.

For future preparation, many schools plan on using alternative ways to get certain information for the new system, such as sending questionnaires home with students and asking questions during parent conferences.

Uploaded January 31, 2000