L.A. Pupils Rank Low in Test Scores

Standardized test scores released by the Los Angeles Unified School District indicate local students rank in the bottom quarter nationally in reading and spelling and below the bottom third in language and math.

Students tested in Spanish fared slightly better in reading but worse in math, continuing a trend for bilingual students.

The Stanford Achievement Test, Ninth Edition, was given in English and, in Spanish, a test called Aprenda was given. On both exams, students had particular trouble with basics such as vocabulary, reading comprehension, capitalization, punctuation and spelling based on principles of phonics. In math, decimals and fractions proved the most difficult.

“There are no excuses. We have to do better,” said Supt. Ruben Zacarias. “Educators tend to argue over tests and we know there isn’t a test that can judge any human being, not to mention a child, but we also know we are judged by these tests.”

Zacarias spent much of last week meeting individually with principals of the 100 lowest-scoring schools, which he targeted for improvement during his campaign for the top district post. Each school is setting its own achievement goals, he said, but most are shooting to improve test scores this year by five percentage points.

“I told them, one year from now, I expect you to be the toast, not only of L.A., but of the nation,” he said.

A campus-by-campus rundown of scores at crucial grade levels-- third in elementary, seventh in middle school and 10th in high school --is on Page B3.

By region, schools that performed best on the Stanford Nine generally were in the wealthier areas: the west San Fernando Valley and on the Westside --where about half the students scored at or above the national median. Poorer neighborhoods of South-Central and Southeast Los Angeles had the largest number of struggling schools.

The Stanford Nine is a test ranked in national norms, given for the first time last spring in Los Angeles public schools. District officials believed that it more closely matched what was being taught than the Comprehensive Test of Basic Skills, which had been used for the previous 14 years. For example, while the CTBS included a page of math computation, the Stanford Nine features more word problems, mirroring the applied way math is taught in most schools today. Reading passages also were longer and more complex.

Scores on the Stanford Nine are expressed in a median percentile, indicating where the middle-scoring student fell in relationship to students across the country.

At first glance, the scores appear to have dropped significantly from last year, when Los Angeles Unified students ranked slightly above the 30th percentile nationally in language and near the 40th percentile in math. But testing experts --confirming predictions of district officials-- said the dip probably resulted from the change to a new test rather than an actual decline in ability.

Teachers, school administrators and parents who received the new test declared it harder than the old one, particularly for students for whom English is not a first language.

Based on a Times analysis of the third, seventh and 10th grades, schools with the least student turnover tended to score the highest, while those with more often fell toward the bottom.

The analysis also indicated that:

* At third grade, the highest-scoring elementary school was Eagle Rock Highly Gifted Magnet, which gathers students with high IQs from throughout the district. Of regular elementary schools, the top scorers were Topanga Elementary and Roscomare Elementary, located east of the Sepulveda Pass, while the lowest was Murchison Elementary on the Eastside.

* Among middle schools, Portola Highly Gifted Magnet in Tarzana attained the highest percentile for seventh grade and Revere Middle School in Pacific Palisades was the higest-scoring non-magnet. Drew Middle School in South Los Angeles had the worst scores.

* For high schools, the highest percentiles in 10th grade were found at North Hollywood’s Highly Gifted Magnet, followed by Granada Hills High among regular high schools. The lowest were South L.A.’s Jordan, Dorsey and Manual Arts highs.

Parents should have received results for their children, but the broader systemwide tallies were delayed until now by data errors, according to district testing officials.

The changeover to a new test makes year-to-year comparison impossible, blunting efforts to measure the progress of various district reform programs. Even school-to-school comparisons are difficult because the district has yet to produce a combined median score for each campus.

A study intended to equate the two sets of results indicated that the 1997 scores may be slightly higher than last year for elementary and middle grades, but slightly lower for high school.

For instance, the 24th percentile achieved in fourth-grade reading on the Stanford Nine would have been the equivalent of a 36th percentile score on the CTBS--up one point from last year.

But district testing officials cautioned that the equating study was simply an estimate, since no control group of students had been given both tests this year.

As a result, “this is totally unintelligible,” school board member David Tokofsky complained.

Tokofsky also expressed concern that newly tested areas--social science and science--were not reported because of delays in receiving the data in a usable form from test publisher Harcourt Brace & Co.

Although the new data is intended as a baseline by which to gauge future performance, its fate is uncertain because the State Board of Education is poised to endorse a statewide test. If the state does not select the Stanford Nine, L.A. Unified could continue to use it, but there would be no state reimbursement.

At this year’s reimbursement rate of $5 per student, the state will offset about half the $3.7-million cost of developing and giving the exam.