English Skills Still the Key in Test Scores

California students who are not proficient in English improved their scores on the Stanford 9 standardized test at about the same rate as their fluent classmates, but new state data released Monday continue to show an immense disparity between the two groups.

Broken out by fluency, the test results highlight the stark reality that there are two distinct levels of learning and achievement in the state. They also raise knotty questions for policymakers about how to improve academic performance in a state where schools must serve children from more than 80 different language groups and cultures.

John Mockler, Gov. Gray Davis’ interim secretary of education, said the results “demonstrate that we’re making progress . . . but also show us we have a long way to go to close the gap.” He added that the scores justify the Legislature’s decision to allocate $260 million for intensive English instruction after school and during summer school, on top of funding to train 15,000 teachers in the teaching of English to language learners.

In July, the California Department of Education released statewide results showing small to significant gains on Stanford 9 scores, depending on the grade and the subject. The strongest improvement came in the early grades, where the state has concentrated most of its four-year effort to halt the decline in its public schools. Overall, however, California students scored below the national average in most subjects and grades.

When the scores of students with limited English are excluded, the picture brightens considerably. Students who speak and read English well continue to score at or above the national average in almost all subjects and at almost all grade levels, said Delaine Eastin, state superintendent of public instruction. Yet their immigrant counterparts struggle, particularly, and predictably, in reading.

For example, 25% of second-graders who are still learning English scored at or above the 50th percentile, the national average, on the reading portion of the 2000 Stanford 9 test. By contrast, 61% of second-grade students who speak English fluently scored at or above the national average in reading.

The Stanford 9, a basic skills test published by Harcourt Educational Measurement of San Antonio, is graded against a national sample. Whereas about 25% of California’s students have limited fluency in English, only 1.8% of the national sample are still learning English. As a result, many educators have argued unsuccessfully that students should not have to take the test until they are proficient.

Other subgroup data showed few surprises. Females generally scored higher than males, except in science and history/social science, and students from more affluent homes tended to fare far better than economically disadvantaged students.

For the first time this year, the state by law must break out data on test performance by race. Doug Stone, a spokesman for the California Department of Education, said the state expects to release that data next month.

The Stanford 9 stakes are higher than ever. Progress, including that of subgroups such as the vast pool of students who are still learning English, will be used as the basis for rewarding schools and teachers. Hundreds of millions of dollars have been set aside as part of Davis’ highly touted plan to hold schools and students accountable for academic improvement. Schools that do not meet specified targets for improvement are subject to interventions by the state.

In the 711,000-student Los Angeles Unified School District, which is about 70% Latino, both English learners and fluent speakers had scores well below those of their counterparts statewide. In Los Angeles County as a whole, students in both categories also trailed their counterparts statewide.

Ron Unz, a Silicon Valley entrepreneur who in 1998 spearheaded a move to end most bilingual education programs in California, said the improving scores indicated that his Proposition 227 was far from the disaster that opponents had feared.

In districts that have embraced English immersion, he said, the scores of students from immigrant homes, while low, have risen much more dramatically than in Los Angeles Unified, San Jose Unified and Vista Unified, where vestiges of bilingual education remain.

Still, he acknowledged that no single education reform can be credited for the improvements. He noted that the state has also shoved aside “fuzzy math” and whole language instruction in favor of programs that emphasize basic arithmetic skills and phonics.

Intensified teacher training and the shrinking of class size in the primary grades, meanwhile, are also having positive effects. Tougher to evaluate are the effects of students’ growing familiarity with the Stanford 9 and some teachers’ tendency to “teach to the test.”

“The issue is so complex,” said Jamal Abedi, director of technical projects with the National Center for Research on Evaluation, Standards and Student Testing at UCLA. He urged the state to seek out researchers to begin the arduous job of discerning which changes are accounting for improvement--and to what degree.

James R. Brown, superintendent of Glendale Unified School District, suggested that districts will find it difficult to whittle down the language achievement gap.

Once a student’s score hits the 36th percentile on the Stanford 9, that pupil--and his or her relatively higher score--is moved out of the “language learner” category, Brown said. Meanwhile, students new to the district--and to English--move in, and the process begins anew.

“It’s sort of like chasing your tail,” Brown said.

Times education writer Duke Helfand and data analyst Sandra Poindexter contributed to this story.

(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)

Statewide Stanford 9 Results

New state data show the immense disparity in test scores between students who are proficient in English and those who are not.

Percentage of students proficient in English who scored at or above the 50th percentile:

*--*

READING MATHEMATICS Grade 1999 2000 Change 1999 2000 Change 2 56% 61% +5 57% 65% +8 3 53% 57% +4 56% 65% +9 4 53% 56% +3 52% 60% +8 5 53% 55% +2 53% 59% +6 6 54% 55% +1 59% 63% +4 7 54% 55% +1 52% 55% +3 8 57% 58% +1 52% 55% +3 9 41% 41% 0 55% 57% +2 10 39% 39% 0 49% 50% +1 11 41% 41% 0 50% 51% +1

*--*

Percentage of students with limited English proficiency who scored at or above the 50th percentile:

*--*

READING MATHEMATICS Grade 1999 2000 Change 1999 2000 Change 2 19% 25% +6 33% 40% +7 3 12% 15% +3 28% 37% +9 4 11% 13% +2 21% 27% +6 5 9% 10% +1 19% 24% +5 6 9% 10% +1 22% 26% +4 7 7% 9% +2 16% 19% +3 8 8% 9% +1 15% 17% +2 9 3% 4% +1 19% 21% +2 10 3% 3% 0 20% 21% +1 11 4% 4% 0 22% 23% +1

*--*