Test scores released Thursday paint a picture of public school student achievement in California that is at once surprisingly bright and dauntingly grim, with those who speak English doing well compared with national averages and those who do not continuing to be severely outmatched.
Scores in reading, math, spelling and other subjects from the second annual administration of the Stanford 9 test show a deep and predictable gulf between those who speak English fluently and those who do not.
A cloud continued to hover over the results, however, as state officials reiterated that language-fluency designations are missing for nearly 420,000 students, or about 10% of test-takers. State officials said they won’t know for at least a week the effect those scores might have on the reported gains. Full scores are available on the California Department of Education Web site at www.cde.ca.gov.
The results–delayed for three weeks because of scoring and reporting flubs by the test publisher, Harcourt Educational Measurement–dispelled earlier optimism that a shift in state philosophy to emphasize English as the language of instruction would provide a quick fix for the shaky academic performance of the state’s hefty immigrant population.
Much more than school pride is at stake. Unless the scores improve dramatically next year, thousands of individuals could be held back as part of the state’s effort to end the practice of promoting students to the next grade regardless of academic readiness.
Also, beginning next year, test scores will be the key component of a new plan to hold administrators and teachers accountable for students’ progress.
“Everybody knows there is no magic bullet,” said state Supt. of Public Instruction Delaine Eastin. “We had districts where there was fantastic improvement . . . but it doesn’t happen overnight. There’s a huge amount of effort that goes into teaching kids.”
She said the billions of dollars the state has spent in recent years on smaller class sizes, more textbooks, more technology and other improvements is beginning to pay off. But she said the state is still spending about $1,000 less per pupil than the national average of $6,335.
The new statewide data show that the 73% of test-takers who were fluent in English rank above the average of a national comparison group in nearly every grade in reading, math and language.
The only particularly weak spots were in spelling, which has been neglected until recently in many California schools, and in high school reading, which state officials suspect might be related to wrinkles in the test design or norming samples rather than to students’ skills.
Interpreting the Results
Eastin and other education observers cautioned against using the scores to drive any dramatic adjustments in policy in California, where the convergence of half a dozen major school reforms has created a chaotic backdrop to standardized testing.
In particular, she was leery of drawing any conclusions about the effects of the voter-approved Proposition 227, which dismantled most bilingual education programs in the state.
“It’s too early to celebrate or throw stones,” she said. “I wouldn’t rush to any conclusions one way or the other.”
State Education Secretary Gary K. Hart said the stronger-than-expected scores for English-speaking students demonstrate that many California students are doing well, although overall results show California students as a group scoring below the 50th percentile in almost every subject across the board.
“It will give those people who take pride in trashing the public schools pause, that they need to be more judicious and careful,” he said. “But I don’t think this should reduce our resolve that there are some serious problems out there that need to be dealt with.”
He said the scores for students who are not fluent in English highlight the need to work harder to boost their performance. The new state budget includes $50 million for additional after-school and weekend classes for non-fluent students in fourth through eighth grades; an additional $10 million will go for training seminars next summer for teachers with significant numbers of non-fluent students.
Officials in the Los Angeles Unified School District, the state’s largest, were enthusiastic about gains of 3 to 4 points made by English-speaking students in reading, which they said reflected improved teaching. But nearly half the district’s 700,000 students are not fluent in English, and those students made only tiny gains, which kept overall scores in the bottom quarter nationally.
“I’m particularly gratified that our reading scores in the first few grades are improving,” said Los Angeles Unified Supt. Ruben Zacarias. “Early literacy was our instructional focus this year and it had the impact we envisioned.”
The statewide data include scores for most schools and districts, supplementing the overall state numbers that the California Department of Education released on schedule June 30. In the intervening weeks, several snafus came to light.
First, Harcourt acknowledged that it had mistakenly lumped the scores of about 257,000 students who had been re-designated as being fluent in English with the results of youngsters still learning English. That blunder artificially inflated the results of English learners in many school districts.
Next it was revealed that Harcourt had applied incorrect national norming samples to 44 districts with year-round schools, tainting those data. As of Thursday, that and other problems continued to plague the scores from a limited number of year-round schools in three districts–Long Beach, Glendale and Ontario/Montclair in San Bernardino County.
The problems with the year-round scores were discovered in the Long Beach Unified School District by a suspicious teacher. In the process of fixing that error, others were discovered.
Lynn Winters, an assistant superintendent in Long Beach, said Thursday that the district had calculated correct scores showing increases in every grade level and every subject.
Still, she said, the problems reinforce the idea that judgments about schools should not be made based on test scores alone.
Taking the Stanford 9 involves “seven hours out of the 1,000 hours a child spends in school in a year,” she said. “The measurement of any human characteristic . . . is not precise. There are errors and they always need to be interpreted in the larger context of what we know about individual children and schools.”
Meanwhile, the state realized last week that, for a reason still to be determined, 420,000 students’ test forms did not specify whether the students spoke English fluently or not. That blunder clouds the state’s efforts to break down results by language ability.
Brad Sales, a spokesman for Los Angeles Unified, said only about 15,000 of the district’s test-takers could not be categorized. That meant that the bulk of the 420,000 students with missing information came from elsewhere.
Despite all the caveats, a new coalition of business leaders said the state must stick with the testing program.
“California now has a consistent, annual measure of how every student is improving,” said Bill Hauck, chairman of the California Business for Education Excellence coalition. “An apples-to-apples, year-to-year comparison is an invaluable gauge of the health of our public schools.”
Large Gains in Early Grades
Scores were up sharply in the early grades in several of the state’s large cities. In San Diego, for example, third-grade scores went up 8 points in reading and 11 points in math. First-grade scores also were up in Sacramento, which launched a phonics-based reading program and assigned coaches to schools to help train teachers.
In San Francisco, however, scores in the elementary grades were mostly flat in reading, though they were up slightly in math.
Although the numbers are up across the board from last year, some experts discounted that as a sign of improvement, saying scores typically rise as students and teachers become more familiar with the test.
Students not fluent in English made the most progress in math and reading in second and third grades, but scores were up at least slightly in every grade and every subject. The overall changes were slight, making it impossible to gauge the effect of Proposition 227.
Opponents of that measure had predicted it would wreak educational disaster; proponents, meanwhile, promised that the end of bilingual education, if achieved, would lead to improvements.
The Mexican American Legal Defense and Educational Fund, a vocal opponent of the measure, said the scores “do not speak for or against any particular educational method” for students with limited English skills, but rather demonstrate the complexity of evaluating the effectiveness of California’s varied programs.
Among the trends in test scores:
* Girls scored higher–often substantially–than boys in reading and language in every grade and were at least on par with them in math in every grade except second and 11th.
* Reinforcing well-established patterns, low-income students ranked 20 to 30 percentile points below their more fortunate peers, depending on grade and subject.