Scoring Glitch Clouds Impact of Prop. 227
It's been two weeks since the results of California's statewide achievement test were released. But the answer to a closely watched question--what those scores say about the effects of Proposition 227, the ballot initiative that curtailed bilingual education in the state's classrooms--remains unclear.
Because of a computer error, in which scores for students who are no longer considered limited-English-proficient were included with the scores of those who are still in that category, the data were returned to the test publisher, Harcourt Brace Educational Measurement.
The company is expected to provide corrected files to the state this week.
The mistake did not affect the overall results, which state schools Superintendent Delaine Eastin said "are moving in the right direction--upward bound." But, she said, the error "may have resulted in overstating the achievement of our LEP students."
The state last week estimated that between 230,000 and 250,000 students were misclassified. There are roughly 1.4 million LEP students in California.
Across the state, news conferences pegged to the release of the test results were postponed, and officials from districts such as Oceanside Unified, which appeared to post remarkable gains in scores for LEP students on the Stanford Achievement Test-9th Edition, announced that they would have to wait and see whether their results were affected by the error.
The Oceanside district, where 21 percent of the 21,500 students are LEP, is being followed closely by people on both sides of the bilingual education debate because of its strict interpretation and implementation of Proposition 227, which California voters approved last year. ("Calif.'s Year on the Bilingual Battleground," June 2, 1999.)
The law, in most cases, requires schools to teach non-English-speaking children in sheltered-English-immersion classes until they acquire a working knowledge of English, which advocates of that method say should take about a year.
Students are then to be moved into mainstream classes.
Improvements Still Expected
While some districts have given parents waivers to keep their children in bilingual classes, which the law allows, Oceanside denied most of those applications and had no bilingual classes this year.
Last month, district officials released their 1999 state scores, which showed increases as high as 475 percent over the 1998 results.
But even with Harcourt's error, supporters of Proposition 227 expect to use the test results to show that teaching LEP children almost exclusively in English is best.
"I feel very confident that the results, when corrected, will show a strong pattern of the success of Prop 227," said Ron K. Unz, a wealthy California software entrepreneur and the architect of the state initiative.
With no LEP results to analyze, Ms. Eastin and local district leaders tried to turn attention toward increases for all students, which are reported as the percentage of students scoring at or above the 50th percentile.
Mathematics scores--with increases ranging from 1 percentage point in 9th grade to 8 points in 3rd grade--and scores for students in the early grades in all subject areas improved the most.
The scores were flat for 7th, 9th, and 11th graders in reading and for 9th, 10th, and 11th graders in social science.
The Stanford-9, which students were required to take for the first time in 1998, tests students in grades 2-11 in reading, language, and mathematics. Children in grades 2-8 are also tested in spelling, and those in grades 9-11 are tested in science and social science.
Teaching Methods Criticized
Districts with the highest percentages of LEP students are eagerly awaiting the corrected test results.
Shel Erlich, a spokesman for the Los Angeles Unified School District, where almost half of the nearly 700,000 students have limited English proficiency, said the scores would help school officials determine which of two instructional models is more effective.
In Model A classes, core academic instruction is in English, with the student's primary language used only for clarification. In Model B, children are taught in both English and their primary languages.
The second method, however, is the subject of a Los Angeles County grand jury investigation into complaints that teachers "are using their positions and influence to circumvent the intent of the Proposition ."
The 23-member panel, which released its findings June 30, the same day the test scores were first made available, recommended that the district discontinue the use of Model B and acquire English-language textbooks and other materials "to assist in implementing a full program of 'structured' English-immersion instruction." The panel also said that the school board should "actively discourage" the use of waivers.
Brad Sales, a spokesman for Los Angeles Superintendent Ruben Zacarias, said he was "concerned about the implications of the recommendations."
"We feel that we are on solid legal ground," he said, adding that the district just underwent a "state compliance review" and that no concerns were raised during that process over how the district is teaching LEP children.
He also noted that district administrators are interested in finding out more about which schools the grand jury visited in preparation for its report.
"We don't know how many schools they went to or how they picked them," Mr. Sales said. "And they didn't ask for testimony from the people who designed the program."
While Proposition 227 didn't address testing, critics of bilingual education see the two issues as related and are closely monitoring a bill in the state legislature that would exempt LEP children from the state's Standardized Testing and Reporting program for their first 30 months in a California public school.
Sponsored by Assemblywoman Carole Migden, a Democrat from San Francisco, AB 144 has passed the Assembly--the legislature's lower house--and is now before the Senate appropriations committee.
Sheri Annis, a spokeswoman for English for the Children--the name of the California campaign for Proposition 227 that has since been adopted by organizers of a similar push in Arizona--said AB 144 is one of several proposals that "would allow us not to know what's going on in the classroom."
"If there is no accountability," Ms. Annis said, "there is no way to assess the new law."
But some school officials argue that it's inappropriate to test students in a language they are still learning.
"When you test children who don't understand the test, you get a false baseline," said Sandina Robbins, a spokeswoman for the San Francisco school district, which won a court battle last year when the state education department tried to force the district to test LEP students in English.
As for any increases in scores that might show up for LEP students, Ms. Robbins said, "people claim great victory, but it's still not an accurate measure of the basic skills, which is what these tests are supposed to do."
Vol. 18, Issue 42, Page 3Published in Print: July 14, 1999, as Scoring Glitch Clouds Impact of Prop. 227