Los Angeles Times

Wednesday, July 22, 1998

Success in Any Language
Experts seek to explain how certain groups of bilingual students in L.A. County and elsewhere outscored English-only pupils on Stanford 9 tests.
By NICK ANDERSON, Times Staff Writer

In many public schools in Los Angeles County and other urban areas of California, select groups of bilingual students outscored their "English only" peers on this year's state tests of reading and math.
     Some of the better scores in these school districts were turned in by two types of students: those who entered their school systems already fluent in English, even though it was not their native language, and those who became fluent while in school, advancing from the ranks of the "limited English proficient."
     Consider a sampling of scores on the test of fourth-grade reading skills:
     * In Los Angeles Unified School District, those who had become fluent after having been branded "limited English" scored at the 49th percentile. That's much higher than the 35th percentile average of students who came from families that speak only English.
     * Similar gaps were found in San Jose Unified School District (the 64th percentile compared to the 58th), San Francisco Unified (66th compared to 49th) and Garden Grove Unified in Orange County (71st compared to 48th).
     The tantalizing findings are drawing notice among educators as a counterpoint to the extremely low marks posted in many school districts by students who are not fluent in English.
     "It's counter to the common stereotype that we're not doing anything good with these children," said Joan Herman, associate director of the Center for Research on Evaluation Standards and Student Testing at UCLA.
     But the relative success of the two types of bilingual students, especially in urban areas, is largely predictable--just as is the poor showing of those who speak limited English.
     One reason is pure statistics. The very criteria used to classify these students as fluent in English guarantees that they are capable in the classroom.
     How do they win the "fluent" designation? Often by taking--and scoring reasonably well on--standardized tests.
     Many school systems require such students to score at the 36th percentile in language skills to be declared fluent. In Los Angeles Unified, that's considerably higher than the districts' overall average.
     Educators suggest the "successful" bilingual students are likely to come from households with advantages not shared by many classmates. Certainly, students entering the school system fluent in two languages are likely to have had help from mom and dad. They may live in homes that can afford books, newspaper subscriptions and the like.
     "I'd like to know who these children are," said Linda Kaminski, testing coordinator for Glendale Unified School District, referring to the success stories.      "One possibility is that they are students who have access to more language, live in neighborhoods where English is spoken more, speak more with their peers in English. These are unknowns."
     That they perform better than their English-only peers in some districts may reflect more on those other students--and the social and economic problems they face.
     Indeed, when you look beyond select urban districts, the "successful" bilingual students--those who have made the transition to English fluency at home or at school--no longer outpace English-only students the same way.
     New statewide data obtained by The Times show that the gap between the successful bilingual students and the English-only students virtually disappears when all of California's students are averaged together. The students whom schools had successfully taught English as a second language scored on the fourth grade reading test at the 51st percentile and the English-only students at the 50th.
     Both were right at the national average, in other words.
     More than 4.1 million California students from grades two to 11 took the Stanford 9 exams this past spring, making it the most ambitious testing effort in state history.
     Tests in grades two to eight covered reading, written expression, math and spelling. In grades nine to 11, they covered reading, written expression, math, science and social sciences.
     The tests were given only in English, and virtually all students were supposed to take them, no matter what their language abilities. Parents could request that their children be exempted, however, and some activist groups encouraged limited-English students to use the waivers to avoid taking the test.
     Still, state officials hope the scores will shed some light on how California schools are doing in their struggle to teach English to 1.4 million students who are not fluent. That issue has been magnified since June 2, when voters approved Proposition 227, the initiative to replace most bilingual education programs in the state with English-immersion classes.
     Later this summer, officials plan to analyze the scores by ethnicity, which could reveal differing patterns of achievement by various groups. The analysis may also indicate how much the encouraging performance of the English-fluent bilingual students is influenced by the well-documented success of Asian Americans on standardized tests.
     Though most of the attention in public debates over bilingual education has focused on limited-English students, there are a large number who go on to achieve fluency.
     In Los Angeles County, about 11% of the test-takers were students who entered the school system fluent in both English and another language. They generally did better on the Stanford 9 than children who grew up speaking only English. That was the case in every grade in math and in all but the high school grades in reading.
     Another 10% of the test-takers in the county were students who were reclassified as fluent in English during their school years. They too bettered or equaled their English-only counterparts in nearly every grade in math. In reading, they repeated that performance in grades two through five.
     English-only students made up about 40% of the county's test-takers and limited-English students about 27%. The language ability of the other 12% was unknown.
     In Los Angeles Unified, the nation's second-largest school district with 681,000 students, reading scores for the successful bilingual students were better or on par with those of English-only students in most grades.
     But percentile rankings for the district's limited-English children lagged badly, falling into the single digits in high school reading.
     Many of the district's limited-English students had been in bilingual classes at some point in their schooling. Those classes emphasize teaching students in their native language in the early grades, most often in Spanish, and gradually phase into instruction in English.
     Even so, it's hard to quickly draw conclusions from the test data. Both opponents and advocates of bilingual education--usually quick to put their own spin on events--have been cautious in this case.
     Reynaldo Macias, an education professor at UC Santa Barbara, said that a promising showing by some bilingual students might indicate that bilingual education is working well--"assuming that those kids were in prior bilingual programs."
     But that, too, is up in the air. While three of 10 limited-English children in California are in bilingual education each year, the other 70% are either in programs that already stress English immersion or are students who get little special attention at all in learning English--they're left to sink or swim on their own.
     In San Francisco, one researcher this year found that students who had gained English fluency through the school system not only tended to outscore their English-only peers on standardized tests but also had better grades and attendance rates. J. David Ramirez, a professor of education at Cal State Long Beach, suggested that "the cognitive benefits of multilingualism" explained the trend in his study, which was based on the records of 12,000 students from 1994-95 to 1996-97.
     Most school officials say they will have to take apart the new test data school by school, class by class--perhaps even student by student--before they can draw conclusions.
     Nevertheless, Brad Sales, a spokesman for Los Angeles Unified, called the pattern of scores for the successful bilingual students intriguing.
     "It is conceivable that we're doing something in the bilingual programs that is creating a boost, particularly in the early grades," Sales said. But it will take "substantial research and time," he added, "to make that determination with certainty."
Times staff writers Doug Smith and Tina Nguyen contributed to this report.


In Los Angeles County, about 11% of the test-takers were students who entered the school system fluent in both English and another language. They generally did better on the Stanford 9 than children who grew up speaking only English. That was the case in every grade in math and in all but the high school grades in reading.

Successful Bilingual Students

Results from the new statewide standardized public school tests--the Stanford 9 exams--showthat some bilingual students in urban areas outperformed their English-only counterparts in reading and math. Statewide, however, the students who were raised speaking onl English tended to do as well or better.

Scoring: All scores are expressed on a 99-point percentage scale, with 50 being the average.

Key: The charts show the scores for three categories of students:

English only: Students who speak no other language at home.

FEP: Students who speak another language at home but entered their school district as "fluent English-proficient."

R-FEP: Students who speak another language and were classified as "limited English-proficient" at first but were reclassified as fluent in English during their school yeas.

Average Scores on Stanford 9 Tests, 1998

STATEWIDE   Reading     Math   
Grade 4
     English only      50 46
     FEP 47 49
     R-FEP 51 60
Grade 8
     English only 54 51
     FEP 49 52
     R-FEP 44 48
Grade 10
     English only 40 47
     FEP 34 48
     R-FEP 29 44
 Reading     Math   
Grade 4
     English only      43 40
     FEP 46 49
     R-FEP 52 60
Grade 8
     English only 46 44
     FEP 50 52
     R-FEP 41 43
Grade 10
     English only 35 42
     FEP 33 47
     R-FEP 28 42

Test Takers

About 4.1 million students from grades 2 to 11 took the Stanford 9 tests last spring--more than 1 million in Los Angeles County alone. Only certain students in special education programs or those whose parents objected were exempt. Chart at right shows the percentage of students who took the test in Los Angeles County, by language ability.

L.A. County

English only: 40%

Unknown: 12%

R-FEP: 10%

LEP (Limited English- proficient): 27%

FEP: 11%
* * * Source: Harcourt Brace Educational Measurement reports of Stanford 9 scores.