Does bilingual education work?
Some studies say it does. Others are inconclusive.
These days, all are a migraine in the making for state schools chief Lisa Graham Keegan, who is deciding what sort of bilingual education legislation to support in 1999.
And University of Texas researcher Jay P. Greene, who recently examined the validity of 75 bilingual education studies, has no cure.
He says that trying to draw conclusions about bilingual education from most studies is like trying to figure out whether acetaminophen cures headaches by giving 500 milligrams to one study group and 10,000 to another.
“For every study that says one thing, there’s another that says, “No, no, no,’ ” Keegan said. “I’d like a study that says, “Here is where a student was in reading, writing and mathematics at this time and here is where they were at the end of the program’, and there virtually is none of that.”
Bilingual education advocates most commonly cite research by George Mason University scholars Virginia P. Collier and Wayne P. Thomas. Their studies show that non-English-speaking children can score as high on standardized tests as native English speakers — if they spend several years in classrooms that let them learn math, science and social studies in their first language.
In one Collier study, children in long-term dual language immersion classes scored an average of 11 percentage points higher on standardized tests than an English-speaking norm group.
But critics of bilingual education call Collier’s research distorted because it mixes together scores of Spanish- and English-speaking children.
In a March 1998 study, University of Texas’ Greene found that only 11 of 75 major academic studies of bilingual education were sound. He discarded the rest because they did not study children for long enough periods of time and lacked control groups.
Greene said, however, that the 11 good studies do suggest that non-English speakers who get some instruction in their best language do “significantly better” on standardized tests than non-English speakers who get no bilingual education.