Agustin Gurza accuses me (“Bilingual Ed: The Truth Behind Test Gains,” July 22) of using “statistical sleight of hand” in suggesting that the early Stanford 9 test scores indicate the superiority of English immersion over bilingual education.
Gurza is misrepresenting the facts.
First, I had emphasized that until the mid-August release of all immigrant test scores, we are merely seeing “straws in the wind,” but I was encouraged that San Juan Elementary’s English-immersion program had doubled the mean percentile scores of its immigrant students in less than two years.
To rebut my example, Gurza selects a particular school in Santa Ana, Walker Elementary, in which half the immigrant students have been kept in bilingual programs, and points out that in some particular areas, Walker’s students are doing somewhat better than those at San Juan.
But consider the 1998-2000 trends in reading, math, language and spelling even at just those two schools. Walker’s scores rose by 5, 12, 9 and 9 percentile points, while San Juan’s rose 14, 18, 16 and 16 points.
Half of Walker’s immigrant students are in English-immersion programs, compared with all of those students at San Juan, and perhaps coincidentally, the rise in Walker’s test scores was about half as large.
San Juan students began 1998 far behind Walker students but have passed them in most subjects.
Furthermore, a statewide study by the San Jose Mercury News found that after one year, students in Proposition 227-type programs had reading and math test scores 20% to 100% higher than students kept in bilingual programs.
If this continues, perhaps even Gurza will finally admit the obvious: English immersion works and bilingual education does not.