More on Jill Stewart's superficial and biased writing

See the prior posting on March 15 for background.
As a long-time resident of California, I suppose that I am one of the ‘everyday Californians’ that Jill Stewart speaks for in her diatribe, “Bilingual doesn’t work in any language.” I am writing to exclude myself from this group that she claims to represent, since I take great offense at her piece, one of the worst examples of polemic disguised as journalism that I have read in a long time. This brand of journalims seems sloppy at best, irresponsible at worst. First, Ms. Stewart does not reveal any of her data sources, other than to reference “scores released by O’Connell’s office” and “national English reading tests.” Thus, it is difficult to verify Ms. Stewart’s assertions or to do an independent analysis of these data. Typically, “scores” are related to specific tests like the California English Language Development Test (CELDT), the Stanford Achievement Test-Edition 9 (SAT-9), or the California Achievement Test-Edition 6 (CAT-6).
Second, even though she gives us little upon which to verify her claims of improved English proficiency, she further errs by confusing English proficiency with improvement “in all subjects.” A comparison of the percentage of Sacramento County students scoring at or above the 50th percentile on the SAT-9 (2001 & 2002) and the CAT-6 (2003), available on the California Department of Education’s website (www.cde.ca.gov), tarnishes the rosy picture Ms. Stewart paints for English Learners. This website has information about different subgroups of students, including English Learners with less than 12 months instruction (“recently arrived”) and English Learners with 12 months of instruction or more (“established English Learners). Using Ms. Stewart’s logic, one would expect that the established English learners would consistently outscore their recently arrived English learner peers, since they were “taking English immersion” and were improving “in all subjects.”
My brief analysis included data from 2001, 2002 and 2003 for seven grades and 4 test domains (84 data points). I found 30 instances (36%) in which the percentage of recently arrived English learners reaching the 50th percentile equalled or outscored the percentage of their established English learner peers attaining this mark. Almost half of these appeared in the 2003 CAT-6 scores, or fully five years into the structured English immersion program. In some cases, like 10th grade math, 7% more of these recently arrived English learners (about 250 children) reached or exceeded the 50th percentile than did their established English learner counterparts. Moreover, in 2003, a higher percentage of the recently arrived English learners than established English learners achieved the 50th percentile in reading in 4th, 5th, 6th, 8th and 10th grades, in language in the 5th, 6th, 8th and 10th grades, and in math in the 10th grade, though they all still fell well below the County’s average at all points.
I’m not sure what Ms. Stewart would deduce from these data since they fly in the face of her claims about the benefits of “taking English immersion.” Friends of bilingual education would not be surprised by these examples, however, since a primary logic of bilingual education is that the more basic instruction in core content that students receive in a language they can understand, the better they will understand that content and, eventually as instruction in English increases, the more prepared they will be to utilize this knowledge in an English-only context. So, those who understand the premises of bilingual education would explain that these recently-arrived English learners may outscore their established English learner peers because they had some instruction in their native language prior to coming to the U.S. Thus, unlike their established English learners peers, they are not learning English and learning new content at the same time, a difficult task for anyone, let alone a child adjusting to the challenges of survival in a new culture.

Ms. Stewart also finds a remarkable correlation between the number of teachers teaching Spanish and the ‘gains’ among “kids who can read and write in English now.” Most educational researchers would be ecstatic to find such clean relationships, but that nothwithstanding, what Ms. Stewart omits to report is that teachers with a Bilingual/Cross-cultural Academic and Language Development (B/CLAD) certification earn this certification by passing language, culture and history exams and taking 15 additional college units. So, these teachers gain salary increments much like they and/or their peers do when they earn extra graduate level units in M.A. and other post-certification programs.


However, it seems a distraction to argue about “best models” at a time when over 90% of the state’s English Learners are served by one approach only. The real story is not how quickly immigrant students are learning English and “improving in all subjects,” but how great the gap still is between “all students” and English Learners. The numbers reported above notwithstanding, disturbing achievement gaps exist for recently arrived and established English learners in all the areas measured by these two tests (reading, language, math and spelling). Spelling is the one domain where the gaps show some decreasing trends. Spelling is certainly important, but I would argue that our democracy and economy will not thrive if English learners simply master English spelling. In-depth coverage of the realities of English learners in Sacramento County – with the perspectives of their teachers, the students themselves, and their parents as well as transparent analysis of available achievement data – would be a service to the readers of the News and Review who surely want more than the superficial and biased coverage provided in Ms. Stewart’s piece.


Dr. Pia Wong