9486 in the collection
Peter Schrag: The test results: Lots of numbers, not much light
By Peter Schrag
When school people talk about educational achievement gaps -- generally the difference between whites and Asians and other groups -- blacks and Latinos are conventionally mashed together as "poor and minority." But earlier this month, as he was releasing two sets of test scores, state school superintendent Jack O'Connell emphatically rejected the old mush.
"These are not just economic achievement gaps," he said. "They are racial achievement gaps." The evidence, part of it anyway, lay in the fact that white and Asian kids classified as poor scored better on the California standards tests in math than African American and Latino students classified by the state as not economically disadvantaged.
Those kinds of numbers have long been familiar. The news is that O'Connell emphasized the racial element.
The immediate question is: What does "racial" mean? Is it parents and home environment? Is it peer culture? Is it genetics? Is it the test itself? And, more important, how do schools address it if, indeed, they can address it at all?
O'Connell, who's planning a major conference on the achievement gap in November, also seems to be opening the door just slightly on the radical idea that part of the solution lies outside of school -- in health care, housing policy, community renewal, recreation, counseling, quality preschool programs, maybe even residential schools for young kids, none of which he mentioned.
There are lots of questions about the numbers. The state categorizes any student who's not eligible for free or reduced priced lunches and has at least one parent who finished high school as not economically disadvantaged. So at least some kids in dysfunctional homes could look middle class in the state's official numbers.
And despite the fact that poor Asians and whites do as well or better than non-poor blacks in both math and English, it's still true that, over all, middle class students do far better than the economically disadvantaged. In English, 60 percent of nondisadvantaged students score at proficient levels or above compared to 29 percent of economically disadvantaged students. In math, it's 52 percent vs. 31 percent.
There are other questions hidden in those numbers: The definition of "proficient" is totally arbitrary. And what's the difference in academic achievement between African American boys, often regarded as the most troubled group in this picture, and African American girls? A study due in the fall may tell us.
To what extent, moreover, does the lack of any clear incentive to do well influence performance on the California standards tests? The ability to "walk" at commencement is motivation to pass CAHSEE, the California High School Exit Exam, a minimum competency test pegged at ninth or 10th grade for English and eighth grade for math. There's no similar spur on the standards tests, which cover most major subjects in grades two through 11.
Even so, the test results released last week have more encouraging news than O'Connell acknowledged. As the early headlines indicated, there was no overall score change from 2006 to 2007 in the percentage of students rated proficient in math and a small decline in the percentage of English learners who had passed the exit exam in 10th grade. (They'll have several more chances to pass it before they're scheduled to graduate in 2009.)
But the flat math numbers hide the impressive increase in the number of students taking what, at least in name, are more rigorous courses and the steady decline in middle- and high-school Mickey-Mouse math course enrollments. In 2003, the first year it was given, 505,000 students took the algebra I test; in 2007, 741,000 took it.
In 2003, 270,000 took the geometry test; in 2007, 370,000 took it. In 2003, 162,000 took algebra II; in 2007, 230,000 took it. In the same period, the number who took "general mathematics" declined from 451,000 to 305,000. The surprise is that math proficiency levels didn't go down.
Something similar holds for the decline in the percentage of English learners who passed the high school exit exam in 10th grade. The wonder is that so many officially classified English-learners passed even this easy test.
And since there's no way to know who these students are -- how many just arrived speaking no English, how many of the most capable were recently reclassified -- the whole category is fatally fuzzy. Nor do we know the relationship between tests, test scores and dropouts.
All of which indicates that under the avalanche of numbers, there's a serious shortage of good data. We still don't have the statewide student tracking system allowing California to measure achievement growth for individual students over time -- and thus the real productivity of the system -- and not just the difference between this year's class and last year's.
We should have had CALPADS, the California Longitudinal Pupil Achievement Data System, going long before now. But last week, the great legislative sausage machine again crunched local assistance funding for it out of the budget. Maybe we like the dark cave we live in.
— Peter Schrag
Sacramento Bee
2007-08-29
INDEX OF OUTRAGES
Pages: 380
[1] 2 3 4 5 6 Next >> Last >>
