Don't Be Misled by Tests
"More schools sanctioned," the Free Press reported on its Nov.16 front page, one day after the Vermont Department of Education released results of federally required "adequate yearly progress" determinations.
Based on spring '04 testing, the number of public schools not meeting No Child Left Behind progress reports jumped from 31 to 39. What's going on here? Is Vermont school quality actually declining?
Some readers are bound to have such doubts, an evitable outcome of flawed No Child Left Behind guidelines.
Standardized tests are misused. Students are reduced to test scores. Students' scores are being analyzed within a distorted framework. How is the framework distorted? No Child requires as many as 37 different measures for each Vermont school. That's separate math and reading/language arts exams given at six levels (grades 3-8) and then sub-divided again into student racial, income, disability, and native language groups, plus several other indicators.
A school's failure to hit its progress target goals in only one category for two consecutive years starts a chain of worsening sanctions. That's like answering one question incorrectly on a 37-item test and receiving a failing mark.
Moreover, absolutely no differentiation is made between a school with insufficient progress in only one student category and another school with insufficient progress in all 37 measures.
No Child Left Behind unrealistically demands perfection in another manner. All students in all public schools must achieve state standards for math and reading/language arts by 2014.
That means all learning disabled students. It means all students living in poverty and suffering through dysfunctional relationships at home. It means all students of limited English fluency.
Schools receive no particular tribute for the learning growth of students from their starting points, or for excellent student achievement in other academic subjects and the arts. There is no credit for schools that consistently narrow the achievement gap within and among student sub-groups, but just not as fast as No Child demands.
There generally is not enough flexibility to use other valid and appropriate student assessments. There are no tangible rewards for schools, only punishments. Mandatory progress determinations are purely numbers-driven.
They presume all children in all schools can and should make the same achievement gains within one year. Laws of statistical probability operating for up to 37 different progress measures simply increase the likelihood annually that more and more schools will not attain their progress goals every year in all categories.
Additionally, Vermont's relatively low school enrollments are more prone to sharp volatility of annual school-level scores. Student scores are used to set annual progress goals and also to gauge whether they have been reached. Yet, basic statistical theory attests to the unreliability and inconsistency of conclusions drawn from small samples.
It's only a matter of time, probably beginning next year, before the number of Vermont schools not attaining progress goals rises dramatically. For instance, only two schools exited No Child's progress sanctions this year.
Fourteen -- seven times more -- didn't make progress goals for the first time. Of 31 schools that didn't make it in 2003, only five made the goals in 2004. Twenty-two others -- over four times greater -- didn't make the goals again in '04 and now must accept "School Improvement" sanctions. No Child Left Behind's exclusive reliance on standardized test scores and skewed progress calculations heavily stacks the odds against local community schools and educators. Ultimately, it threatens to undermine public confidence in them. Instead, we should support teaching and schools that recognize children as unique learners.
Vermont has great public schools and educators. We need and deserve great public support for them.
Angelo J. Dorta is Vermont-NEA president.
Burlington Free Press
INDEX OF NCLB OUTRAGES