The droll headline marched across the newspapers of Vermont, “More Schools Failing.” Few reporters went beyond the press release and list of schools “not making adequate yearly progress” to ask how the state (under federal direction) made these judgments. Perhaps it is time we hold the federal and state accounters accountable.
Unfortunately, it is impossible to know how many schools improved test scores, stayed about the same or how many actually got worse. Here’s why:
Different tests with different content were used in different grades with different students. Further, some high schools were evaluated using one set of measures and others were evaluated using a different set. Elementary schools had yet a different set of measures.
Last fall, “progress” results were released on the basis of a second grade reading test, an attendance rate and a graduation rate. This summer, these results were compared to entirely different reading and math tests in grades three through eight plus a tenth grade test. Further, this summer, the graduation rate was computed in a different way which knowingly showed greater “failure” rates. Astonishingly, students that are still enrolled are now counted as drop-outs if they don’t graduate in four years.
Nevertheless, the two very different data sets were compared. Proponents of test-based accountability say this kind of comparison is 70% error. We might as well say you were 5’8” last year; weighed 160 pounds this year and, therefore, you did not grow enough. . .
"A child's learning is the function more of the characteristics of his classmates than those of the teacher." James Coleman, 1972
Tuesday, September 12, 2006
A clip from Bill Mathis's commentary at Forum for Ed and Democracy: