On Oct. 23 I introduced speakers at a revealing countywide forum on the Virginia Standards of Learning (SOL). The forum, held at George Mason University, was sponsored by a half-dozen business and educational associations in Fairfax County and was attended by more than 1,800 people.
Gerald Bracey, director of research, evaluation and testing for the Virginia Department of Education from 1977 to 1986, spoke, and written testimony from Lawrence Cross, professor of educational research, evaluation and policy studies at Virginia Tech, was presented. Both men found that problems with the SOL are egregious and counter to the stated purpose of the state board of education to promote quality education in Virginia for all students.
For example, the SOL tests at grades three, five, eight and 11 are multiple choice. This may mean that they are inexpensive to administer and grade, but it also means that they measure little other than the ability to take tests and memorize information. Also, an inordinate amount of classroom time is spent teaching to the test.
Some of the standards also are ridiculously high. My own reading of some of the standards leads me to believe that 70 out of 100 adults could not meet many of them. Both Bracey and Cross believe that the cutoff scores for passing or failing are absurdly high. Bracey compared Fairfax County Public Schools with schools in 41 developed nations using the Third International Mathematics and Science Study. He found that Fairfax schoolchildren "would outscore all 41 countries in science and perhaps all 41 in math." Nonetheless, on the second administration of the SOL tests, 78 percent of Fairfax Public Schools -- and 88 percent of schools statewide -- did not pass enough of the SOL to meet state accreditation standards.
Cross's written statement asserted that the SOL reflect the socioeconomic status of students more than the quality of education they are receiving. Numerous studies have cited socioeconomic status as the single best predictor of test scores. Cross obtained the mean SOL scores for each school division in Virginia and was able to predict two-thirds of the variations in SOL scores across school divisions on the basis of three indicators of socioeconomic status.
On top of all these problems with the SOL tests themselves, the way they are being used violates standards for test use promulgated by the American Educational Research Association, the American Psychological Association, the National Council for Research in Mathematics and the National Council for Measurement in Education. These standards say that standardized tests should not be used in isolation for making education decisions. Yet for high school graduation and school accreditation in Virginia, nothing counts but the SOL scores.
Under the buzzwords of "accountability" and "high standards," the SOL tests are being used to do a great injustice to Virginia's young people.
-- Brett Cramer