( This article is responded to by John Butcher in the next column.)
Maybe it’s time for Virginia to scrap the Standards of Learning (SOL) tests.
The SOLs arose in the mid-1990s as a way to provide feedback to the community on how well local schools were performing. It was a worthy experiment. Despite massive increases in spending in preceding years, the quality of education in United States was widely seen as deficient. Backers hoped that transparency would provide teachers, principals, school boards, parents and citizens data they could use to work toward the betterment of their schools.
As with so many reforms enacted with the best intentions, this initiative has gone terribly awry. There is little evidence that SOLs improve anything. Indeed, insofar as the standardized exams encourage teachers to “teach to the test” — more on that in a bit — they may do actual harm.
In short order after their enactment, the SOLs morphed into a means to hold schools “accountable” for poor performance. Schools with low levels of academic achievement were highlighted in local media reports and shamed for failing their students. Newspapers published SOL data for schools within their circulation zones, and parents used the data to guide home purchasing decisions. As parents voted with their feet, affluent households displaced poor households in “good” school districts, and poor households gravitated by default to the “bad” schools. In sum, the tests arguably had the unintended effect of aggravating residential inequality and making it harder for poor schools to improve.
Comparing schools with one another was problematic anyway because educational achievement is strongly correlated with socioeconomic status, the mix of affluent and poor children varied widely by school, and average scores reflected socioeconomic status as much as the quality of the teachers and staff. The Virginia Department of Education (VDOE) possessed the data to adjust scores for socioeconomic status so as to not unfairly penalize low-income schools but, for reasons that remain obscure to me, the department stopped publishing it.
Meanwhile, many teachers, principals and school boards learned how to game the system to dress up scores and avoid the shaming. John Butcher, author of Cranky’s Blog, and I have chronicled numerous scandals around the state — the relaxed test standards, the teacher coaching, the classification of sub-performing students as disabled, and sometimes the outright cheating. Newspaper accounts tend to treat these phenomena as isolated instances that happen to occur in their back yard, but they are in fact commonplace.
Perhaps the most insidious gaming of the system involves teaching to the test. As has been explained to me, the SOLs do not test students’ comprehensive knowledge in a particular subject. Rather, they sample knowledge in sub-topics and assume that if a test-taker gets the answers right for those sub-topics, they will demonstrate the same mastery across the board. Over the years, teachers have learned, to pick an example, that the math SOL will address regular polygons but not irregular polygons, so they spend more time teaching regular polygons and perhaps even skip the irregular polygons. Thus, insofar as teachers teach to the test, meaning that they emphasize certain topics over others, SOLs actually may encourage educational malpractice.
As the emphasis has shifted to holding schools accountable for poor performance, VDOE began using SOL scores to declare schools accredited or unaccredited. (There are various flavors of being unaccredited, depending upon whether schools are deemed to be making progress.) While the state can declare a school “unaccredited,” under the state constitution, schools answer to their school boards. The state does negotiate “Memoranda of Understanding” with chronic laggards but, as Butcher has documented, MOUs consist of educratic mumbo jumbo and are useless in turning schools around. At the end of the day, and not for any lack of trying, accountability remains elusive.
Virginia has doggedly tried to make SOLs work for more than twenty years now. We have enough experience under our belts, I would argue, to draw some broad conclusions. SOLs are deficient as a means of measuring students’ academic achievement; if anything, the teaching-to-the-test phenomenon hurts students. SOLs are useless as a means for improving schools’ academic performance or holding administrators accountable for results; teachers and principals are endlessly creative at gaming the system. And, by influencing people of means to buy houses near “good” schools
, SOLs arguably have become an unwitting driver of socioeconomic and racial segregation.
I’m not saying that things will miraculously improve if Virginia did away with the SOLs. They exist for a reason. But we must acknowledge that tweaking and nudging a broken system won’t work either. I don’t have any great suggestions for what we put in place of SOLs. I can say that reform should be bottom-up, not top-down, and it should encourage creativity and experimentation. Failed experiments should be shut down, and successes should be replicated. We cannot afford more business as usual.
(This article first ran in Bacon’s Rebellion on October 10, 2017)