Study says fairer accountability system means more tests
August 05, 2008
An Ohio State University sociology professor says the state's new value-added method for measuring student academic progress is an improvement to the accountability system but still doesn't go far enough.
Sociology professor Doug Downey told the Cleveland Plain Dealer that Ohio's new approach is "substantially better" than the old method but he still doesn't think it goes far enough. He believes the accountability system would be better if, instead of one test, students are administered two tests (one in the autumn and one in the spring). Downey argues that one test does not measure how much students retain from the preceding year and a test should be administered in the autumn after students return to school to gauge that knowledge.
Further, in the current issue of Sociology of Education (see here) Downey and researcher Paul von Hippel argue that the common method many states use for evaluating schools, called the status model, is biased and unfairly labels schools that serve predominantly disadvantaged children as "failing."
The researchers believe the "status" model, which measures student achievement at a single point in time-as required by No Child Left Behind-fails in comparison to an "impact" model . An impact model attempts to measure how children build on what they have previously learned year to year. Impact models attempt to account for learning lost over the summer months but disadvantaged children are more likely to lose ground during that break (see here).
While Ohio's achievement tests can and should be improved, the problem isn't the tests. The real problem is that we have such huge disparities in test results and that not enough is being done to intervene with underachieving students as identified by the tests. For example, why are we still telling children who are academically behind to take the summer off? Ohio's standardized assessments remain the best indicator we have for telling whether children are proficient in basic subjects like math and reading.
Downey is correct, though, that achievement tests should be supplemented with other factors, like growth over time. The Ohio General Assembly, State Board of Education, and Ohio Department of Education are to be commended for bringing value-added indicators on-line and including them in the state's accountability system. Even under our current assessment system, there are schools that succeed in helping disadvantaged youngsters reach high levels of achievement. Perhaps, instead of adding tests and revamping the accountability system, the state should put that money and effort toward following the lead of KIPP and others by offering longer school days and an extended school year.