|
Interpreting Test Scores: More Complicated Than You Think
Interview with Daniel Koretz by Susannah Tully, Chronicle of Higher Education, August 15, 2008
The utility of standardized test scores for both college admission decisions and assessment of student learning is a much-debated but not always well-understood topic. Daniel Koretz, author of Measuring Up: What Educational Testing Really Tells Us spoke to the Chronicle of Higher Education in an August 15 interview about some of the nuances of understanding standardized test scores.
The key testing issue today, Koretz says, is the expanding use of standardized tests to hold institutions accountable. Under the No Child Left Behind Act, standardized tests have become “the cornerstone of K-12 education policy in America,” and many leaders are pushing to expand such testing to higher education. The problem, Koretz explains, is that “test scores describe but don’t explain: They tell us what students know but not why they know it. It’s become routine to treat differences in scores—for example, among schools or districts—as simple indicators or the relative quality of education. That’s nonsense.”
There are two major pitfalls to interpreting test scores, Koretz says: not realizing that they provide only an incomplete description of what students can do, and ignoring score inflation. While some scores are reasonable, others are highly inflated, and educators don’t often know which are which because testing programs are so seldom evaluated. Koretz argues that standardized testing could be much improved by removing the reliance on benchmarks and performance standards, by more frequently evaluating tests, and by ultimately developing new ways to assess students and schools.
The entire interview is available on the Chronicle’s Web site. Subscription or login may be required.
|
