Standardized testing: Cheating and other problems | Bleader

Standardized testing: Cheating and other problems

by

comment

Test security is no better today than it was 20 years ago.
  • Getty Images/iStockphoto
  • Test security is no better today than it was 20 years ago.
Put a lot on the line with standardized tests and you'll have cheating, critics of high-stakes testing say. They point to the scandal in Atlanta. And they cite Campbell's Law: "The more any quantitative social indicator is used for social decision-making, the more subject it will be to corruption pressures."

Education researcher Richard P. Phelps maintains that this isn't necessarily so. Higher stakes may in fact be leading to less cheating, he says. Phelps has written extensively on standardized testing and is an ardent advocate of it, but also recognizes its limitations. I referred to him in another post last week about high-stakes testing.

Writing in the Wilson Quarterly in 2011, Phelps noted that states and school districts used to buy standardized tests off the shelf, and, to save money, reused them year after year. "Even if educators did not intentionally cheat, over time they became familiar with the test forms and questions and could easily prepare their students for them," Phelps wrote. This raised scores, which made everyone happy: teachers, administrators, and elected officials could talk about how well their students and schools were doing.

Tests are still reused this way, but far less often, Phelps tells me. One reason is that test results are being used to evaluate teachers as well as students, and with these higher stakes "there tends to be more scrutiny."

Phelps believes that test publishers also "did clean up their act some" as a result of a 1987 study that showed rampant test-score inflation across the country from the reuse of tests. But the author of that study, John Cannell, came to a different conclusion. "I fear that little has change and my work was for naught," Cannell wrote in 2006. No Child Left Behind "undoubtedly made the cheating worse" because it "increased the consequences" of doing poorly on tests while there was no reform of how tests were administered.

"The tests are better, but test security often is not," Phelps wrote in his 2011 essay. "Although the fixes are simple and obvious, test security is effectively no better today than it was in Cannell's time."

Teachers shouldn't administer tests to their own students, administrators shouldn't handle test materials in their own schools, and test materials should arrive just before the test, Phelps went on. "Neither teachers nor principals can coach students on specific test items in advance if they don’t have them in advance."

Newer "computer-adaptive" tests make it much harder to drill a class for a test, because questions vary from student to student. But with each new kind of test, a new kind of integrity problem emerges. As I noted in my recent story on standardized testing, Chicago elementary school teachers are now being evaluated in part on how their students improve on Measures of Academic Progress (MAP), a computer-adaptive test administered three times a year.

But in January, teachers at a Seattle school refused to administer MAP, saying it took away too much time from instruction. The Seattle teachers later acknowledged another concern with the test. As in Chicago, teachers in Seattle are being evaluated by the improvement their students show on MAP. Nothing's riding on the outcome for the students, though—so they tend not to take the test seriously, especially at the end of the year, the teachers say, which undermines their "value-added" scores.

The company that produces MAP, the Northwest Evaluation Association, has acknowledged that student motivation can be an issue. "Researchers at NWEA have extensively studied the impact of student effort on MAP results and found that as students get older, the test-taking effort of an increasing number of students decreases," a company press release says. "Decreased compliance of students as they enter adolescence is not confined only to test taking (as any teenager's parents can attest), but to many areas of their lives."

That's probably little comfort to teachers being evaluated in part by MAP results.

I've heard concerns from Chicago teachers about the student-motivation problem, so I asked what CPS was doing to make sure that teachers' value-added scores weren't confounded by it. A spokesperson provided a statement from John Barker, chief accountability officer, which sidestepped the question: "MAP testing provides valuable information to teachers to ensure we are providing our students with a 21st century education. I've found that teachers across the district are genuinely committed to using data from MAP and other assessments to inform their teaching practices throughout the year."

Add a comment