D.C. Test Score Gains Too Good to be True

Standardized tests remain wildly capricious things. But is the pressure proving too much for some schools?

By Julianne Hing Mar 28, 2011

Washington, D.C. schools are feeling the heat after a USA Today investigation found that local schools posted large test score gains that may have been the result of impropriety, rather than improved student performance.

USA Today found that at Crosby Noyes Education Campus, just one of a list of Washington, D.C. schools’ test scores the paper analyzed, only 10 percent of Noyes students were considered "proficient" or "advanced" in math in 2006, but by 2008, 58 percent of the school’s students were at that level. Noyes also posted alarmingly high erasure rates that were higher than D.C. averages–test score scanners track not only the final choice students bubble in but the marks they made and erased.

From USA Today:

On the 2009 reading test, for example, seventh-graders in one Noyes classroom averaged 12.7 wrong-to-right erasures per student on answer sheets; the average for seventh-graders in all D.C. schools on that test was less than 1. The odds are better for winning the Powerball grand prize than having that many erasures by chance, according to statisticians consulted by USA TODAY.

USA Today found that not only did Noyes have high erasure rates, but that students consistently changed their answers from wrong to right. In the meantime though, Noyes was honored as a National Blue Ribbon School in 2009, and was hailed by then-superintendent Michelle Rhee as a model for other schools trying to make similar gains. Noyes faculty were rewarded for their work with $8,000 bonuses in 2008 and 2010.

The district’s new superintendent Kaya Henderson resisted drawing any causal relationship between the findings.

The district "has investigated all allegations of testing impropriety," Henderson told USA Today. "In those situations in which evidence of impropriety has been found, we have enforced clear consequences for the staff members involved, without hesitation."

McGraw-Hill, which runs Washington, D.C.’s testing and was hired to do erasure analysis of D.C. schools, said high erasure rates were not proof of cheating, and that students are more likely to change their answers when teachers urge them to review their work. But USA Today found that the company acknowledged that high erasure rates could identify "possible cheating incidents for follow-up investigation," according to a McGraw-Hill document.

Standardized tests are supposed to be reliable methods to measure the quantifiable aspects of a student’s education. Under the Obama administration’s reform agenda test scores have become even more crucial to gauging a student’s mastery of a subject, and an educator’s own teaching proficiency. It turns out, though, that rather than being sound and valid measures of either, standardized test scores are wildly capricious things. They jump from year to year, even within the same grade and among classrooms taught by the same teacher. The reason, education advocates argue, is that there is so much pressure to raise test scores that education has shrunk. Teachers must "teach to the test," and increasingly, teachers’ jobs depend on their ability to raise their students’ test scores. Under No Child Left Behind and Race to the Top, the competitive grants program started by the Obama administration, schools that don’t post adequate test score gains can get shut down and replaced by brand new staff or charter schools.

Earlier this month a Detroit Free Press and USA Today investigation found similarly improbable testing performance in Michigan schools, suggesting that some kind of test tampering may have taken place. The Michigan State Board of Education opened an investigation into 34 schools whose test scores had made improbable leaps in their test score performance.