Test erasures: The unsolved mystery
By Jay Mathews,
The Office of the State Superintendent of Education (OSSE) serves as the equivalent of a state education department in the District. It provides reams of information, including data that suggest that principals and school district leaders are clueless or disingenuous about test tampering.
The latest example is a large spreadsheet (the printout stretches four feet across my dining-room table) with a detailed accounting of erasures on the 2011 D.C. Comprehensive Assessment System tests at 20 schools. It points to trouble spots ignored in a recent cheating investigation that D.C. Schools Chancellor Kaya Henderson, in a mystifying statement, said gave the community “a renewed sense of confidence in the work that we are doing here.”
For instance, the OSSE spreadsheet said one class of 19 students at J.O. Wilson Elementary School averaged more than nine wrong-to-right erasures per child on the math test and more than eight on the reading test. This was three times the average number of wrong-to-right erasures per student in that grade in math and four times the average number in reading.
Professional educators tell me children almost never make that many changes in those exams, and certainly not so many to their advantage. Many make no changes at all. The explanations offered by D.C. officials for students’ extraordinary success in turning wrong to right — that they had checked their work, had second thoughts about their initial answers or discovered they’d lost their place on the answer sheets — make no sense to classroom veterans. These pros, both teachers and principals, instead say it’s more likely school administrators tampered with the answer sheets after the students went home.
J.O. Wilson has a history of students experiencing amazing flashes of intuition and changing many wrong answers to right ones. In 2008, according to testing company data released by OSSE, 93 percent of classrooms tested at Wilson had wrong-to-right erasures far above the average. In 2009, the percentage was 83 percent. In 2010, it was 100 percent, the highest in the city.
D.C. officials told the team from the Alvarez & Marsal consulting firm that investigated Wilson in March to ignore the 2008, 2009 and 2010 erasure results. The investigative guidelines had been changed, they said, and only one classroom at Wilson — it is unclear what grade — had enough erasures in 2011 to qualify for the probe.
So did the investigators address the erasures and explain what happened? Nope.
The team’s two-page report on Wilson does not mention erasures at all. Wilson was given a clean bill of health: “No potential testing violations were identified,” the report said.
Did the investigators ask the students in that classroom at Wilson if they remembered erasing any answers, and if so how many? Alvarez & Marsal declined to tell me. The D.C. schools and OSSE also had no answer. Did the investigators compare the students’ recollections with the actual number of erasures on their answer sheets to see if they matched? No one will say.
About a dozen other classrooms on the OSSE spreadsheet show two or more times as many wrong-to-right erasures as the D.C. average, but the report offers no explanations of those erasures, either. The investigators seemed uninterested in how the answer sheets acquired so many smart corrections.
One other investigating team has not been heard from. The D.C. Inspector General, with help from the U.S. Education Department, has been examining pre-2011 erasures and scores for more than a year. Don’t expect too much. Even the D.C. state superintendent, Hosanna Mahaley, has praised the Alvarez & Marsal report that ignored her agency’s data.
If neither Henderson nor Mahaley recognizes the problem, I don’t think anyone else with the power to get at the truth will, either. The mystery of who made all those erasures will remain unsolved.