D.C. Reveals 26-School Cheating Probe; Results Are 'Inconclusive'

By Bill Turque
Washington Post Staff Writer
Friday, September 4, 2009

District officials revealed Thursday that they commissioned an investigation last summer into possible cheating at 26 public and public charter schools where reading and math proficiency on 2008 standardized tests increased markedly.

The probe, an analysis of incorrect student answers that were erased and changed to correct answers, found "anomalies" at some of the schools that administered the D.C. Comprehensive Assessment System (DC-CAS) test. But officials called the investigation, conducted by the test's publisher, CTB McGraw-Hill, "ultimately inconclusive."

District officials did not name the schools that were investigated, and they did not release a copy of the CTB McGraw-Hill report, which was requested by The Washington Post on May 29 under the Freedom of Information Act. Officials also offered no explanation for the interval between the conclusion of the investigation in March and their decision to disclose it at a news conference called by Mayor Adrian M. Fenty (D) on Thursday.

D.C. Attorney General Peter Nickles said the full report would be released Friday.

One document from the probe was disclosed Thursday: a March 2009 memo from CTB McGraw-Hill research scientist Steve Ferrara. According to the memo, the "erasure analysis" was requested in August 2008 by then-D.C. State Superintendent of Education Deborah A. Gist.

Accompanying Gist's request was a list of 26 schools with proficiency gains of 20 percent or more on the 2008 test. It is not known whether Gist, who resigned in April to become state education commissioner in Rhode Island, was acting on specific information about possible cheating when she requested the analysis. She declined to comment Thursday.

According to 2008 District test results, the list Gist submitted included Aiton, Hearst, Raymond, Thomas, Hendley, Garrison, Maury, Reed, Draper, Powell, Bowen, Young and Cleveland public elementary schools, along with Winston Education Center, Sharpe Health School and Mamie D. Lee School, which are also public schools. The public charter schools included Community Academy, SAIL, St. Coletta, Tri-Community and Kipp DC Will Academy.

Fenty and D.C. Schools Chancellor Michelle A. Rhee hailed the 2007-08 academic year gains as evidence of genuine progress in a public school system regarded as one of the nation's lowest-performing. The proficiency level for public elementary students rose an average of 11 percentage points in math and eight points in reading; the level for secondary students increased by nine percentage points in reading and math.

Teachers and administrators at public schools with gains of 20 percent or more in both reading and math received cash awards. Public school scores rose again this year but by a smaller amount.

Erasure analyses have been used by other school districts to monitor the integrity of high-stakes standardized tests, which are used by the federal government to determine whether schools have made adequate academic progress under the federal No Child Left Behind law. It is not known whether this is the first time the District has studied erasures.

Student answer sheets are electronically scanned to detect erasures and then examined for suspicious patterns in classrooms and across entire schools. The objective is to determine whether the average number of wrong-to-right answer changes is inordinately higher than the average number of changes established for each classroom, each school and each school district.

Ferrara's memo does not elaborate on the findings, but it recommends that Gist "not draw conclusions about cheating behavior on the basis of these analyses." A follow-up study, conducted by another firm, led to "divergent data," according to a statement by Deputy Mayor for Education Victor Reinoso.

CONTINUED     1        >

© 2009 The Washington Post Company