SAT Score Error Cover-Up Continues

Ten months after a scanning error resulted in inaccurate scores for more than 5,000 students who took the October 2005 SAT, the test’s sponsor, the College Board, still has not provided a coherent explanation of how the foul-up took place or why it took so long to detect and report.

A long-promised analysis of the incident by College Board consultant Booz Allen Hamilton was made public in late July after Sen. Kenneth LaValle, chair of the New York State Senate Higher Education Committee, threatened College Board President Gaston Caperton with a subpoena and a contempt citation. A provision of New York’s “Truth in Testing” law, which LaValle sponsored, mandates the disclosure of any “report cited in memoranda of support or opposition to legislation.” The College Board referenced the forthcoming “consulting firm . . . comprehensive review of SAT scoring” in a memo opposing a bill Sen. LaValle authored to increase state regulation of the testing industry.

But the Booz Allen report failed to address most fundamental questions about the scoring error. It did not seek to identify the root cause of moisture contamination that allegedly caused scanning machines operated by contractor Pearson Education Measurement to misread students’ responses. Instead, it focused on praising recent technical improvements in the scoring process, such as letting answer sheets dry out and scanning them twice, while making such modest recommendations as providing all test-takers with number 2 pencils to ensure responses are legible.

Other yet-unanswered questions include:

o Where did the moisture contamination take place, and why were answer sheets from multiple states and at least two continents affected?

o How long did it take the College Board to detect the error from the time it received the first request for hand re-scoring of the October 2005 SAT?

o Why were students, colleges and high school guidance counselors not notified of the problem until the very end of the spring admissions season?

o What quality control measures did the College Board have in place to prevent scoring errors?

In retrospect, the weakness of the Booz Allen Hamilton report was not surprising. The firm has long done consulting work for the College Board, receiving more than $5 million for its services in 2004-2005. A critical analysis may have put such a lucrative contract at risk.

The College Board’s ongoing unwillingness to conduct or sponsor the truly independent examination that a problem of this magnitude demands leaves a comprehensive investigation up to the courts and legislative bodies. A lawsuit filed by students claiming damages for harm from the scoring error will be heard by a federal judge in Minneapolis in late August. In addition, Senator LaValle is planning fall hearings on proposals to require stricter public oversight of university admissions exams. In the 2006 legislative session, his bill was adopted by the New York State Senate but has not yet been considered by the state Assembly.

The scoring error and its aftermath continue to undermine the credibility of the College Board and its flagship exam. In an editorial titled “SAT’s Own Aptitude in Question,” published shortly after the Booz Allen Hamilton report was released, the Cleveland, Ohio, Plain Dealer wrote, “College Board leaders seem unable to grasp the validity of the growing skepticism about their work. The organization erred and then lagged in letting people know. Those facts mean few will tolerate an attitude of ‘trust us, we’re the experts.’ The stakes for college admission and aid are simply too high to expect the public to accept the College Board’s high-handed behavior. If it doesn’t change tactics quickly, it can expect further censure — and rising popularity among alternatives to its tests.”