The College Board’s scanning process for scoring the SAT college-admissions test is reliable, a report released last week concludes.
The study by Booz Allen Hamilton, a consulting firm based in McLean, Va., was commissioned by the New York City-based nonprofit organization this past spring after more than 4,000 students were found to have received incorrect scores on the October 2005 exam. (“SAT Glitches Prompt Broader Testing Worries” March 22, 2006.)
Released July 20 by the College Board, the study concludes that “the current process is reliable and has prudent controls in place to safeguard scoring accuracy.” It notes that some of the changes to the scanning process introduced after the scoring errors were discovered have made the results more reliable.
Those changes include more-frequent scoring checks and an environmental-acclimation process to ensure that answer sheets are not affected by humidity. The errors in the October administration of the SAT may have occurred when moisture caused the answer sheets to expand, according to Pearson Educational Measurement, the Iowa City, Iowa-based company that scored the tests.
The Booz Allen Hamilton report identifies 16 “secondary risks,” including the possibility of software bugs in test-scanning equipment.
“The College Board is addressing every one of those risks,” Gaston Caperton, the College Board’s president, said in a statement.
A version of this article appeared in the July 26, 2006 edition of Education Week