A University of Kentucky business professor says it's all too easy to figure out standardized test performances of minorities at schools across the country using publicly available data and a little bit of math.
That, essentially, would violate federal laws that require individuals' test scores remain private, said Krish Muralidhar, a professor of decision science and information systems at UK's Gatton School of Business.
Only the performance of groups of students should be released publicly, according to the Federal Educational Rights and Privacy Act.
While school systems don't publish scores for small groups with fewer than 10 students in them, Muralidhar said its often simple to calculate certain individuals' results when a school has a few small groups of ethnic minorities or other minorities, such as students who receive free lunches.
Never miss a local story.
"To me, it's a very prevalent problem," said Muralidhar, who has researched privacy issues with personal data in businesses.
Muralidhar first looked into how school systems publish test result data when his daughter brought home her test scores from Paul Lawrence Dunbar High School last year.
Large and diverse schools, such as Dunbar, don't have the same problems as smaller schools, Muralidhar found as he checked out the Kentucky Department of Education's test results Web page.
As an example, he pointed to 2008 results for an elementary school in Adair County, where the third grade had 60 students, in which 56 were white and four were Hispanic.
Data for the performance of the four Hispanic students was listed as "N/A"
"By saying 'N/A' you're literally inviting someone to say, 'Hmm. I wonder if I can figure this out,'" Muralidhar said.
Using multiplication and subtraction, Muralidhar determined how many of the Hispanic students performed at novice level on the test.
Muralidhar pointed this out to Kentucky Department of Education officials, who last year awarded Muralidhar and his research partner, Rathin Sarathy, a professor of business at Oklahoma State University, with a $20,000 grant to assess how widespread the problem was.
Ken Draut, associate commissioner of the Office of Assessment and Accountability at the Kentucky Department of Education, said as a result of their work, officials decided not to release the number of students in classes this year — just what percent of students scored in each of the categories: novice, apprentice, proficient and distinguished.
The change didn't sit well with some education watchdog groups, such as the Prichard Committee for Academic Excellence, which want more transparency, Draut said.
That underscores the delicate balance between public information and the individual's right to privacy.
"There's tension between protecting that individual's score and the public's right to know," Draut said. "Krish has pointed out that someone who wanted to take a little bit of time could easily figure out the scores of these kids. So we said we'd rather err on the public release of data and protect the privacy of students."
As a result of the changes, Muralidhar said Kentucky is "at the forefront" of protecting students' privacy.
He and Sarathy made presentations across the country to point out privacy problems with test data from California to Indiana.
"The demand for information comes from the public at large," he said. "But the demand for privacy comes from individuals. Unless you are one of the individuals affected, you're never going to demand it."