State releases new student achievement statistics. Experts balk | KUOW News and Information

State releases new student achievement statistics. Experts balk

Sep 22, 2016

Parents in Washington state may notice another measurement along with their students’ state test scores this year. The student growth percentile (SGP) is meant to compare a child’s learning to their peers. 

"You can kind of think of it like when your child goes to the doctor and your doctor says your child is in the 75th percentile for height," said Nate Olson, spokesman for the Office of Superintendent of Public Instruction. In other words, your child is taller than 75 percent of children their age.

Similarly, SGPs are supposed to show in relative terms how much your child’s state test score improved over last year compared to kids around the state with a similar test score history. That might help parents interpret whether a low score is a sign a child is struggling in a subject - or that one year's test was just unusually difficult.

But as a measurement tool for individual student performance, SGPs have raised the ire of some researchers. Steve Sireci is a professor of psychometrics and directs the Center for Educational Assessment at the University of Massachusetts Amherst. Sireci started researching SGPs after he first received them in the mail for his own son. He and several colleagues studied how SGPs are constructed, and the research regarding their use.

"I hate to hear myself give you this advice, but my advice to parents is to ignore the student growth percentiles," Sireci said. Measurement error is much greater on standardized tests than on a doctor's office scale, he said.

"These growth percentiles are based on two or more tests administered over two or more years, and that measurement error actually compounds," Sireci said. "There is a need to come up with new and better ways to measure students' progress over time, but this is not the way."

The federal government encourages states to use SGPs to show how well schools and districts are doing. Doing so can help qualify states for certain federal grants.

The Office of Superintendent of Public Instruction posts school and district-level median percentiles on its website, but leaves it up to individual districts whether to send families their students' SGPs. OSPI had taken a three-year hiatus from posting SGPs because roll-out the new state test, first introduced in some districts in 2014, didn't allow year-to-year comparison until this past spring.

Individual districts, like Seattle, have also calculated SGPs independently over the years. In Seattle, the measurement was used to assess teacher effectiveness, and used as part of teacher evaluations. Teachers later bargained that controversial practice out of their contracts.

Olson advises caution in interpreting your child’s student growth percentile, which hs said is just one tool. "We don’t want to narrow it down to just one particular measure. I think the entire picture is important," he said. That means test scores – grades – and hopefully feedback from your child’s teachers.