Standardized tests have been battered and bruised lately. In a 2014 Phi Delta Kappan/Gallup poll, more than half of respondents nationwide indicated that standardized testing was “not helpful.” Some states have stepped away from commitments to tests aligned with the Common Core State Standards, while others have experienced disruptions and delays in implementing new assessment systems. Even U.S. Secretary of Education Arne Duncan, long a proponent of test-based accountability, has lamented that standardized testing can be “a distraction from the work it is meant to support.”
Meanwhile, parents in many states are taking matters into their own hands, requesting that their children be exempt from state tests altogether—a process known as “opting out.” In New York, for example, as many as 165,000 students opted out this year alone. National advocacy groups provide sample letters and state-by-state guides for parents interested in opting their children out of tests. In some cases, including in our own backyard of Philadelphia, pockets of educators have joined the conversation and actively encouraged parents to opt their children out. The Philadelphia school district also makes opt-out information available on its website.
As researchers, we don’t take a position on the opt-out movement specifically or standardized testing more broadly. But, as parents, educators, and others debate the role of standardized tests, it’s important to assess whether test-based accountability holds up in the face of a growing number of opt-outs. In the most practical terms, how sensitive are school rating systems to the opt-out movement? And how many students must opt out before creating too much statistical noise for test-based systems to be useful?
To inform these questions, our Philadelphia-based nonprofit group Research for Action ran a simulation using Pennsylvania data and the state’s new school rating system, called School Performance Profiles, or SPP. Our analyses showed that SPP scores were indeed sensitive to opt-outs. For example, and not surprisingly, if high-achieving students opt out, schools’ scores will fall. Moreover, it doesn’t take too many opt-outs to affect a school’s ratings. In cases where a school was hovering just above the threshold for “acceptable” performance—the Pennsylvania Department of Education has said it’s an SPP score of 70—it might take only a dozen or so high-achieving opt-outs to lower the school’s rating.
These school-performance-rating changes could have far-ranging effects—most notably for school employees, as the state transitions to a new evaluation system that ties building-level scores to ratings of teacher and principal effectiveness. Although SPP scores are specific to Pennsylvania, using standardized tests as the primary measure of school performance is the norm—and has been for years. The 13-year-old No Child Left Behind Act assesses schools exclusively on test scores, and even those states that received NCLB waivers from federal officials were required to set up similar test-heavy systems.
Opt-outs have special implications for Title I schools. Both the No Child Left Behind law and the waivers require that schools test 95 percent of students, and face sanctions if they do not. Colorado is the first state to specifically ask for leniency from the U.S. Department of Education for not meeting the 95 percent threshold because of the increasing numbers of student opt-outs. The department is currently considering its position, but it is likely that other states will make similar requests soon.
This special collection of Commentaries reflects a range of perspectives on parents’ opting their children out of tests, from researchers who are studying the phenomenon, to parents who have long embraced testing boycotts, to teachers whose opinions on the subject vary widely.
This Commentary special section on parent empowerment is supported by a grant from the Walton Family Foundation. Education Week retained sole editorial control over the selection and editing of the content; the opinions expressed are those of the authors.
Second, to comply with waiver provisions, Title I schools must meet a set of annual measurable objectives, or AMOs, for tested grades and subjects. In many states, these AMOs are based largely on standardized-test scores, which means that student opt-outs could result in failure to meet some of these objectives. Failing to meet AMOs carries with it certain state-level sanctions. In Pennsylvania, for example, the state department of education requires that the state’s lowest-performing schools implement seven “turnaround principles,” such as replacing, or receiving state approval of, a building principal; undergoing a curriculum audit; and school-schedule redesign. Schools that do not meet all four AMOs for three consecutive years are subject to even more intensive sanctions.
How state education departments will address opt-outs in their evaluations of schools and districts is still an open question. Complicating the issue is the fact that there is little hard data available about which students are opting out. Gathering and disseminating such data would provide parents and other education stakeholders with valuable perspective on the impact of opt-outs, and should be a top priority of state education departments. It would also help educators and district administrators understand the degree of sensitivity inherent in those ratings.
Finally, it is imperative that policymakers understand the potential impact of opt-outs as they continue to craft legislation—in particular, the reauthorization of the Elementary and Secondary Education Act, the law currently known as NCLB—that could attach even more incentives and sanctions to school performance.
