A better way to grade schools
April 06, 2005
As winter turns toward spring, we turn toward a perennial spring event: student testing. With that testing comes the inevitable anxiety as states brace themselves for the annual status races.
My state, Wisconsin, is no exception. We look ahead to this testing season with concern about how our performance data will measure up to results from other states, other districts, other schools. As a result of NCLB, which requires students in numerous subgroups to move toward "proficiency" in reading and mathematics (and, soon, in science), schools and school districts will either "make AYP" or be labeled "in need of improvement." Unfortunately, these assessments do not by themselves tell the full story about how well a school is performing.
That the current system does not fairly depict the quality of a school or district became clearer than ever in a study released last year by the Northwest Evaluation Association (NWEA), a nonprofit organization comprised of some 1,500 school districts. (See "Individual Growth and School Success.") NWEA has one of the largest repositories of longitudinal student achievement data in the world.
The study authors posed the question: If School A and School B had identical state test scores, would they have similar success with students? Consider that School A started the year with low-performing students and caused every one of them to grow twice as much as students in School B. We immediately realize that end-of-year testing data can tell us where students are at a point in time, but not where they started or how far they traveled to get there.
Some of the findings of this important study require us to challenge our current perception of school success. It turns out that:
- Schools with similar status levels differ substantially in the achievement gains of their students.
- More than 20 percent of "high-achieving" schools fall into the bottom quarter of schools in terms of the value they add to their students' achievement.
- Many "low-achieving" schools actually cause as much growth in their students' learning as the best high-achievement schools.
Impact of Focusing on Status
If you are appraising a school's success, it matters a lot whether it is adding academic value to its pupils and at what rate. Such growth information is critical in developing a more complete picture of school performance and school effectiveness. But the current NCLB requirements do not factor in measures of growth and, as a result, cause unfair consequences for schools in several ways.
Fiscal Impact: Schools that effect substantial growth in low-performing students may still be subject to needless disruption (intervention, reconstitution, etc.) if they don't bring students all the way up to proficiency. This may result in loss of students, staff, and funds, not to mention morale, from schools that are, in fact, positioning low-performing students for future success.
Impact on High Performers: Under NCLB, two schools with mostly proficient students will be labeled as equally successful, even though one's pupils are making scant academic headway while the other's are moving toward superior performance. This can give parents and educators in high-achieving but low-growth schools a false sense of confidence and hope.
School Choice Impact: If only final achievement levels are used to identify schools that make adequate yearly progress, students who change schools won't have all the information they need to make informed decisions about which schools are most effective. Not even parents and students, let alone teachers, have access to student growth data to know whether a school is really propelling students forward academically.
Redefining School Success
Accountability for school success is important, but we need to include high-quality growth measures (i.e., value-added assessments) if we are going to gauge how successful a school or district truly is.
Parents with children in a number of Wisconsin school districts are beginning to see the power of looking not just at children's baseline achievement levels but also at their growth. These school districts have recently begun to use NWEA's computer-adapted and Internet-enabled test that measures growth in learning in reading, language use, and mathematics. Nationally, more than a dozen state school chiefs recently endorsed the concept of using a growth model as a measure of accountability.
Without some indicator of individual and collective student growth, we cannot identify schools that lack the required percentage of students scoring "proficient" yet are making significant gains in reading and math. Conversely, we are not able to distinguish between "proficient" schools that are maximizing student learning potential and those that are merely maintaining the status quo.
NCLB should be a "floor" for states to judge school accountability, ensuring a minimal level of accountability while allowing states to develop additional indicators of school success. One of those indicators should be growth measures for each and every student - the only way we can begin to understand if a school or school district is truly successful in leaving no child behind.
Bill Breisch is director of instruction for the Monona Grove School District in Monona, Wisconsin. He can be reached at firstname.lastname@example.org