How We Can Survive—Even Thrive—With Less Testing
Will Lorié and I recently wrote about a forthcoming study where we examined several different approaches for reducing the footprint of state testing. We know that many people will push back against any decrease in state testing. One common concern is that parents won’t get honest, yearly information about their student’s achievement. Another is that cutting back on state testing will lead to reduced accountability pressure, which some see as crucial to improving our educational systems. I’ll focus today on the first concern and address the accountability issue in my next post.
How Well Is My Student Learning?
Twenty-plus years ago, when federal law began requiring testing students every year in grades 3-8, the intention was to make sure that no child was left behind. Parents, caregivers, and educators would be able to receive valid and reliable reports each year about their children’s academic achievement. To be fair, before the No Child Left Behind Act (NCLB) (and subsequently the Every Student Succeeds Act, or ESSA) required schools to test 95 percent of students—and to disaggregate results by various demographic subgroups—many students were conveniently “missing” from both state and district testing.
But we’ve come a long way since 2001. The assessment field has a much better understanding of how to create assessment experiences that are fair and accessible to all students, and policies have changed to make sure essentially all students are included in testing. This tends to be true of district and classroom assessments as well. This has been one of the most important positive outcomes of the NCLB era.
Unfortunately, the increase in state testing required by NCLB was associated with lower-quality tests than the rich performance-based assessments that many states were using in the 1990s. The NCLB era also ushered in the proliferation of commercial interim assessments. The massive use of these two types of tests— tests that relied almost exclusively on multiple-choice items targeted to recall and procedural knowledge—led to well-documented negative effects: Classroom teaching ended up mimicking the kinds of questions on these state and interim tests. So yes, parents were getting annual reports, but of what value?
The consortium assessments—Smarter Balanced and PARCC—first administered in 2014 and 2015, respectively, marked a notable increase in the quality of tests presented to students. They included many innovative types of test questions that reflected the deeper thinking expected of students as they engaged with complex learning expectations. Unfortunately, the perceived need to fully assess the richness of the new content standards made for really long tests. I say “perceived” need because we believed (and I was part of the problem) that we needed to signal to educators and leaders all of the important things they needed to teach.
It wasn’t long before states started abandoning the consortia (although Smarter Balanced is still doing well) or demanding much shorter tests, which meant jettisoning the rich item types many of us think are most important. Most state tests now include higher-quality items than during the NCLB era, but they’re not nearly as strong as during the heyday of the consortia of the 1990s. Again, parents are getting annual information, but of what value?
Filling the Assessment “Void”
If we cut back on state testing to something like every other year or every other grade, states and school districts would have the space to support performance assessment initiatives or other rich assessment experiences in the grades/years without state testing. States and districts could build these rich testing experiences now, but as we regularly hear from local educators, state testing sucks the air out of the room: state and district leaders are already maxed out supporting state testing.
Reducing state testing is not just a fantasy. We have recent experience doing this sort of work. The New Hampshire Performance Assessment of Competency Education (PACE) project limited state testing to once per grade span and supported local educators in developing high-quality performance-based assessments in the non-state-tested grades.
PACE struggled to survive as an accountability assessment system, as required by the federal Innovative Assessment Demonstration Authority, but it delivered two significant victories: it expanded meaningful assessment by using performance tasks in participating schools, and it fostered deeper understanding of assessment practices among educators and school leaders. Parents found the information they received from the performance tasks much richer than the more obscure score reports based on multiple-choice state tests. (This parental preference for performance assessment over standardized test results was documented nearly 30 years ago by Lorrie Shepard and Carribeth Bliem.)
Similarly, the Hawaii Department of Education is leading a project, along with several of my Center for Assessment colleagues, that supports educators in developing high-quality performance tasks and other curriculum-embedded assessments. Hawaii is doing this while still administering its annual statewide assessment, but they’re using a much shorter version of that test. Their goal is to use the richer classroom-based tasks to report higher-quality and more specific information than the subscores on state tests could ever do.
My point here is that we won’t fall off an information cliff without state test scores for every student every year; in fact, we can deliver much better information about student learning. I’ve cited just two examples here, but exciting work like this is going on in California, New York, Kentucky, and several other states and school districts. If we create more space, I believe we can provide much more information to parents about what their students really know and can do than they can learn from a score of 362 (for example) or even a global rating of “proficient.”
The Elephant in the Room: Accountability
Chris Domaleski and I have both written about the need to reform our accountability systems if we are going to be able to truly support assessment innovation. Any state that proposes an alternative testing plan should also include a proposal outlining how its accountability system would need to be adjusted to accommodate the change in state-directed testing. It would also need to provide transparent assurances that its new approach would preserve—or even enhance—the equity goals of NCLB and ESSA, especially enhancing the educational opportunities of students who might need the most support to succeed.
Don’t We Need Some Testing Guardrails?
Yes, definitely. The current testing schedule—annually in grades 3-8 and once in high school— should be the default for states that are not yet ready for—or interested in pursuing—alternatives.
But Congress must authorize the U.S. Department of Education to invite states to propose a reduced testing plan that includes clear proposals for how parents would be informed of each child’s achievement in the subjects identified by the state for the years when a state test is not administered. The proposals should include a defensible theory of action (hey, this is a Center blog; I have to say “theory of action” at least once) for how the state will support assessment literacy and related capacity-building at the state and local levels. These proposals should likely undergo some sort of public peer review process, but we can discuss these details at another time.
If we allow some breathing room for alternatives to testing every student every year, parents benefit from richer information about their children’s learning. But there are many more benefits I did not detail here, including, but not limited to, building educator and leader assessment capacity, increasing local educators’ sense of ownership of an important process, and reporting to parents about how their children are doing without the shadow of the consequences that flow from state testing.
I’m not naïve enough to think that such approaches are without challenges, such as trying to increase educator assessment literacy at scale, but we’ve been trying the same state-mandated approaches for more than 20 years. It’s time to create space for some new ideas.