This past spring, the U.S. Department of Education gave states permission to cancel federally mandated state testing and accountability reporting because of pandemic-induced lockdowns. As the new testing season approaches, many advocacy groups are urging the department to reinstate testing requirements.
As an assessment researcher who has studied both high-stakes statewide tests and very different classroom-assessment processes, I am alarmed when testing advocates claim that test data will automatically serve equity goals. Advocates do not acknowledge any potential harm from testing for the very students in communities of color most traumatized by COVID-19. If the downsides were factored in, I believe most, even all, state tests would be canceled for 2021.
Even under normal circumstances, high-stakes testing has negative consequences. State assessment programs co-opt valuable instructional time, both for weeklong test administration and for test preparation. Accountability pressures often distort curriculum, emphasizing testlike worksheets and focusing only on tested subjects.
Recent studies of data-driven decisionmaking warn us that test-score interpretations can lead to deficit narratives—blaming children and their families—instead of prompting instructional improvements. High-stakes tests can also lead to stigmatizing labels and ineffective remedial interventions, as documented by decades of research.
Most significantly, teachers report that they and their students experience high degrees of anxiety, even shame, when test scores are publicly reported. These stressors would undoubtedly be heightened when many students will not yet have had the opportunity to learn all of what is covered on state tests. A high proportion of teachers are already feeling burnt-out.
Some advocates, alert to the potential for harm, have argued in favor of testing but without accountability consequences. Clearly it would be unfair to hold schools and teachers accountable for outcomes when students’ learning opportunities have varied because of computer and internet access, home learning circumstances, and absences related to sickness or family disruption.
Testing this year is counterproductive because it potentially demoralizes students and teachers without addressing the grave problems exacerbated by the pandemic.
Others are insisting on accountability for spring 2021, saying that schools and districts had plenty of time this school year to prepare for COVID circumstances. In a recent letter, 10 civil rights, social-justice, disability-rights, and education advocacy organizations urged the Education Department to maintain federally mandated testing requirements so as “to hold districts and states to account.”
That impulse looks very close to blaming educators, who have given so much during the pandemic. It is counterproductive because it potentially demoralizes students and teachers without addressing the grave problems advocates have in mind.
One of the main arguments for testing this spring is to document the extent of learning loss, especially disproportionate losses affecting poor children and communities of color. We are told those data would then be used to allocate additional resources to support students who have fallen the furthest behind.
Indeed, massive investments are needed—especially for summer school and individual tutoring to redress educational inequities exacerbated by the pandemic. Marc Tucker at the National Center on Education and the Economy, for example, urges that we invest in a national tutoring program like that being developed in the United Kingdom.
We already have enough evidence of COVID impacts to warrant federal investments. At the state level, there may not be new monies to allocate because of budget cuts.
Testing advocates should also consider the technical difficulties of testing during a pandemic. Remote testing requires security protocols that would violate privacy laws in some states, and even with such protocols, remote and in-person test results could not be aggregated or compared as if they were equivalent. Bringing all students into schools for testing when some are still learning remotely is unfair.
Consider, too, that the many students who are now absent from remote learning would likely be absent from testing, skewing results compared with previous years. Given the likely inaccuracies in 2021 state test scores, other data sources might be just as good depending on the intended purpose for testing.
If state policymakers need tests to convince them that learning needs are dire, then periodic, computerized tests already administered by districts will work, as shown in new studies by NWEA and Renaissance Learning.
To allocate resources to districts facing the greatest obstacles in meeting the needs of their students, states could use opportunity-to-learn indicators. Data on device and internet access, absences, and time spent in remote (as opposed to in-person) learning, especially for emerging bilingual and special education students, could provide direct evidence of disproportionate needs across districts.
Bear in mind that state tests do little to guide instruction for individual students. Knowing which students are below proficient does not tell teachers what skills they have already mastered nor what understandings students still need. Assessments embedded in high-quality curriculum or key assignments are the best way for teachers to gain substantive insights about children’s thinking, plan instruction, and share information with parents.
Some parents, it is argued, want independent evidence as to whether their children are meeting grade-level standards. But parents lobbying for testing are in the minority. University of Southern California researchers Daniel Silver and Morgan Polikoff reported nationally representative survey results showing that, in October, 64 percent of parents were in favor of canceling standardized testing for spring 2021.
Even with flawless rollout of COVID-19-vaccination programs, it is unlikely that all or even a large majority of students will be back to full-time, in-school learning by the end of February. I think a good rule of thumb would be that students should have been back to normal school for at least a month before being asked to spend a week taking state tests. The new federal education secretary should grant states testing waivers if the condition of a month of in-person schooling cannot be met.
If such a condition can be met, state officials should still weigh the value of going ahead with testing, making explicit how test results will be used to benefit students. What actions will be taken based on the test scores?
This all feels depressingly like the same argument I had with civil rights leaders around the testing provisions of the No Child Left Behind Act of 2002. My colleagues believed that calling out test scores for Black and brown children would force attention to inequities. Some admit today that congressional appropriations did not follow on the scale that was hoped for or needed and that negative side effects were greater than intended.
Before the pandemic, inequities were as great or greater than in 2002 and now are devastatingly worse. Let’s not repeat old testing mistakes.