As soon as schools open for the 2020-21 school year — whether at school or online — teachers are going to want to try to figure out where each student is academically so they can, to the best of their ability, help them try to cover lost ground. How teachers can and should assess that is, however, not as obvious as it may seem.
There are some pro-standardized testing advocates who think it is a no-brainer: Give all the kids a standardized test or two.
On that list is Chiefs for Change, an organization formed years ago by former Florida governor Jeb Bush (R), who was a founder of the movement to use standardized tests to hold schools “accountable” for student performance (no matter how poorly the exams were designed). On Monday, the group, made up of state education and school district leaders, issued a statement urging districts to “implement standards-based state tests to every extent practicable in the 2020-2021 school year.”
Others, including a lot of educators, say that standardized tests won’t help teachers get the information they need on each student in real time.
This post looks at what makes sense for teachers as they start a new school year during the novel coronavirus pandemic. It was written by Lorrie Shepard, professor and dean emerita of the School of Education at the University of Colorado at Boulder.
A former president of the National Academy of Education, the American Educational Research Association and the National Council on Measurement in Education, Shepard has studied the use and misuse of tests in educational settings. She consulted with Kathy Escamilla, a professor in the BUENO Center for Multicultural Education at the University of Colorado at Boulder, and Jorge García, chief executive officer of the Colorado Association for Bilingual Education, in composing this commentary.
By Lorrie Shepard
Students will be going back to school soon — either in person or at home — and there are growing calls for them to be given tests so that parents can know whether their children are meeting grade-level standards and teachers will have “diagnostic” information needed to plan instruction.
Teachers, of course, have to determine where each student is, but giving them a standardized test, as some testing companies and pro-testing advocates are calling for, is not the answer. Responding to student needs can be done much more effectively with curriculum-embedded assessments and benchmark instructional tasks.
Unfortunately, the benefits of the existing test products have been greatly exaggerated. Summary scores from multiple-choice questions covering an entire year’s worth of content do not help teachers know what specific skills students have mastered nor what they are next ready to learn. Because these are generic measures, they are not connected to local curriculum.
Then there are the psychologically troubling effects of taking a test to begin school. Many students are already traumatized about the pandemic. Many children, confronted with these tests, will still be in the midst of staggering experiences related to illness, economic hardship and racism. Many of these same children have prior testing experiences of anxiety, frustration, boredom or even shame, when, for example, their low scores were posted on data walls.
At this time of national trauma, we should not forget that tests have a long history of being the cause of inequity rather than a resource for educational opportunity. For decades, tests — first IQ tests and then achievement tests — have been used to sort children of color and English-language learners into low-track classes where learning opportunities and outcomes have been worse than in regular classrooms. The same disproportionality and negative effects for children from low-income households and communities of color have been shown for grade retention, which some are saying should be part of the testing strategy.
The high-profile pushes for beginning this school year with a test do not represent the views of many assessment researchers. Lynn Olson’s recently released “Blueprint for Testing” report summarizes advice from a panel of distinguished testing experts. They stress that educators must “first, focus on student wellness and relationships.” This advice — especially germane after a summer of pandemic isolation and harrowing reminders of racial injustice — is also grounded in learning sciences research from the past 20 years. Relationships and student identities matter for learning and development. Recognizing who students are and the strengths they bring from home and community are critical to purposeful academic learning. (See “asset models for learning in school,” National Academies of Sciences, Engineering and Medicine.)
In contrast, consider the position taken by advocates for formal testing, Nicholas Munyan-Penney and Charles Barone at Education Reform Now (ERN). In a report, they say that “statewide diagnostic assessments could help determine the impact of various types of distance learning on student achievement” and that standards-aligned tests would provide teachers with immediate and actionable data to plan instruction. The ERN report includes a list of 15 commercial test products, which the authors identify as diagnostic assessments.
First, it should be apparent that a fall standardized achievement test could not reasonably be used to evaluate the relative effectiveness of various remote learning practices last spring. It would be impossible to disentangle the effects on achievement outcomes of district policies and practices versus home learning circumstances. No amount of statistical adjustment could make such results interpretable.
The claim that the ERN-endorsed products provide “diagnostic” information is also misleading. To be truly diagnostic and useful to guide the next steps in instruction, assessments need to be closely tied to local curricula, and they need to be responsive to strengths children bring from their homes and communities, thus bringing together the cognitive and relational aspects of effective and caring pedagogy.
The computer-delivered test products reviewed by ERN provide quantitative scores about how far behind or ahead students are (one grade level or two) on broad sub-test dimensions, such as number and operations or algebra, which may not have sufficient reliability to identify relative strengths and weaknesses. Because they rely on machine-scored, multiple-choice or short-answer test questions, they do not provide insights into children’s thinking. They have not been built, as are more fine-grained diagnostic assessments, to identify students’ misconceptions.
Most of these products were designed originally as interim tests to help schools prepare for No Child Left Behind accountability tests. Typically, they cover a whole year’s worth of learning objectives, which often leads to remedial re-teaching of a long list of decontextualized skills, an intervention strategy that has been shown to be ineffective.
A more purposeful and coherent approach for both students and teachers was recommended in a Council of Chief State School Officers report authored by Scott Marion and other experts from the Center for Assessment. They explained how curriculum-embedded assessments and instructional tasks could be selected to target specific skills that students most need to know to engage in the first major unit or two of instruction in their fall grade level. This honors the pedagogical principle of “scaffolding up” to grade-level standards (rather than teaching down to deficits). Revisiting prerequisite skills in the context of a specific unit of study is more manageable for teachers and more meaningful for students.
Such an affirming instructional approach can best be supported through the use of open-ended instructional tasks to help identify learning needs in ways better attuned to children’s social-emotional needs.
Teachers can, for example, invite children to read and talk about grade-leveled texts, with particular attention to vocabulary and comprehension. In mathematics, rich, well-structured performance tasks are available as well as curriculum-embedded assessments that can be used to elicit children’s thinking as well as tapping mastery of prerequisite skills.
Importantly, these activities do not take time away from instruction and they place teachers in a helping role with their students at the outset of the school year, rather than in an evaluative or judgmental one. This is a more equitable and effective approach than the endless, objective-by-objective remediation that would likely follow from interim assessment results.
The calls for fall testing do not seem to be lessening. Education leaders and parents should be wary of administering tests to vulnerable students to find out what they don’t know. Districts already invested in computer-delivered testing systems should consider their use carefully. Those without such products in place should, instead, consider investing in curriculum-embedded assessment tasks and professional development.
It makes no sense for districts to undertake separate social-emotional initiatives and separate diversity and racial-justice initiatives without upholding an approach to instruction and assessment that resonates with these goals — one that recognizes students’ whole selves, what they are experiencing, and how school learning can contribute to their sense of well-being, hope for the future, and contributions as participating citizens.
Yes, students need reading and math for this. Right now, our instructional and assessment practices must show them how all of these things are connected.