Editor鈥檚 Note: Associate Editor Stephen Sawchuk covers curriculum and instruction. This analysis is part of a special report exploring pressing trends in education. Read the full report: 10 Big Ideas in Education.
Ask anyone from a school accountability expert to a parent of a school-age child, and you will get near universal agreement that we have a dysfunctional standardized-testing system in the United States.
Educators do not like the annual statewide tests: They inform school penalties, not learning, because the results come so late in the school year. They fail to match any specific curriculum, and generally don鈥檛 deeply measure students鈥� analytical capabilities or the dispositions employers and colleges value.
Our nearly 20-year experiment with yearly federally required exams has boosted math scores, research indicates, but those gains have petered out as accountability pressures have grown more acute. And efforts over the last decade to produce better tests have been met with tepid enthusiasm.
Could classroom assignments really measure student progress as effectively as standardized testing? Scroll down for more from Keith Dysarz.
For most teachers and students, testing remains the educational equivalent of a root canal. It鈥檚 endured, and then recalled with relief. It鈥檚 not something deeply integrated with teaching that reinforces long-term educational goals like developing students鈥� ability to engage in thoughtful, reasoned argumentation.
The opt-out movement was a warning sign, the first rumble of a sentiment that has spread, just like the teacher protests of the past year. If policymakers want to preserve the long-term health of our testing ecosystem鈥攁nd help teachers feel invested in, rather than victimized by it鈥攖he time to act is now.
The main questions on the table: Should we rethink the annual testing of every student? Could such a revision be balanced against the clear need to maintain a focus on those students who have been least well-served by K-12 education?
For now, annual testing isn鈥檛 going anywhere; it鈥檚 enshrined in federal law. But if the politics of the last two years prove anything, it鈥檚 that longstanding policies, even those long thought inviolable, can be rewritten.
With testing, you can鈥檛 have your cake and eat it too: Year-end testing offers longitudinal, comparable data in each state on student performance. But it鈥檚 severely limited in scope and depth. The alternative鈥攖esting that better captures student learning, including portfolios and extended writing prompts鈥攃osts far more in money and time and is harder to make valid and comparable.
The Obama administration attempted to square this circle by providing seed funding for richer annual tests, but the results were less revolutionary than proponents had hoped, and they were quickly caught up by the anti-common-core and opt-out movements.
What would a world with less-frequent testing look like? There are alternatives worth considering, although each contains significant tradeoffs:
鈥� Do what the National Assessment of Educational Progress does, by measuring samples of students rather than every student. The resulting savings could be put into test formats that better measure students鈥� ability to integrate knowledge and analyze.
鈥� Continue to test every student, but do so only every two to three years, again investing the cost savings into more useful systems.
鈥� Integrate testing with curriculum practices, as is the case in the United Kingdom and most European countries鈥攚here mandatory tests used to grant 鈥渓eaving certificates鈥� and other student credentials directly reflect the actual curriculum taught in schools and require the extensive use of essays.
鈥� Revisit the notion of drawing from classroom artifacts of student work to inform less-frequent testing results.
No discussion of these alternatives can be complete without pointing out their very real disadvantages, for equity and transparency, at a time when education advocates are pushing harder for both.
The point of testing every child, every year, was to expose the serious performance gaps among and within schools. That鈥檚 why many civil rights groups support the practice so strongly, and fear what could happen if it鈥檚 lost.
Doing away with annual testing could also hurt efforts to study what鈥檚 working at scale. After all, annual testing led to the development of data systems that could track kids鈥� trajectories over time, prompting a new wave of empirical research on school and teacher quality.
Finally, the ideas above are also ambitious. Linking classroom practice to testing is rarely done in the United States, and previous attempts to use informal data from classroom work samples have faltered. Technology could help, but in the age of Big Data, it also raises all kinds of privacy challenges.
Nevertheless, there are signs of the beginnings of a transformation: States and school districts are experimenting with several ideas.
Georgia is revisiting the idea of using a series of smaller, classroom-based tests better aligned to everyday lessons under its own state-level initiative.
A few other innovations are emerging thanks to a small (and not terribly popular) pilot run by the U.S. Department of Education. Louisiana, for instance, is hard at work creating a reading test based on the actual novels and texts students read, rather than dull, random reading passages that favor students with the most background knowledge. And with federal permission, New Hampshire allows districts to use shared performance-based tasks developed by teachers for some grades, and traditional tests in others.
Some individual districts are also charting their own paths: Prince William County, Va., among others, uses a British exam system more tightly integrated with curriculum goals in a handful of its schools.
What will happen next in testing? It鈥檚 largely anyone鈥檚 guess. But our current divorcing of standardized testing from the everyday aims of schooling seems untenable for the long haul. It鈥檚 time for smart policymakers, educators, and testing experts to lay out a new path.
PERSPECTIVE
How to Make Classroom Assignments Count
By Keith Dysarz
Measuring student learning is a critical component of education. Assessments, when done well, provide essential data for guiding students鈥� learning and inform teachers, parents, and school leaders of students鈥� progress. But when done haphazardly and without strategy, assessments waste time and have limited value.
As frustration with standardized testing regimes continues to roil educational waters, districts and schools should pay attention to an option that can be just as informative in measuring student progress: classroom assignments. The tasks that teachers ask students to complete daily are an important piece of the larger assessment puzzle.
Assignments that challenge students provide an opportunity for simultaneous learning and assessing. Tasks that reflect grade-level standards can help students build their critical thinking skills. Such high-quality tasks are a type of formative assessment, yielding immediate and actionable data on students鈥� proficiency.
The problem? As significant as assignments can be, research from the Education Trust shows that most student assignments fall far short of demanding college- and career-ready standards. In the more than 6,000 assignments we analyzed, less than 12 percent pushed student thinking to higher levels. The majority of assignments were low-level, provided little-to-no opportunity for students to communicate their understanding in writing or through discussion, and did little to engage them in content. Similarly alarming results were found in research by TNTP.
Notice the contrast, for instance, between a 10th grade English assignment that asks students studying a Susan B. Anthony essay, 鈥淲hat does 鈥榮uffrage鈥� mean?鈥� and 鈥淲hat was the central idea of the article?鈥� with another assignment that calls for 10th graders to write an essay reflecting on power differences linked to cultural accommodation using the short story they have just read and other specific evidence.
If what teachers are asking students to do is not on grade level and does not push their thinking or provide opportunities to justify their reasoning, there is little value in using it for formative assessment.
Educators must be equipped to recognize the features of high-quality classroom assignments, whether garnered from highly rated instructional programs, district curriculum, or teacher-created tasks. When such assignments are the norm, educators will also have the kind of meaningful formative assessments our students deserve.
Keith Dysarz is the director of P-12 Practice at the Education Trust.
