Classroom assessment for learning is a marvelous, cost-effective way of enhancing student learning. Solid research evidence confirms that it works, assessment experts endorse it, and teachers adore it. Nevertheless, I fear for its survival.
To explain, let me wind back the clock. In October 1998, two British researchers, Paul Black and Dylan Wiliam, reported on the work of their colleagues in the United Kingdom's Assessment Reform Group. Black and Wiliam described a powerful distinction between two approaches to classroom testing. They characterized the traditional approach to assessment, in which a teacher tests students chiefly to determine what those students know, as assessment of learning. In contrast, they referred to a more instructionally oriented approach, in which testing plays a pivotal role in helping students learn, as assessmentfor learning (see the November 2005 issue ofEducational Leadership on “Assessment to Promote Learning” for a variety of viewpoints on the topic).
By adroitly juggling a pair of prepositions, the Assessment Reform Group succinctly highlighted the difference between these two approaches to assessment. In their 1998 article, Black and Wiliam drove home the significance of this assessment distinction by presenting a meta-analysis of previously reported empirical research regarding the effects of classroom assessment for learning. Their analysis indicated striking test-score improvements for students, not only on classroom assessments but on external examinations as well. Subsequent meta-analyses by other researchers have confirmed the idea that classroom assessment for learning can be a wonderful way of boosting students' scores on external achievement tests.
Assessment for learning involves the frequent, continual use of both formal and informal classroom assessments. It can be as simple as requiring students to respond to a lesson-embedded, one-item quiz as a way of gauging student understanding of what's being taught. Ideally, this innovative approach to classroom assessment is based on a careful analysis of the enabling knowledge and subskills that students must first acquire to master a higher curricular aim.
Unlike assessment of learning, which attempts to get a fix on what students know for the purposes of giving grades or evaluating schools, the array of test-like events in assessmentfor learning is always linked to the question “What's next instructionally?” Students are key players in this endeavor because one aim of this assessment approach is to empower students to monitor their own progress toward clearly understood curricular goals. Moreover, prominent assessment specialists, such as Rick Stiggins and Lorrie Shepard, have given ringing endorsements to this classroom-testing strategy. There's lots to love and little to hate about classroom assessment for learning.
Why, then, am I fearful for the long-term survival of classroom assessment for learning? In the United States, there is currently one big, federally installed school-evaluation scoreboard in the sky, and it reports whether a school's students have made adequate yearly progress (AYP) on the standardized achievement tests that each state uses to implement No Child Left Behind (NCLB). This highly visible NCLB scoreboard officially indicates whether a school should be regarded as successful or unsuccessful.
Although these key state standardized tests vary substantially, the vast majority of them are instructionally insensitive—that is, they're unable to detect even striking instructional improvements when such improvements occur. This significant shortcoming arises because these tests are so strongly linked to students' socioeconomic status that they tend to measure what students bring to school rather than what they are taught there. Adding to this regrettable reality are the levels of AYP test-score improvement that NCLB requires. According to University of Colorado professor Robert Linn, one of the world's leading measurement experts, current AYP expectations are altogether unrealistic.
Given insensitive standardized tests and unrealistic AYP targets, more and more teachers are apt to abandon sound instructional strategies like classroom assessment for learning. You see, even this powerful classroom assessment strategy won't be able to increase students' scoresenough so that most schools can avoid AYP assassination.
Pressured teachers, then, will most likely succumb in desperation to any sort of quick-fix score-raising techniques that offer the promise of AYP success—even though some of those techniques, such as relentless test-prep drilling using practice items practically cloned from the state's standardized exams, are educationally unsound.
As this pressure continues to mount, I fear that the big tests will drive out the little tests that can demonstrably help students learn. States' NCLB assessments of learning will swamp teachers' classroom assessments forlearning.
- Measure students' mastery of only a modest number of curricular aims or of exceptionally important ones—so teachers aren't overwhelmed by too many instructional targets.
- Provide teacher-palatable descriptions of the skills or bodies of knowledge being assessed—so teachers can direct their instruction toward those assessment targets rather than toward particular test items.
- Contain a sufficient number of items measuring each curricular aim—so the tests effectively report on each student's achievement in terms of each assessed aim.
U.S. educators need to lobby for the installation of instructionally sensitive accountability tests. They can do so by first learning more about such tests, and then using whatever energy and influence they have—individually or in league with colleagues from their professional associations—to urge key policymakers to install more defensible accountability tests. The survival of instructional interventions like classroom assessmentfor learning will depend on it.
