<i>Formative assessment</i> is now at risk of being understood merely as testing that is done often.
Sometimes the vocabulary we use as educators starts out with one meaning, but morphs over time into something entirely different. We speak, for example, of wanting a curriculum that is aligned, intending that our district and classroom content will be matched to state standards. This alignment would further mean that there would be congruence among the written, taught, and learned curriculum. Enter textbooks, supplemental materials, and software programs, now presented to schools with the promise that publishers’ series or products are aligned to states’ standards. The term “aligned” here becomes relative: In some cases, the correlation is tight and specific; in others, it may be loose to vague. Whatever the level of alignment, it now seems that everything purports to be aligned to everything else, and the term itself begins to lose value, if not meaning.
Standards-based reform started out being about all students’ learning well, about pre-identified content and performance standards, and about student attainment of standards—not seat time. Today, in a form that some argue is greatly debased, it is just as recognizable as what Lorrie Shepard, the dean of the school of education at the University of Colorado at Boulder, has described as a “heavy-handed system of rewards and punishments.”
Currently being reshaped in today’s No Child Left Behind environment is the term formative assessment. It now is at risk of being understood merely as testing that is done often. In some extremes, it is little more than frequent summative assessment: testing that doesn’t originate in the classroom, that creates another mark for the grade book or a set of data to be analyzed, and that, in theory, tracks individual and/or group progress toward the ultimate summative test—the high-stakes test that quantifies the school’s adequate yearly progress.
And that’s the whole point, isn’t it? When this is the way formative assessment is defined and used, clearly what we’re trying to “form” is higher test scores. The need to raise test scores is of course real, and unlikely to go away any time soon. Still, we may want to watch out for what could be sacrificed unintentionally on our way to adequate yearly progress. That is, formative-assessment practices used by classroom teachers to inform—to provide information on what an individual student needs to practice, should be retaught, or is ready to learn next, and how students themselves might become formative users of assessment information.
Good assessment starts with a clear purpose. If a test constructed far from the classroom has as its primary purpose to sort students from highest to lowest, it has some, but only limited, value in shaping what happens next in the classroom. It is even less able to help students decide what steps they can take to improve.
Any test can be used to some extent in a formative way: to help shape instruction, to identify curricular strengths and weaknesses, or to inform students of what they know and don’t know, or can and cannot do, at a given point in time. Any formative benefit for teachers and students, however, is dependent upon the quality of the test itself. If the test has been constructed with items and tasks adhering to certain standards of quality, the likelihood is increased that the test will measure student learning accurately.
But what if it hasn’t? And what if we are simply administering more and more tests in the name of formative assessment?
We may want to watch out for what could be sacrificed unintentionally on our way to adequate yearly progress.
Tests do not automatically yield dependable results; the purpose of the test and the uses of its results, the type(s) of achievement targets to be assessed, the method(s) to be used, individual item quality and clarity, sampling, and control for bias are just some of the necessary considerations if we expect dependability. The more we use and rely on short-cycle, common, benchmark, or interim assessments to generate the data we need to track student progress toward the state standards, the more it is incumbent upon all of us to ensure their accuracy. The quality of the decisions we make about and on behalf of students is directly related to the quality of the source of the information.
Should we begin to lose sight of what formative assessment really looks like, and what it can do to help teachers and students improve learning, we can remind ourselves of the way it is practiced in countries other than the United States: as assessment for learning. Assessment for learning is based in the classroom. It involves students in every aspect of their own assessment, including record-keeping and communication, and helps teachers teach and students learn. Assessment for learning rests on the understanding that students, not just adults, are data-based instructional decisionmakers. In that sense, assessment for learning cannot be reduced to a package purchase. It is a human process: teachers and learners together generating accurate information about student learning and then using it effectively to promote even greater learning.
We should not attempt to teacher-proof the assessment process by removing the classroom teacher’s assessment responsibilities and replacing them with a series of ready-made tests. Should we do so, we risk closing the door on assessment’s being anything more than an end in itself. We’ll miss seeing assessment become the integral, daily part of teaching and learning it can be, when used by teachers as both instruction and measurement.
When done right and used well, short-cycle and common assessments have value and contribute to the improvement of learning and of schools. But no number of summative tests, even those of high quality—and especially those mislabeled as formative classroom assessments—can replace the information an assessment-literate teacher can produce about and with students through the classroom assessment process.