Email Status

Volume 27, Number 4
July/August 2011

The Common Core State Standards

Challenges for assessment


Our nation is currently caught up in an enormous educational challenge: to see whether the vast majority of our states cannot only adopt identical curriculum standards in mathematics and English language arts (ELA) but also devise suitable instructional and assessment systems linked to those standards. Because these assessments will most likely become the accountability tests routinely used in evaluating the success of most U.S. schools, the impact of these tests on American education is potentially enormous.

The Common Core State Standards (CCSS), produced at the initiative of the Council of Chief State School Officers (CCSSO) and the National Governors’ Association, have been adopted by all but a few states. Currently, two assessment consortia, each composed of several dozen states, have received substantial federal funding to create assessments that can determine the degree to which students have mastered the intended learning outcomes embodied in the CCSS. These outcomes—the knowledge and skills the CCSS authors believe students should achieve as a result of schooling—can be labeled content standards, instructional targets, or, as I will call them here, curricular aims.

In the past, we have seen inappropriately designed standards and assessments torpedo many promising school reform efforts. Such setbacks frequently arise when teachers are asked to teach more than they can cover in the allotted instructional time or to spend too much time on testing. As currently conceived, the CCSS represent an iceberg awaiting its Titanic. To illustrate, in a single academic year, a fourth-grade teacher is supposed to ensure that her students master about 65 significant math and ELA curricular aims. Because the common standards cover so many curricular aims, the two consortia charged with developing assessments—the Partnership for the Assessment of Readiness for College and Careers and the SMARTER Balanced Assessment Consortium—will surely run aground if they try to measure every single one. Instead, they need to carefully conceptualize what’s expected of students based on these standards, identify a small number of key curricular aims, and build their assessments around those.

Curricular Eyes Meet Instructional Stomachs

In almost every state, when curriculum specialists identify what they want their state’s students to learn, those specialists come up with far too many targets. Indeed, most states’ officially approved standards have tended to resemble wish lists rather than realistic collections of what students can actually learn in school. The resultant numbers of wished-for curricular aims invariably turn out to be too many to successfully teach (at any depth) during the available instructional time or to assess appropriately during the available assessment time. I have bemoaned elsewhere the negative results when educators’ curricular eyes are bigger than their instructional and/or assessment stomachs.

With much applause from most bystanders, the architects of the CCSS decided to create curricular targets that were “fewer, clearer, and higher.” Yet, while creators of the CCSS have done reasonably well on the “clearer” and “higher” dimensions, they appear to have scored below basic when it came to “fewer.” Put candidly, there are simply too many curricular aims—represented at varying levels of generality—now contained in the CCSS.

One significant function of curricular aims is to help teachers know where to direct their instructional efforts. If too many curricular aims exist, the profusion of goals will most certainly overwhelm both teachers and students. By the same token, when assessment specialists are presented with too many curricular aims, they cannot include enough items to provide accurate information about students’ mastery of each aim without making the test prohibitively long. This is the problem that now confronts the two CCSS assessment consortia.

Four Steps to Formulating Appropriate Aims

There are four steps the consortia should follow to devise a set of assessment-appropriate curricular aims.

1. Choose a manageable number of curricular aims to assess. Teachers—and students—benefit from having a clear idea of what should be learned by the end of a semester or a school year. For example, a fourth-grade teacher who wants her students to master eight important mathematics aims and seven significant ELA aims can use these targets as a guiding curricular framework for an entire school year. Typically, a teacher can make sense of, accurately comprehend, and keep track of ten or so curricular targets. Few teachers can make sense of, accurately comprehend, and keep track of four or five dozen curricular targets.

Another substantial payoff from focusing on some, not all, CCSS aims is that it will allow these aims to be assessed more accurately. With fewer aims to assess, tests can include more items per aim. By including six to ten items per aim, we can often get a reasonable fix on a student’s mastery of each assessed aim—or at least one that is more valid than an inference based on one or two items, as is currently the case in most high-stakes tests. Nor will computer-adaptive assessment necessarily solve this problem (see sidebar “Are Adaptive Tests Instructive?”). Are Adaptive Tests Instructive?

The use of computer-adaptive testing has been promoted particularly by the SMARTER Balanced Assessment Consortium. In this kind of testing, computers magically (that is to say, algorithmically) choose test items based on a student’s previous answers so that questions are matched to the appropriate level of difficulty for each student. However, for this kind of testing to work, all of the items on the test must be linked to a single variable, such as “mathematical understanding” or “reading comprehension,” and arranged on a single vertical scale. Weak students will get easy items on that scale and strong students tough items that have a higher place on that same scale.

The only purpose of computer-adaptive testing is to rate a student’s performance on the vertical scale. There is no interest—at all—in the isolation of subskills or bodies of enabling knowledge. This sort of instructionally pertinent information is irrelevant. Indeed, if one were to try to include enough items to reveal how students had scored on particular subdimensions of the test, it would almost certainly distort the nature of the vertical scale and therefore defeat the purpose of adaptive testing. As a result, we can’t tell which math or reading skills a student is struggling with.

Computer-adaptive tests do save time: students spend less time answering questions that are too easy for them or struggling with questions that are too hard. But for computer-adaptive tests to have a truly positive impact, they must be accompanied by formatively focused classroom assessments. Such classroom assessments are part of the SMARTER Balanced Assessment Consortium’s game plan.
2. Designate only the most important curricular aims for assessment. If one accepts the need to focus on only a modest number of curricular aims, it then becomes imperative that those aims measure the most important things students of a given age should know and be able to do. To illustrate, in ELA we surely want students to be able to write original compositions and to do so with increasing sophistication as they grow older. In mathematics, we want students to be able not only to perform the four basic arithmetic operations properly but to be able to employ those operations to solve problems the students have not previously encountered.

Accordingly, the CCSS aims must be rigorously prioritized. Because of the way the CCSS aims are currently formulated, this process has to involve more than just picking one target over another. In a number of instances, it will be necessary to coalesce some of those targets into more comprehensive curricular goals that reflect, as faithfully as possible, the aspirations and preferences of the architects of the CCSS. Such a coalescence strategy could permit both enhanced instruction and more accurate evaluation. (See attribute 4 in this list regarding how such subsuming might be accomplished.)

3. Ensure that the selected curricular aims can be addressed instructionally. Those who are considering a curricular aim as a potential target for assessment must give at least some thought to the following question: “How can students be taught to master this aim?” Targets should be set by not only ELA and mathematics subject matter experts but by experienced teachers who can judge how much students of a given age can be taught during the instructional time available and who can ensure that the stated aims reflect student learning—not innate ability, life experience, or their family’s affluence.

4. If the curriculum aim includes subcomponents—subskills or bodies of enabling knowledge—ensure that the aim can be assessed both holistically and analytically. This final attribute is particularly important but somewhat difficult to explain. Perhaps an example will help. When we assess writing skills by asking each student to compose an essay, we may arrive at an overall judgment (holistically) about the quality of the student’s writing, but we can also reach separate judgments (analytically) regarding factors such as organization, mechanics, and content knowledge. Our overall judgment hinges on a composite appraisal of multiple distinct dimensions, each of which can be separately addressed instructionally.

Sometimes, however, a so-called curricular aim does not represent a single target but instead embraces several quite different subcomponents. This situation often occurs when curriculum architects rely on “strands” to describe a set of related, but dissimilar, cognitive skills. Too often curricular strands conceal—not clarify—the nature of what we expect our students to accomplish.

The reason this fourth attribute is so important hinges, as usual, on the instructional consequences. Assessment of an overarching curricular aim should help teachers discern which subskills or bodies of enabling knowledge appear to be causing the more general problem. Once diagnosed, teachers can then zoom in instructionally to deal with those areas of weakness.

Only One Bite of This Apple

It is impossible to imagine a higher-stakes set of educational tests than those soon to be created by the two assessment consortia. If done well, these new tests could become catalysts for a new era of more effective schooling. If done badly, however, the new tests will have wasted an enormous amount of money. Even worse, they will have deflected America’s educators from pursuing other potentially effective strategies for improving our schools.

Either way, the two consortia will have only one shot at success or failure. Too much is at issue here for national policy makers to allow an unsuccessful consortium a do-over. Accordingly, the careful selection of curricular aims they plan to measure is altogether critical. Each of the curricular aims chosen as an assessment target must satisfy all four of the attributes identified here.

This wonderful opportunity will come our way but once. We must not muff it.

W. James Popham is professor emeritus at the University of California, Los Angeles, School of Education and Information Studies.

For Further Information

For Further Information

W. J. Popham. Unlearned Lessons: Six Stumbling Blocks to Our Schools’ Success, Cambridge, MA: Harvard Education Press, 2009.

W. J. Popham. “Exposing the Imbalance in ‘Balanced Assessment.’” Evidence-Based Education (Spring 2011): 14–15.