Across the nation, the movement for new forms of student assessment appears to have slowed and in some cases even reversed. Gov. Pete Wilson of California, for example, vetoed the reauthorization of the California Learning Assessment System, which included not only a substantially open-ended state exam, but also funded a wide array of local assessment-development projects.
Now California will pay local school districts to use off-the-shelf tests. That means most will use norm-referenced, multiple-choice exams. Though most educators oppose the use of such tests with young children, California’s new law says that a district must begin testing in grade 2, or it will not be reimbursed for any of its testing.
Arizona also halted its performance exam, for apparently political reasons. The National Assessment of Educational Progress may regress back toward more multiple-choice items. As a front-page story in the Jan. 31, 1996, issue of Education Week noted, “Caution is increasingly the watchword” toward performance assessment.
Moreover, a rising number of states plan to use test scores for very high-stakes decisions, such as determining high school graduation, after a period in which such practices declined. As student needs increase and demands for improvement escalate, budgets are tightening. Policymakers too often respond to this squeeze by focusing on raising scores on norm-referenced tests.
The resurgence of standardized testing may be great for test publishers, but it is not good news for students, educators, parents, or the nation. What seems to have been forgotten is why testing reform became an important issue in the early 1990s.
First, it became clear that traditional standardized tests were not able to measure many vital areas of student learning. Second, the narrowing and limiting effects of these exams on the curriculum became increasingly visible. Third, research on learning and human development indicated that the theory underlying standardized tests was, at best, inadequate. Fourth, research also demonstrated that the damaging effects of testing most powerfully affected low-income children and students of color. Students who tended to do least well on the tests were most likely to have their schooling reduced to drilling and test-coaching. Through its use in tracking and other important educational decisionmaking, and by reinforcing the ideology of the bell curve, testing also played a key role in limiting educational opportunities.
In response to these problems with standardized tests, interest in performance and portfolio assessment grew rapidly. As use of performance exams and open-ended tasks spread at the state level, however, a number of problems appeared. Teachers generally were not prepared to apply these methods, which meant extensive professional development was essential, but often not readily available. Use of performance assessments for accountability purposes required significant technological developments and sufficient time to work out problems, but the time was often not allowed. In many cases, parents and the community were not educated about the nature and purpose of performance assessment and the need for reform. While opponents took advantage of these issues in their attempts to discredit performance assessment, reformers debated how best to further implementation.
Beneath these problems, however, was a deeper unresolved issue: What should be the primary purpose of assessment? At the state level, accountability was usually the stated goal. But often an additional objective was to transform curriculum and instruction, that is, to change classroom practices. However, it was not clear whether, or to what extent, reliance on performance exams to reform curriculum and instruction would produce improved teaching and learning. Furthermore, developing performance assessment for high-stakes purposes within a short time frame often raised the difficult technical issues that opponents of reform used to discredit new assessments.
Assessment reform has thus reached a crossroads. One route leads backward toward multiple-choice, norm-referenced testing, perpetuating all the old problems. The other leads toward an expansion of performance assessment.
To reinvigorate the reform movement’s momentum and to guide the transformation of student assessment, the National Forum on Assessment, a coalition of major education and civil-rights groups, recently released “Principles and Indicators for Student Assessment Systems.” The forum believes performance assessment is essential if schools are to help all students gain rich, complex knowledge and understanding, and if schools are to become hospitable, engaging communities which support learning. Without improved assessment practices, too many students, trapped in schools focusing on multiple-choice tests, will not even learn the “basics” of fluent reading, clear writing, and competent use of mathematics.
The forum is clear on the essential purpose of assessment. The first two “principles” declare, “The primary purpose of assessment is to improve student learning,” and “Assessment for other purposes supports student learning.” To enact these simple statements, the “Principles” make classroom performance assessment central, require accountability to support good classroom practice, and propose constructing assessment systems on the basis of high-quality classroom assessment. This way, assessment can support learning, rather than being an end in itself, or worse, undermining education in the name of accountability.
The forum’s principles represent a profound departure from decades of practice. The traditional model was the externally constructed and mandated multiple-choice test used to measure learning at some end point, rank students based on their scores, and then make important decisions about individuals and programs. Classroom assessment came to mirror this form of externally mandated testing, rendering it not helpful for good instruction.
The paradigm detailed in “Principles and Indicators for Student Assessment Systems” integrates assessment with instruction and curriculum. Observation and documentation of student work over time provide the materials first to assist, then to summarize and evaluate, student learning. Projects, tasks, exhibitions, experiments, demonstrations, essays, performance exams, portfolios, learning records, and more are all part of the assessment mix. Within this context, tests, whether classroom or large-scale, are but one part of an array of assessment tools, rather than the dominant one. Examples of assessments that enhance learning are beginning to appear--indeed, the “Principles” document is based on emerging practice--and many of the recently developed national content standards have adopted this perspective.
In emphasizing that assessment must be reshaped to support learning, the National Forum on Assessment concludes that assessments constructed to rank and sort students should not be a significant part of any assessment system. Additionally, because of their often harmful effect on curriculum and instruction, the use of multiple-choice items should be sharply limited.
The principles also emphasize that important decisions should not be made on the basis of any single assessment. The process of gathering and evaluating evidence of learning over time provides a more powerful basis for understanding student accomplishments and making decisions. It also provides a rich basis for talking with parents about their children’s learning and informing the public about student achievement, as well as for involving the public in dialogues about educational goals and progress.
The process of developing classroom and school-based assessments is vital for ensuring equity. If teachers know how to look at each individual child, to know her strengths and ways of learning, his cultural background and interests, then they can work better with each student. When a teacher sees the full child, it is more difficult for biases based on ignorance or misinformation to persist. Teachers can more readily understand different paths to high-quality outcomes and therefore provide students with multiple ways to demonstrate their learning.
Doing high-quality performance assessment in the classroom does require substantial professional development for educators. But this can also have great value for teaching. For example, through discussing student work together, teachers can improve assessment practices as well as learn how to work better with particular students and improve curriculum. The discussion process itself also helps establish a culture of shared learning in the school.
Use of classroom-based assessment can enhance learning processes and outcomes, strengthen equity, improve instruction, and involve parents and the community more deeply. These processes are all essential elements of accountability. Accountability also requires that the public know about student learning. To meet this need, the principles recommend combining information from classroom-based assessments with data from external assessments such as performance exams. Since accountability does not require individual scores, sampling should be used to the extent feasible. Sampling can also allow expansion of the breadth and depth of learning that is evaluated.
It will take time to learn how to properly use performance exams and information taken from portfolios, learning records, and other in-class assessments. The relationship between classroom and external uses of assessment information will undoubtedly remain complex. But nothing indicates that problems in these areas cannot be resolved.
It is important also to remember that the changes in classroom practices that are at the heart of the principles of the National Forum on Assessment cannot be assumed to flow automatically from changing accountability exams. Simply basing external exams on genuine performances does not prepare teachers for the more fundamental assessment practices rooted in observation and documentation of student work over time.
At a minimum, an exam-driven approach requires extensive professional development and a clear plan for ensuring that changes in on-demand exams translate to improved curriculum, instruction, and assessment in the classroom, as well as ensuring that narrow teaching to new exams does not occur. Such scrutiny would be in line with the principles’ call for regular review of the quality and effects of assessments.
A retreat from high-quality practice--back to forms of schooling and testing that failed many children and rarely expected real thinking or use of knowledge--is a clear and present danger. That threat is most severe for low-income students and those from minority groups that have historically suffered discrimination, but it faces all students, and it must be combatted.
For reform to take root and thrive, a foundation must be built on changed classroom practices and community understanding. The assessment principles, signed by more than 80 education and civil-rights organizations and over 120 individuals, can be the basis for building renewed support for assessment in the service of real learning.
Copies of “Principles and Indicators for Student Assessment Systems” are available for $10 each (with bulk discounts) from FairTest, 342 Broadway, Cambridge, Mass 02139.
A version of this article appeared in the February 28, 1996 edition of Education Week as Assessment Reform at Crossroads