MISSION | CONSTITUTION | EXECUTIVE BOARD & STAFF | REGIONS | ADVOCACY
- NCLB relies too heavily on single assessments administered once each year. One test does not present an accurate or reliable indication of progress.
- Better assessment systems with multiple measures mean better accountability for school improvement and better data that educators can use to help each student succeed.
- We can do better than evaluating schools, students, and reform methods on the basis of a narrow, single-test assessment system. We need increased flexibility to incorporate school-level assessment data that is based on the most accurate indicators of success.
Current Status
Under the U.S. No Child Left Behind Act (NCLB), schools and districts must meet adequate yearly progress (AYP) requirements by demonstrating students' proficiency on mandatory assessments each year. Currently, NCLB requires assessments in math and literacy in grades 3–8 and 10. In 2007–2008, those requirements will increase when science assessment is added once in grades 3–5, 6–9, and 10–12. Although NCLB is not slated for reauthorization until 2007, the testing provisions of the law may be addressed in the coming year through other legislation or regulations.
The U.S. Department of Education, for example, has approved a pilot program that allows 10 states to develop growth models of assessment. Growth models allow schools and districts to meet accountability requirements by tracking individual student progress over time, as opposed to comparing the performance of one year's group of students to that of the following year's students.
ASCD Supports Multiple Measures of Assessment
ASCD supports assessments and accountability based on valid and reliable information from multiple data sources. When done properly, measuring student learning is an essential tool in educating children. The use of multiple and/or formative assessments provides appropriate evidence of student learning to ensure accountability to students and the district community. The data from these assessments can be used to help modify instructional practice to better meet students' individual learning needs.
Unfortunately, many of the assessments used to make high-stakes decisions do not meet these criteria. NCLB places too much emphasis on single assessments administered once each year. These single assessments make up a narrow accountability system that is inadequate for measuring the full breadth of student learning. What's more, the data from these assessments are often insufficient to help teachers improve their instruction. In some states, teachers receive information only about the percentages of students who fail certain tests, without the detailed information they need to help those students succeed.
Researchers from the University of Washington found that "because of the way most high-stakes assessments are designed and the way scores are reported, students with . . . vastly different strengths and weaknesses can actually receive the same score." The lack of complete information means that teachers are unaware of differences among students that would call for dramatically different interventions. According to the researchers, "unless we look beneath test scores, at individual students and the conditions that foster high-quality learning, more and more students will continue to fall below the bar" (Valencia & Riddle Buly, 2004).
The U.S. Department of Education Web site claims that annual testing provides information that helps teachers improve student performance and diagnose problems. This is incorrect. As assessment expert W. James Popham (2006) writes in Educational Leadership, the vast majority of state standardized tests used for accountability under NCLB are instructionally insensitive—"they're unable to detect even striking instructional improvements when such improvements occur."
The U.S. Department of Education Web site also states, "If a single annual test were the only device a teacher used to gauge student performance, it would indeed be inadequate. Effective teachers assess their students in various ways during the school year." If the Department of Education admits that a single test is inadequate as a comprehensive student assessment, then it stands to reason that a single test is inadequate as the sole determinant of school performance under NCLB. According to Stephen Raudenbush (2004), evaluating schools with the tests typically used for NCLB is "scientifically indefensible."
ASCD calls upon the U.S. Congress and the U.S. Department of Education to allow the use of multiple assessments for evaluating student growth and performance. Creating a more sophisticated assessment system is crucial to improving the education of all children. Such a system will ensure not only that important accountability decisions are based on the best possible data, but also that data can be used by educators to determine which students are struggling, what strategies are working to help those children, and how they should adjust their instruction to ensure that each child succeeds.
As Popham argues, the current assessment system may push out forms of assessment that can demonstrably help students learn. A better assessment system would help states employ instructionally sensitive accountability tests capable of detecting the effect of first-rate teaching (Popham, 2006). Additional flexibility that the U.S. Department of Education has provided for states to explore value-added assessment is only a beginning. ASCD also calls for flexibility that will allow additional assessment data to be incorporated at the school level, including portfolio assessments and other authentic assessment data. Improvements to the assessment system must also allow for the use of formative assessments.
As the directors of the National Center for Research on Evaluation, Standards, and Student Testing have said, "A single test cannot address all that is important for students to know and be able to do. Multiple measures are needed to address the full depth and breadth of our expectations for student learning." Beyond multiple-choice and short-answer items that are typical of current assessments, "other types of performance measures—essays, applied projects, portfolios, demonstrations, oral presentations, etc.—are needed to represent and guide students' progress" (Herman, Baker, & Linn, 2004).
ASCD is opposed to federal sanctions that are determined by performance on a single assessment. Using a single test as the determinant for student, individual school, and school district performance does not present an accurate assessment and will result in inappropriately and inaccurately labeled students, schools, and school districts.
References
Herman, J. L., Baker, E. L. & Linn, R.L. (2004). Accountability systems in support of student learning: Moving to the next generation. CRESST LINE. Los Angeles, CA: University of California, National Center for Research on Evaluation, Standards, and Student Testing.
Laitsch, D. (2005, July). A policymaker's primer on testing and assessment. Infobrief #42. Alexandria, VA: Association for Supervision and Curriculum Development. Available at http://www.ascd.org/newsletters
Popham, W. J. (2006, February). Assessment for learning: An endangered species? Educational Leadership, 63(5), 82–83.
Raudenbush, S. W. (2004). Schooling, statistics, and poverty: Can we measure school improvements? Princeton, NJ: Educational Testing Service, Policy Evaluation and Research Center, Policy Information Center.
U.S. Department of Education. Testing for results: Helping families, schools, and communities understand and improve student achievement. Retrieved February 14, 2006, from www.ed.gov/nclb/accountability/ayp/testingforresults.html
Valencia, S. W. & Riddle Buly, M. (2004). Behind test scores: What struggling readers really need. The Reading Teacher, 57(6), 520–531
|
