|
February 2003
| Volume 60 | Number 5
Using Data to Improve Student Achievement
Blind Data
Marge Scherer
How Classroom Assessments Improve Learning
Thomas R. Guskey
Assessments can help improve education, but if we use them only as a means to rank-order schools and students for the purposes of accountability, we will miss their most powerful benefits. The author argues that teachers who develop classroom assessments that are useful for both students and teachers, provide high-quality corrective instruction, and give second chances to demonstrate success can improve their instruction and help students learn.
A Reader's Guide to Scientifically Based Research
Robert E. Slavin
Beginning with the Comprehensive School Reform Demonstration legislation in 1997, and continuing with the No Child Left Behind Act, the federal government is placing more and more importance on scientifically based research. But what constitutes good research? Why is one study more valid than another? Random assignment to experimental and control groups is optimal, but difficult and expensive. Research dependent on cherry picking, bottom fishing, and pre-post comparisons should be avoided. As the education community becomes more and more astute at discriminating research findings, the outcomes of education for students will markedly improve.
Lessons from Research That Changed Education
Gordon Cawelti
New federal legislation has emphasized using rigorous scientific research to evaluate teaching practices and curriculum materials. Research policy changes that might discourage innovative approaches to research by insisting on particular requirements, however, would be a mistake if the result were fewer studies like the most influential studies of the past half-century. The author reviews 11 research and research-based formulations that have influenced U.S. Supreme Court decisions, curriculum and instruction, early childhood education, and policies regarding small class size, standards-based reform, and productivity. He points out that few met strictly rigorous research requirements of research and that many were formulations based on earlier research or seminal studies that influenced later studies. He suggests that the task of improving government-sponsored research should involve replacing bureaucratic inefficiency and political intrusions with swifter and more responsive processes and structures to ensure high-quality research, make findings more readily accessible to practitioners, and encourage developmental and follow-up studies. In the end, he says, we should all recognize that great research, like great art, will always need room for variations, new approaches, initial trials, and later refinement.
First Things First: Demystifying Data Analysis
Mike Schmoker
The experts' tendency to complicate the use and analysis of student achievement data often makes it difficult for educators to understand data or to use it effectively, writes the author. He recommends focusing data analysis on the need of teachers to answer two basic questions: “How many students are succeeding in the subjects I teach?” and “Within those subjects, what are the areas of strength and weakness?” The answers to these questions enable teams of teachers to set a limited number of simple, specific improvement goals. They can then focus their efforts on the real work of instructional improvement: “the collaborative effort to share, produce, test, and refine lessons and strategies targeted to areas of low performance, where more effective instruction can make the greatest difference for students.”
No Schools Left Behind
Victoria L. Bernhardt
What kinds of data are important for continuous school improvement, and how can educators best organize the data for easy access and analysis? The author provides examples of how the careful gathering, intersecting, and analysis of four kinds of data—demographics, student-learning measures, perceptions, and school processes—reveal useful information for improving schools. The author suggests several features that schools should look for when they acquire a data warehouse or data analysis tool. She notes that many schools already have these data and need to work on making them more accessible and usable.
The Effects of High-Stakes Testing on Student Motivation and Learning
Audrey L. Amrein and David C. Berliner
“If statewide high-stakes testing policies actually improve student learning, we should see that improvement reflected not just in the state's own test scores but also in independent measures,” write Audrey L. Amrein and David C. Berliner. The authors look at 18 states that require students to pass state examinations to graduate from high school. For each state, the authors evaluate how student achievement—as measured by the SAT, the ACT, the Advanced Placement tests, and the National Assessment of Educational Progress—related to the implementation of the graduation exams. They found no evidence that in states using their own tests, student achievement increased on those other measures. In addition, the researchers report data on dropout rates, retention in grade, and number of students obtaining alternative high school diplomas to suggest that high-stakes testing pushes low-achieving students out of school rather than helping them succeed.
The Power of Testing
Matthew Gandal and Laura McGiffert
The authors draw on the parallels between medical tests and education assessments to argue that well-designed tests are an essential part of effective education reform. Just as medical tests measure a patient's condition against accepted standards of health (such as cholesterol count or blood pressure), so education assessments should measure students' performance of common, high achievement standards. Just as medical tests provide information that is relevant to the patient's specific condition, so education assessments should be tightly aligned with standards. And just as in medicine, assessments in education are only a means to an end. The important question is what happens when the results come back. Therefore, useful education assessments provide timely, specific, meaningful results that educators can use to target instruction to students' needs.
The Dangers of Testing
Monty Neill
Research and experience have shown that high-stakes testing harms schools rather than improving them, argues the author. He reviews evidence that high-stakes testing does not raise student achievement, but instead narrows the curriculum and dumbs down instruction. As an alternative, he points to the success of schools that use formative assessment practices to encourage skilled teaching and higher-level learning.
The Seductive Allure of Data
W. James Popham
Despite the claims of their proponents, most state accountability tests fail to produce the kinds of data that teachers can use to improve student achievement, says W. James Popham. After trying in vain to make sense of these data, teachers may become disillusioned about the instructional use of data. But the right kinds of assessment data, which most often come from properly designed classroom assessments, can help teachers improve student achievement. Popham discusses five attributes of instructionally useful tests: significance, teachability, describability, reportability, and nonobtrusiveness. In addition, classroom assessments used to evaluate teacher effectiveness must produce credible data that will convince an objective observer that student learning has improved.
Backward Design for Forward Action
Jay McTighe and Ronald S. Thomas
Looking back to the key concepts and essential questions that underlie content standards can help identify learning goals and be the starting point for planning both curriculum and school improvement. This framework works for both curriculum design and school and district improvement efforts. Through this unified process, school teams determine learning goals; collect, analyze, and summarize evidence from multiple sources of data to determine how well students are doing on external accountability tests and the extent to which they really understand what they are learning; consider the root causes of present achievement; and then—and only then—implement systemic actions to address root causes, promote enduring learning, and increase test scores. In this way, schools and districts can identify priorities, monitor results, and target actions that improve student learning.
Using Data: Two Wrongs and a Right
Robert J. Marzano
Being data driven is an admirable goal. Just because a school collects data, however, does not mean the data are being used to improve student achievement. Problems mount if the data are not direct measures of student learning or if no explanatory model articulating the factors that affect student achievement accompanies the data. On the basis of 35 years of research, the author identifies 11 factors as the primary determinants of student achievement, including those tied to school-level policy and practice, teachers, and students.
Data Warehousing: Beyond Disaggregation
Lawrence M. Rudner and Carol Boston
The No Child Left Behind legislation requires both more disaggregated data and a yearly progress report. Educators are scrambling to ensure that their data collection and analyses comply with the new requirements, but the authors recommend that educators should instead seize this opportunity to retool and systematize their data collection, reporting, and analysis. They point out that a well-organized, easily accessible data warehouse allows a wide range of analyses using both cross-sectional and longitudinal data. The authors describe the value of combining these kinds of data for a more complete and precise picture of how the schools are doing and how they can improve.
A Tale of Two Schools' Data
Beverly A. Parsons
Acknowledging that implementation takes time safeguards district from discarding effective practices by mistake. Joseph and Zindzi—two math teachers from similar schools in different districts—received their students' less-than-stellar achievement scores. Joseph didn't know what to do. The instructional changes he made the past year seemed to have no effect, and his district provided little guidance. Zindzi remained optimistic. Her district formed teams to research and advise on best practices and then to evaluate test data according to the degree to which teachers had implemented the new practices in their classrooms. Connecting instruction, professional development, and student learning makes all the difference between successful and unsuccessful data use.
Using Data to Differentiate Instruction
Kay Brimijoin, Ede Marquissee and Carol Ann Tomlinson
One teacher illustrates how using pre-assessments, self-assessments, and ongoing assessments in her classroom enables her to differentiate instruction and meet the needs of individual learners. Informal and formal assessment helps students recognize what they do and don't understand and helps teachers modify instruction so each student is appropriately challenged.
A Global Perspective on Student Accountability
John H. Holloway
Mental Health Specialists in Schools
Steven C. Schlozman
Reviews
Letters
EL Themes for 2003–2004
ASCD Community in Action
Developing an Inquiry-Minded District
Jay Feldman, Gail Lucey, Sarah Goodrich and Dana Frazee
By engaging in ongoing data analyses, teachers can develop plans to address barriers to student learning. The data-based inquiry and decision making process involves a group of teachers setting a vision, collecting and analyzing data, determining strengths and challenge areas, planning action, and making annual assessments of those actions. School districts can support teachers and help create a culture of inquiry by organizing and disseminating data, providing time and other resources for teachers, and sharing information across all district schools.
Study Guide / EL Extra
Web Wonders / Using Data to Improve Student Achievement
Deborah Perkins-Gough
Copyright © 2004 by Association for Supervision and Curriculum Development
|