Test Score Data to Improve Instruction
July 1, 2003 |
Teacher Use of High-Stakes Test Score Data to Improve Instruction
by James H. McMillan and Susan McKelvey (July 2003)
Across the United States most school districts and state departments of education have embraced high-stakes testing for their public schools and students. This has led to widespread school reform since many students have not reached the standards that the states and/or districts have set. Teachers increasingly focus on making sure that their students have enough knowledge to pass these tests, and some states and districts have made passing scores a requirement for graduation. Many states and districts have invested considerable time aligning their standardized tests with the objectives of the curriculum.
One of the consequences of high-stakes testing is that teachers have become more accountable for what they do in their classrooms. Of particular relevance to this study is the professional development that teachers engage in to understand and use students’ standardized test scores. The emphasis is on helping teachers review the scores and data of these tests to help them make informed decisions about their instructional practices. This research, then, addresses the use of standardized test scores by teachers to improve instruction and student learning.
In Virginia, state high-stakes tests are called Standards of Learning (SOL) tests. These tests are currently administered in grades 3, 5, and 8 and in the four major subject areas (English, mathematics, science, and social studies), and at the end of high school courses. Schools and school divisions are provided test result data at the end of the school year or in early summer for school year just ended. Further distribution of data to teachers varies, and is the focus of this study.
The purpose of this research is to determine the extent to which teachers have used SOL test data to change or modify instruction, and to identify procedures that have promoted effective and accurate use of SOL test data.