SOE News

William Sanders describes value-added assessment for measuring student progress

Speaking to a capacity crowd at the School of Education on March 21, 2006, William Sanders described his work over the last 20 years developing a methodology to measure the influence of school systems, schools and teachers on the academic progress of students.  

His system, known more generally as “value-added assessment” is designed to reduce the error of measurement that arises in any single pre- and post-test situation by creating a large information array for each student and populating it with data from many achievement measures. In order to be included in his approach, Sanders recommends that a measure meet three criteria: It must be aligned with the school’s curriculum; it must have appropriate reliability, and it must have sufficient stretch to measure high-achieving and low-achieving children in the same grade level.

“The more information you can pack into the array for each kid, the more precise you can be in following the progress of that kid as an individual,” Sanders explained. “By doing so, you can begin to measure the impact of various educational influences on the rate of progress of populations of kids.”          

The approach is based on rigorous, documented and well-published statistical theory and methodology that have been in the statistical literature since the late 1950s, according to Sanders. He has brought that theory and methodology ─ based on mixed-model equations and techniques ─ to the problem of measuring academic growth over time for students and groups of students and evaluating the effects of teachers, schools and school systems on that growth.

By providing a longitudinal view of student progress, value-added assessment can help school leaders and teachers make better informed instructional decisions. During his presentation, Sanders showed examples of projections of students’ future academic success based on their current progress. He demonstrated how the information can help educators identify students who are not making gains, obtain better diagnostic information about those students and help them succeed. 

He also discussed how the test findings can reveal useful information about teacher effectiveness. “For example, on average, a teacher’s effectiveness improves steadily from the first year of teaching to the tenth; then it plateaus,” he said. This kind of information can be helpful to schools because, as many studies have documented, the sequence of teachers a student has over time will have an enormous effect on the student’s academic career, especially in mathematics.

Sanders encouraged educators not to focus on minimum standards, but rather to strive toward various endpoints. By using a value-added assessment methodology, for example, teachers and educational administrators can determine the likelihood that a student will be successful in multiple academic settings ─ from a remedial college mathematics class to an advanced college calculus course. Sanders urged educators to use this type of diagnostic information to stimulate students’ academic growth and help them reach their highest endpoint.

Sanders is a senior research fellow with the University of North Carolina system and manager of value-added assessment and research for SAS Institute Inc. in Cary, N.C. Also participating in his presentation was Sanders’ wife, Dr. June Rivers, who also works in value-added assessment research at SAS.