Harold C. Doran and Lance T. Izumi, Pacific Research InstituteJune 2004
Like most states, California's current assessment system lacks value-added measures. Rather than tracking test scores of individual students over time to gauge learning gains each year, most states take "snap-shots" of class and school averages, a method that can neither account for changes in student populations (is a school really doing better if this year's fourth graders score better than last year's fourth graders?) nor measure a student's learning each year (a single score reflects a lifetime of learning, not just one year). Equally problematic, California's system, like others, tracks the proportion of students meeting proficiency targets; this ignores gains, or the lack thereof, made by students well above or below the proficiency cutoff (akin to "measuring a child's height with a yardstick but acknowledging growth only when his or her height exceeds 36 inches"). To rectify these problems, California legislated in 2002 that the Department of Education seek proposals for a student achievement tracking database. In 2003, it required creation of a system of unique student identification numbers and created a committee to recommend how to use these in assessment. This paper presents one potential solution: the REACH (Rate of Expected Academic Change) value-added model. It would track student achievement data over time to create a trend line, predicting how a student will fare by the end of his or her schooling compared to a "proficient" score. A student on course to reach 80 percent of proficiency would get a 0.8, while one on track to exceed the standard would score above a 1.0. Because it would track individual students, it would measure how much they gain each year relative to the end goal (proficiency). Much of this paper is technical, and you may not like its whiff of determinism, but it offers the lay reader a useful overview of value-added testing (including the Tennessee model) and of California's existing assessment system. Other value-added models besides REACH may yet be proposed for California, but for now they appear to have at least one tantalizing option on the table. To read more, click here.
