We present a method for data-based decision making at the school level using student achievement data. We demonstrate the potential of a national assessment database [i.e., the University of Oregon DIBELS Data System (DDS)] to provide comparative levels of school-level data on average student achievement gains. Through the DDS as a data source, and the analytic methods we outline, we illustrate one way that schools can engage in system-level formative evaluation by examining their students' gains across an academic year conditional on initial skill level relative to the performance of a large sample of other schools. We provide the empirical Bayes estimates of school-level effects and their associated standard errors for second grade, DIBELS oral reading fluency using a percentile band plot. We illustrate a practical way that schools could use this output to improve their data-based decision making procedures.