Using Data Mining to Identify Actionable Information: Breaking New Ground in Data-Driven Decision Making

The implementation of No Child Left Behind (NCLB) presents important challenges for schools across the nation to identify problems that lead to poor performance. Yet schools must intervene with instructional programs that can make a difference and evaluate the effectiveness of such programs. New advances in artificial intelligence (AI) data-mining software can aid in identifying important indexes of achievement to help teachers and administrators improve these instructional and programmatic interventions. The problem addressed in this study is the difficulty that school leaders face in using the stores of data they have already collected to analyze the effectiveness of interventions focused on improving achievement. The essential question is whether educators can predict student achievement from all of the disparate variables already stored in typical data warehouses. Support for this study was provided by SPSS, Inc., a partner in the research, who provided the University of Connecticut with several licenses for its premier data-mining application-Clementine 8.5.