The Evolution of Mastery Learning: Challenges, Technologies, and the AI Journey Leading to the Intelligent Tutoring System DARTS

Abstract Bloom’s groundbreaking theory of Mastery Learning (1968) and his “2 Sigma Problem” study (1984) revealed that one-on-one tutoring could produce learning outcomes two standard deviations above conventional classroom instruction. However, while pedagogically compelling, the approach was economically and logistically unfeasible at scale (Guskey, 2007). Despite its proven effectiveness, widespread implementation of Mastery Learning remained difficult to operationalize at scale due to practical constraints—chiefly, the high cost and logistical challenge of providing personalized instruction and timely feedback. The emergence of artificial intelligence marks a pivotal point in this evolution, offering the capability to replicate the adaptive, responsive, and personalized instruction traditionally delivered by human educators (VanLehn, 2011; Ma et al., 2014). After decades of incremental innovation, the evolution comes full circle with DARTS—Dynamic Academic Response and Tutoring System—an AI-powered mobile Intelligent Tutoring System (ITS) designed to deliver the benefits of Bloom’s model through personalized, conversational tutoring. This study culminates in the conceptualization of DARTS, which bridges the original promise of Mastery Learning with modern scalable technology. The authors plan a three-part article series that will examine DARTS in depth—its theoretical foundations, technical design, and implementation—illustrating how it addresses longstanding challenges in delivering scalable, individualized instruction.