How to Schedule Multiple Graphical Representations? A Classroom Experiment With an Intelligent Tutoring System for Fractions

Abstract

Providing learners with multiple representations of the learning content has been shown to enhance learning outcomes. When designing problem sequences with multiple representations, designers of intelligent tutoring systems must decide how to schedule the representations. Prior research on contextual interference has demonstrated that interleaving different types of learning tasks can foster a deep understanding of the underlying concepts. Do the same advantages apply to interleaving representations? In a classroom experiment, we compared four conditions that varied the practice schedules of multiple graphical representations between interleaving and blocking. The multiple-representation conditions were compared to three single-representation control conditions. During their regular classroom instruction, 290 4th and 5th-grade students worked for five hours with versions of an intelligent tutoring system for fractions. On several dependent measures, interleaving multiple graphical representations led to better learning results than blocking multiple graphical representations. Findings from a think-aloud study give insight into the underlying cognitive processes.