EXP: Modeling Perceptual Fluency with Visual Representations in an Intelligent Tutoring System for Undergraduate Chemistry

Students learn science through visualizations. For example, students use pie charts to visualize fractions and ball-and-stick models to visualize chemical molecules. To help students learn with visual representations, research has focused on one representation skill: conceptual understanding of representations. However, a second representation skill is equally important for STEM success: perceptual fluency, or the ability to rapidly see meaning in representations and to effortlessly translate among multiple representations. For example, expert chemists can translate between visual representations as fluently as bilinguals translate between languages. Perceptual fluency frees cognitive resources to engage in higher-order conceptual thinking, inference making, and creativity. 

The goal of this project is to develop adaptive support for perceptual learning tasks. The project will build cognitive modeling techniques that have well-documented advantages for supporting conceptual learning. We will expand cognitive modeling to perceptual learning. To this end, we will combine machine learning and experimental psychology using Chem Tutor, an adaptive educational technology for undergraduate chemistry. We will test the hypothesis that adaptive perceptual fluency trainings enhance undergraduate chemistry learning.