Analyzing Handwritten Log Data for Evaluating Computational Thinking: Investigating State Transitions in Drawing Tasks Using Pen-Based Tablets

This study proposes a method for estimating computational thinking abilities using handwritten logs in a drawing task with a layer-based tool. Computational thinking, vital in twenty-first-century education, involves problem-solving and strategic planning. The approach involves assessing participants’ foresight by observing transitions between layers during sketching, line drawing, coloring, and parts. Handwritten logs, obtained from an iPad and Apple Pencil, include position coordinates, pressure, altitude, azimuth angles, and speed. A machine learning model is trained to distinguish users with and without foresight, identifying features indicating computational thinking abilities. Experiment results suggest a potential correlation between foresight and computational thinking during drawing tasks. Further exploration is necessary to assess the method’s generalizability across diverse tasks and participants.