Representing learning progression through activity monitoring with a focus on “learning from failure”

Abstract

This study addresses the challenge of characterizing learning progression in self-regulated learning environments only by activity monitoring. Existing studies often overlook self-regulated learning environments without predefined mastery criteria. The study fills a gap there by emphasizing learning from failure and offers a novel perspective on understanding learning dynamics in less structured settings. It empirically analyzes whether graded self-reporting can adequately capture students’ learning progression using available student logs, and shows that graded self-reporting better reflects grades and the learning process quantitatively. It also introduces learning progress graphs generated from action logs to quantify and visualize learning progression, aiming to capture individual learning trajectories and to identify wheel-spinning in unguided contexts. Through two evaluations, involving educational technology researchers and experts, the study assesses the adequacy and interpretability of learning progression graphs for detecting wheel-spinning. Findings suggest that learning progression graphs effectively identify wheel-spinning on the basis of learning progression graphs, highlighting the importance of understanding student active steps without predefined target criteria. However, limitations include reliance on self-assessment, limited system recommendations, and the absence of teacher feedback. The study also notes challenges in interpreting learning progression without predefined guidance.