Evaluating Knowledge Gain and Retention in IoT Circuit Assembly Using Mobile Augmented Reality Technology

• Published in: Computer applications in engineering education Vol. 33; no. 3
• Main Authors: Lam, Meng Chun, Hadi, Hadi Bashar Khalid, Dahnil, Dahlila Putri, Suwadi, Nur Asylah, Abd Majid, Nazatul Aini
• Format: Journal Article
• Language: English
• Published: Hoboken Wiley Subscription Services, Inc 01.05.2025
• Subjects: Animation; Augmented reality; Comparative studies; Completion time; Effectiveness; Errors; human‐computer interface; Internet of Things; Knowledge; knowledge retention; learning; mobile‐based augmented reality; Retention; Taskload; Three dimensional models; Usability; user evaluation
• ISSN: 1061-3773; 1099-0542
• DOI: 10.1002/cae.70045

ABSTRACT Augmented Reality (AR) offers potential benefits in assembly training, yet there is a scarcity of research on knowledge retention when utilizing 3D model and animation overlays through Mobile Augmented Reality (MAR). This study investigates the influence of MAR applications, leveraging the signaling principle through 3D animated models, on knowledge gain and retention in a complex Internet of Things (IoT) assembly task. In this regard, this study developed a MAR framework and application to facilitate IoT assembly training. A comparative study was conducted with 40 participants, equally distributed between the MAR and paper manual groups based on prior knowledge and AR familiarity. The evaluation consisted of three phases: a pre‐test, an immediate post‐test, and a delayed post‐test. Data collection involved knowledge tests, task completion time, error rates, usability and subjective feedback. Results showed significant knowledge gain in both groups, with the MAR group achieving a 21% increase and the paper group 15%. In terms of knowledge retention, both approaches were equally effective in helping users retain knowledge and improve task completion performance by reducing task completion time. Notably, the MAR group (0.5 error rate) made fewer errors than the paper group (1.35 error rate). Additionally, MAR demonstrated higher effectiveness based on Perceived Usefulness, Ease of Use, and the NASA Task Load Index. These findings suggest that while both methods support knowledge retention, MAR with better accuracy and usability, making it a valuable tool for IoT assembly training.