UX Design Guidelines for Teaching Procedural Knowledge in Educational VR: Medium-Specific Message Design Strategies, Learning Effects, and Consistent Improvement Across Learners

Abstract

Research on educational virtual reality (VR) has focused on learning outcomes, but less on the design strategies that create those outcomes or whether benefits are consistent across different learners. This study created user experience (UX) design guidelines for educational VR by analyzing strategies from print, video, and digital media, then adapting their functions for immersive VR. A two-dimensional framework combined five instructional functions from Gagné’s events of instruction and Merrill’s four content types. Guided by cognitive load theory and multimedia learning theory, five guidelines were developed: Space-Centered Stimulus Presentation, Environment-Intrinsic Context Learning Guidance, Body-Tracking-Based Direct Performance Induction, Integrated Structural Real-Time Feedback, and Multi-Scenario Retention and Transfer Training. A CPR prototype focused on compression-only resuscitation was tested with 40 adults in experimental and comparison groups. The experimental group showed significantly larger gains in knowledge (r = 0.704, large), performance (r = 0.395, medium), and usability (r = 0.385, medium). An exploratory analysis found that all experimental participants improved, compared to 45% and 75% in the comparison group (Fisher’s exact p < 0.001 and p = 0.047). Semi-structured interviews with 22 participants revealed that both groups experienced hand-tracking errors; however, the comparison group also reported issues such as feedback separation, redundant presentation, and unclear flow, causing extraneous cognitive load, whereas the experimental group cited the spatially integrated guidance (heart character, sternum display) as beneficial for learning. These results suggest that VR design guided by these guidelines can lead to greater and more consistent learning improvements among users.