arXiv subjects

Benjamin A. Clegg

Publications and source records attributed to Benjamin A. Clegg.

4 recordsLinked to original sources

arXiv2026

Leveraging fNIRS to Evaluate Workload for Adaptive Training in Virtual Reality

Advance in technology offer the potential for future adoption of a combination of virtual reality (VR) and real-time adaptivity to enhance training and education. Providing a valid neuro-ergonomic measure of cognitive load can enable an adaptive training regime to continuously adjust tas difficulty to an optimal level as training progresses. The current study validated the functional near-infrared spectroscopy (fNIRS) measure of cognitive load to reflect the demands of two different forms of lad within Cognitive Load Theory: extraneous and intrinsic to he task to be mastered. Thirty-six participants completed a VR shape assembly training task followed by a test of their skill retention They wore near-full head coverage fNIRS and provided subjective ratings of ther workload. The fNIRS findings largely corroborate intrinsic workload literature with significant activation in cortical regions (dorsolateral and rostral prefrontal cortex and left angular gyrus) associated with working memory, short term memory buffers, multisensory integration, and attention. These fNIRS results were tracked closely by NASA TLS measures of mental workload. The results also revealed far less brain activity associated with extraneous load, namely just the right angular gyrus, deemed irrelevant to the mastery of the task.

cs.HC↗

arXiv2025

The Impact of Simple, Brief, and Adaptive Instructions within Virtual Reality Training: Components of Cognitive Load Theory in an Assembly Task

Objective: The study examined the effects of varying all three core elements of cognitive load on learning efficiency during a shape assembly task in virtual reality (VR). Background: Adaptive training systems aim to improve learning efficiency and retention by dynamically adjusting difficulty. However, design choices can impact the cognitive workload imposed on the learner. The present experiments examined how aspects of cognitive load impact training outcomes. Method: Participants learned step-by-step shape assembly in a VR environment. Cognitive load was manipulated across three dimensions: Intrinsic Load (shape complexity), Extraneous Load (instruction verbosity), and Germane Load (adaptive vs. fixed training). In adaptive training (experiment 1), difficulty increased based on individual performance. In fixed training (experiment 2), difficulty followed a preset schedule from a yoked participant. Results: Higher Intrinsic Load significantly increased training times and subjective workload but did not affect retention test accuracy. Extraneous Load modestly impacted training time, with little impact on workload or retention. Adaptive training shortened overall training time without increasing workload or impairing retention. No interactions were observed between the three types of load. Conclusion: Both Intrinsic and Extraneous Load increased training time, but adaptive training improved efficiency without harming retention. The lack of interaction between the elements suggests training benefits can be worth seeking within any of the components of cognitive load. Application: These findings support the use of VR adaptive systems in domains such as manufacturing and military service, where efficient assembly skill acquisition is critical. Tailoring difficulty in real-time can optimize efficiency without compromising learning.

cs.HC↗

arXiv2024

Multimedia and Immersive Training Materials Influence Impressions of Learning But Not Learning Outcomes

Although the use of technologies like multimedia and virtual reality (VR) in training offer the promise of improved learning, these richer and potentially more engaging materials do not consistently produce superior learning outcomes. Default approaches to such training may inadvertently mimic concepts like naive realism in display design, and desirable difficulties in the science of learning - fostering an impression of greater learning dissociated from actual gains in memory. This research examined the influence of format of instructions in learning to assemble items from components. Participants in two experiments were trained on the steps to assemble a series of bars, that resembled Meccano pieces, into eight different shapes. After training on pairs of shapes, participants rated the likelihood they would remember the shapes and then were administered a recognition test. Relative to viewing a static diagram, viewing videos of shapes being constructed in a VR environment (Experiment 1) or viewing within an immersive VR system (Experiment 2) elevated participants' assessments of their learning but without enhancing learning outcomes. Overall, these findings illustrate how future workers might mistakenly come to believe that technologically advanced support improves learning and prefer instructional designs that integrate similarly complex cues into training.

cs.HC↗

arXiv2023

Fast, Accurate, but Sometimes Too-Compelling Support: The Impact of Imperfectly Automated Cues in an Augmented Reality Head-Mounted Display on Visual Search Performance

While visual search for targets within a complex scene might benefit from using augmented-reality (AR) head-mounted display (HMD) technologies helping to efficiently direct human attention, imperfectly reliable automation support could manifest in occasional errors. The current study examined the effectiveness of different HMD cues that might support visual search performance and their respective consequences following automation errors. Fifty-six participants searched a 3D environment containing 48 objects in a room, in order to locate a target object that was viewed prior to each trial. They searched either unaided or assisted by one of three HMD types of cues: an arrow pointing to the target, a plan-view minimap highlighting the target, and a constantly visible icon depicting the appearance of the target object. The cue was incorrect on 17% of the trials for one group of participants and 100% correct for the second group. Through both analysis and modeling of both search speed and accuracy, the results indicated that the arrow and minimap cues depicting location information were more effective than the icon cue depicting visual appearance, both overall, and when the cue was correct. However, there was a tradeoff on the infrequent occasions when the cue erred. The most effective AR-based cue led to a greater automation bias, in which the cue was more often blindly followed without careful examination of the raw images. The results speak to the benefits of augmented reality and the need to examine potential costs when AR-conveyed information may be incorrect because of imperfectly reliable systems.

cs.HC↗