Investigating visuomotor control and adaptation in response to constant and varying visual feedback delays

Abstract

When interacting with network-supported technologies, we experience delays between our actions and their corresponding outcomes within the virtual space. Previous research shows that performance immediately declines at first exposure to the delay, but this is followed by adaptation to recover performance. Much of this research has used constant delays, but in practice, these delays continuously fluctuate in response to geographical, mechanical, and infrastructural factors. Performance and adaptation are believed to be poorer under varying delay conditions, but there is an absence of research comparing visuomotor control and adaptation between constant and varying delay conditions.
Experiments 1 and 2 explore how changes in the mean delay, jitter amplitude, and jitter frequency affect performance in a mouse-based acquisition task where the participants had to click on a series of static targets. The mean delay was the largest determinant of performance, with delay variability only having a small impact.

After observing a minimal effect of delay variability, Experiments 3 and 4 investigate adaptation within the same acquisition task. The participants adapted to the constant and varying delays at a similar rate and showed similar after-effects.

Finally, Experiments 5 and 6 investigate adaptation and perceptual recalibration to constant and varying delays in a driving simulator task across multiple sessions. With each session, the participants in both delay conditions demonstrated faster adaptation and greater perceptual recalibration.

This thesis shows that delay variability is not as universally disruptive as previously considered. The participants showed similar performance and adaptation to the constant and varying delays in both tasks. Moreover, the two delay conditions showed a similar savings effect. With each session, the average performance level improved, the learning slopes flatted, and the participants reported that they eventually stopped perceiving the delays. The general discussion explores the practical implications of these findings for human-centred network optimization.

Date of Award	10 Dec 2024
Original language	English
Awarding Institution	University of Bristol
Supervisor	Iain D Gilchrist (Supervisor) & Danae E B Stanton Fraser (Supervisor)