Human Performance And Motor Behaviour Models For Distal Positioning Tasks In Virtual Reality

It is essential to understand how humans interact with a virtual environment and build motor behavior models to guide user interface design.One of the basic and the most common task in a virtual environment is the distal positioning task.The task requires a user to place an object that cannot be reached directly by his or her arm.It remains unclear what factors determine the performance of the distal positioning task.In this dissertation,we conducted eight studies to explore the effect of factors in three aspects: tasks types,interaction devices,and environments on object positioning performance in a virtual environment.In the first study,we asked participants to perform a typical distal positioning task by a hand-held controller in the ray-casting technique.We measured the movement time(MT)in three phases: acceleration,deceleration,and correction.The results demonstrated different determining factors in three phases.In the acceleration phase,MT was inversely related to object size and positively proportional to movement amplitude.In the deceleration phase,MT was primarily determined by movement amplitude.In the correction phase,MT was affected by all three factors.We observed similar data patterns at different backgrounds and using different interaction devices(head tracking control).We thus proposed a three-phase model which fits the participants' performance in distal object positioning very well.In the second study,we conducted two experiments to explore the effect of depth on MT.Our results demonstrated a significant effect of depth on MT in the deceleration and correction phases when using physical length.Interestingly,the effect of depth was normalized if all distances were represented in an angle unit instead of a physical length.The results indicate that the angle unit was more appropriate than the physical length in our three-phase model.Study 3 investigated the effect of movement directions on the distal positioning task.We observed a fluctuation of MT in a sine function pattern across the movementdirections.We thus revised the three-phase model by adding a parameter of the movement direction.Study 4 compared the efficiency of two interactive devices: head tracking control and eye tracking control.Each of them has its own advantages.Head tracking control demonstrated good stability in the correction phase while eye tracking control showed high efficiency in the acceleration and deceleration phases.In particular,eye tracking control outperformed the head tracking control when the movement distance was between 30° and 50° and the target tolerance exceeded 9°.Study 5 proposed an approach to combine the two interactive approaches by using eye tracking control in the acceleration and deceleration phases,and head tracking control in the correction phase.The combination approach demonstrated a better performance than either eye tracking control or head tracking control only.In the 6th and 7th studies,we explored whether altering the display-control gain could improve the operational efficiency of head-tracking control and eye-tracking control.The results showed that setting gain values in a range of 1.3 and 1.5 in the acceleration and deceleration phases could significantly improve the operating efficiency of the head tracking control while altering gain values in the correction phase could not significantly improve the performance of eye tracking control.Study 8 explored the effect of body position on MT in head tracking control.The results showed MT was shortest in leaning forward and the longest in leaning backward.In addition,when the human body was titled more than 45°,the performance of the head tracking control was significantly degraded.In summary,this study provides a quantitative framework for researchers to evaluate 3D interfaces for the positioning task in a virtual environment.Moreover,our model can provide valuable guides to designers who would improve efficiency in designing virtual reality applications.