טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
Ph.D Thesis
Ph.D StudentShorr Yifat
SubjectThe Effect of Multimodal Feedback on the Acquistion
of Rowing Skills
DepartmentDepartment of Industrial Engineering and Management
Supervisors Professor Emeritus Daniel Gopher
Ms. Maria Korman
Full Thesis textFull thesis text - English Version


Abstract


This study focused on the selection and evaluation of different modes of information feedback in a virtual environment to facilitate acquisition and transfer of the  complex motor-cognitive skill of rowing. Rowing skill is based on a periodic movement, where stroke is a unit cycle, and requires coordination of several limbs to make the oars cover the right cyclic trajectory in the right direction as smooth as possible. Skill acquisition necessitates multi-session practice and thus can benefit from simulator environment empowered by virtual reality technologies that can enhance indoor training. Expensive efforts have been made to develop virtual reality simulators to mimic ecological tasks with high fidelity. However, it remains unclear whether the information available in the real world is integrated in the same way in a virtual environment. Thus there is no straightforward answer to whether the high fidelity between an ecological and a virtual task should be preserved. To probe this crucial issue we designed a simplified indoor rowing platform with similar visual and haptic feedback modalities as the ecological task. These modalities  were designed to provide feedback on the accuracy of the arm movement per stroke: a. the haptic sense was engaged by creating an air flow at three critical points along the trajectory of arm movement; b. a visual representation that indicated whether the movement went through or missed the three critical points along the hand path. Changes in hand trajectory during stroke were analyzed in reference to the”ideal” performance signatures extracted from expert rowers' performance data.  The training performance of 4 experimental groups: Visual, Haptic, Visual Haptic augmentation, and a Basic feedback condition that only included a two- dimensional display of the boat's progress on the screen, and a metronome that provided auditory cues.     


The results showed that the performance of all 4 experimental groups improved with training and converged to similar levels of error rates and variance of performance in the retention test at the end of training. Surprisingly, the Basic feedback group, with no augmented feedback, showed better learning results, as reflected in high rates of Drop proportions on the retention test and a high level of transferability to new tasks. Thus overall, enriched augmented feedback in similar modalities as the ecological task environment may not lead to better results. The implications of these findings are discussed.