Abstract
Previous studies have shown that advance information (more certainty) about target static location increases movement time in reaching tasks. On the other hand, fast movement times are observed when we know when to intercept a moving object and have little time. It is unclear, however, how uncertainty will affect movement time when we have to rely on uncertain predictions based on early parts of a trajectory. I address this question in one Experiment in which subjects wearing a data glove had to catch balls in a VR setting. Four horizontal velocities (HV) and four initial vertical velocities (VV) resulted in 16 different trajectories with four flying times (0.61, 0.78, 0.94 and 1.11 seconds). The same trajectories were shown in 3 viewing conditions (early, late and full path). In half of the sessions feedback on the action success was given. Results show that accuracy was similar across viewing conditions but the contribution of movement initiation and movement time to reducing temporal error was different. Subjects started to close the hand earlier and performed slower closing movements when they saw the initial part of the trajectory only. This was especially so for longer flying times which implied larger VVs and, therefore, prediction relied more on elevation angle than image expansion. On the contrary, action was initiated later and hand closing was faster in the late and full vision conditions. Furthermore, action started according to a rate of expansion threshold only when the last part of the path was available and VVs were small (smaller flying times). Absence of feedback led to longer movement times especially in conditions that subjects had to rely on predictive information. I suggest that longer movement times increase temporal accuracy under the uncertainty of predictive information and this strategy is further promoted when feedback is missing.
Meeting abstract presented at VSS 2012