Follow the Leader The Role of Local and Global Visual Information for Keeping Distance in Interpersonal Coordination

Abstract

Despite many years of research into human movement, how humans deal with information in dynamical situations is still subject to debate. The current research programme examined an individual’s ability to coordinate their actions with others in invasion sports (‘interact-ability’) using the ecological dynamics theoretical framework to address this general aim. In its essence interact-ability describes how humans make sense of their sensory world in order to tie in - some form of - action output to serve goal-directed behaviour. A series of experimental studies examined the law-like relationship between agent and environment (cf., perception-action) when one has to coordinate with others. A virtual reality task was chosen that involved maintaining distance in locomotion: follow the leader. The first study examined whether optical expansion by itself would enable distance keeping in a follow-the-leader locomotion task in forward-backward direction. In one condition, participants coordinated their locomotion with an expanding and compressing sphere, whilst in another with a fully animated avatar (i.e., with moving limbs). The coordination of the follower with the leader was analysed using response times (RT) and the point-estimate relative phase (φ) to quantify the temporal synchrony. Additionally, the spatial accuracy was estimated by testing to what extent the rate of change in visual angle was nulled over time. Findings showed a decreased temporal synchrony, but no significant decrease in spatial accuracy, when no limb movement was present in the leader stimulus (i.e., sphere compared to avatar). Additionally, it appeared that regulating distance based on global motion information was affected by a direction-based visual angle bias. The second study then investigated if information for action could be situated along a spectrum from local (i.e., segmental) to global (i.e., expansion-compression) visual information sources. It also was examined how the perception-action coupling was mediated by key task constraints (i.e., regularity and viewpoint). Extending the analysis procedures of the first study, the virtual distance between follower and leader was estimated to further quantify the spatial accuracy. It was put forward that followers may benefit most from flexibly switching between information sources governed by task constraints. Participants followed more irregular leaders better when local motion information was available. Although information for action may not be easily situated along a linear spectrum, various relative benefits were put forward for each component of the proposed spectrum. In the third and final study, these follower-leader dynamics were examined in a lateral follow-the-leader task. It was shown that a point-light display provided information to tighten the temporal synchrony. However, as the spatial accuracy was not significantly affect by the information presented, it may be that early responses were as often facilitating as detrimental for keeping distance. Overall, this body of work may contribute towards understanding how action and perception are linked in dynamic interpersonal situations. Local sources of information were shown to contribute to a tightened temporal synchronization and global sources of information were consistently shown to provide pertinent distance-related information. The main contribution is substantiated by showing how agents in a whole-body interaction task can flexibly use different sources of information and do so as a function of task constraints.