Abstract
Introduction: Reflexive orienting is the automatic direction of spatial attention to a location indicated by a cue. Previous research has shown that these effects can be elicited by the eye or body orientation of a social agent. Here, we tested whether the biological-ness of an agent influenced social attentional orienting. Methods: We used a standard attentional orienting paradigm with humanoid agents as cue stimuli. Subjects were asked to perform a speeded target detection task. Cue direction was not predictive of the target's location. Three types of agents were used: a human, and two non-human agents: one with highly humanlike appearance (android), and one with a clearly non-biological appearance (robot). Each trial started with an image of an agent with face and upper body in clear view, followed by another image of the same agent with the head and upper body turned to the right or left (directional cue), or continuing to face forward (neutral cue). A target letter appeared to the right or to the left of the agent after a variable SOA (200, 400, or 60ms). Results: There was a main effect for cue validity; subjects responded faster to valid cues for all agents. There was also a main effect of agent; responses in trials with the robot cue were slower relative to the other agents, whereas the android and human conditions did not differ. There was also a main effect of SOA, but no interaction with agent. Conclusions: Humanoid agents can automatically direct spatial attention regardless of whether they have biological or non-biological status. However, there is a small but consistent RT cost imposed by an agent with non-biological appearance that occurs independently of cue validity. Subsequent studies will use videos to examine potential effects of both motion and appearance on reflexive orienting.
Meeting abstract presented at VSS 2014