Implicit learning of gaze-contingent events

In a novel implicit learning task, participants responded to a target stimulus that could appear in one of three locations. Unknown to participants, the location in which the target appeared was probabilistically determined on the basis of the location of eye-gaze immediately prior to the appearance of the target. Participants’ response times to the appearance of the target in a high-probability location were faster than when it appeared in a low-probability location, revealing that participants were able to learn these gaze-contingent events. Furthermore, there was no difference in the cuing score between those participants classified as aware or unaware of the contingencies on a subsequent forced-choice recognition task. These data suggest the task involves implicit learning of instrumental (action-outcome) contingencies, which has potential implications for our understanding of gaze-contingent processes in social interaction.