Figure 3. Performance accuracy ratios for each experiment (1 and 2), by target type (self vs. non-self) and set size (4 vs. 6).

Experiment 1 (top): Participants were more accurate when the target was the item they controlled than when it was not. Adding distractors impaired accuracy only when the target was not controlled by the participant. Experiment 2 (bottom), in which participants viewed the same displays as in Experiment 1 but did not control the motion of any of the stimuli. There was no advantage for targets that had been controlled by the participant in Experiment 1, thus removing any concern that physical differences between self and non-self motion might underlie the findings from Experiment 1. Error bars represent SEM.