Does the presence of social agents improve cognitive performance on a vigilance task?

Abstract

Humans are historically bad at performing cognitive tasks that require sustained attention. Social facilitation theory states that the real, imagined, or implied presence of other social agents can improve cognitive performance. Typically these agents are other humans, but the current study explores the possibility that robots can trigger social facilitation effects. We hypothesized that more humanlike social stimuli are (1) more likely to be ascribed internal states (e.g., having a mind, having emotions, having preferences), and (2) more likely to induce social facilitation on a vigilance task. Experiment 1 investigates the relationship between physical humanness and attributions of intentionality by comparing ratings of internal states for three agents (human, robot, and nonsocial). In Experiment 2, we examine whether physical humanness results in improved performance on a vigilance task through social facilitation. While Experiment 1 showed the expected positive relationship between human appearance and mind attribution, the degree to which mental states were attributed to an agent did not influence performance on the vigilance task. The implications for social robotics are discussed.