What Triggers Social Responses to Flattering Computers? Experimental Tests of Anthropomorphism and Mindlessness Explanations

The present research evaluated two explanations for the computers are social actors (CASA) paradigm: anthropomorphism and mindlessness. Using flattery effects as an example of social responses, two experiments examined how humanlikeness of the interface, individuals' rationality, and cognitive busyness moderate the extent to which people apply social attributes to computers. In Experiment 1, anthropomorphic cartoon characters elicited more positive overall evaluations of the computer, but they significantly reduced low rationals' self-confidence, suggesting social facilitation effects. Moreover, low rationals were less likely to accept the computer's suggestions when flattered, whereas high rationals showed no corresponding tendency. In Experiment 2, although participants attributed greater social attractiveness to the flattering than generic-comment computer, they became more suspicious about the validity of its claims and more likely to dismiss its answer. Such negative effects, however, disappeared when they simultaneously engaged in a secondary task. Theoretical implications for CASA are discussed.