Gender, more so than Age, Modulates Positive Perceptions of Language-Based Human-Robot Interactions

Prior work has shown that a robot which uses politeness modifiers in its speech is perceived more favorably by human interactants, as compared to a robot using more direct instructions. However, the findings to-date have been based soley on data aquired from the standard university pool, which may introduce biases into the results. Moreover, the work does not take into account the potential modulatory effects of a person’s age and gender, despite the influence these factors exert on perceptions of both natural language interactions and social robots. Via a set of two experimental studies, the present work thus explores how prior findings translate, given a more diverse subject population recruited via Amazon’s Mechanical Turk. The results indicate that previous implications regarding a robot’s politeness hold even with the broader sampling. Further, they reveal several gender-based effects that warrant further attention.