Gender-Emotion Stereotypes in HRI: The Effects of Robot Gender and Speech Act on Evaluations of a Robot

Abstract

Humanlike design of social robots can potentially reproduce societal biases. This paper presents an online video experiment (n = 194) examining the stereotyping effects of speech act (assertive vs. affiliative speech) on people's evaluation of gendered robots (masculine vs. feminine) in terms of warmth, competence, and discomfort. Results show that feminine robots are rated higher in competence than masculine robots, regardless of speech act, and assertive robots are rated higher in competence than affiliative robots, regardless of robot gender. Additionally, women rate robots as more competent than men do. The results for warmth and discomfort are insignificant. This study emphasizes the need for theory-driven experiments addressing robot gendering and highlights the importance of avoiding the reinforcement of gender bias in social robot design.