Bachelorarbeit Lisa Heidenreich: Influence of implicit gender cues on perception of virtual agents and interaction – an experimental study in Germany

Given the growing integration of AI into daily life, understanding how gender perceptions influence interactions with (seemingly) artificial intelligent interaction partners is crucial to prevent stereotypes from being reproduced through this interaction. This paper explores the role of implicit gender cues in user perception and satisfaction when interacting with virtual agents. Varying levels of agency in the agents thereby serve as the implicit gender cue. Data analysis from 93 participants indicate no clear gender-based divergence in preferences over all participants, but a preference for low-level agency agents from older male participants. High communion emerged as a significant positive predictor of satisfaction across all conditions, whereas low communion had a detrimental effect. These results underscore the critical role of gender considerations in AI design, suggesting that AI systems should be developed with an awareness of these preferences to enhance user satisfaction and promote gender equality. The study calls for a thoughtful and ethical approach to AI development, ensuring that technology serves diverse user needs without perpetuating stereotypes.

Betreuerin: Anja-Liisa Gonsior, M.A.

Prüfer: Prof. Dr. Dr. Christian Reuter