Why Female Virtual Conversational Agents Exhibit Greater Social Presence: Insights From the Stereotype Content Model

In recent years, companies have increasingly adopted virtual conversational agents (VCAs) to interact with their consumers. A significant number of these agents are assigned to a specific gender, typically female, thereby reinforcing gender-related stereotypes. This work employs the stereotype content model (SCM), especially focusing on the two subdimensions sociability and morality, to investigate how the VCA's gender impacts social presence in two studies (Study 1: avatars; Study 2: chatbots). Both studies show an indirect effect of the VCA's gender (female) on social presence through sociability, but not through morality. Contrary to expectations, the indirect effect through sociability is significant for both male and female consumers. The findings indicate that the indirect effect is robust across different product categories (i.e., finance, fashion, holiday trip, and toothbrush) and types of VCAs (i.e., visual and verbal). This research demonstrates how established stereotype theories (i.e., SCM) can serve as a valuable framework for exploring human-machine communication. Moreover, the results provide a fundamental basis for addressing and reducing the persistence of gender stereotypes, initiating a more extensive discussion about the ethical considerations related to interactions between humans and technology.