Paper ID

1567

Paper Type

full

Description

Current technological advancements of conversational agents (CAs) promise new potentials for human-computer collaborations. Yet, both practitioners and researchers face challenges in designing these information systems, such that CAs not only increase in intelligence but also in effectiveness. Drawing on social response theory as well as literature on trust and judge-advisor systems, we examine the roles of gender stereotyping and egocentric bias in cooperative CAs. Specifically, by conducting an online experiment with 87 participants, we investigate the effects of a CA’s gender and a user’s subjective knowledge in two stereotypical male knowledge fields. The results indicate (1) that female (vs. male) CAs and stereotypical female (vs. male) traits increase a user’s perceived competence of CAs and (2) that an increase in a user’s subjective knowledge decreases trusting intentions in CAs. Thus, our contributions provide new and counterintuitive insights that are crucial for the effectiveness of cooperative CAs.

Share

COinS
 

Mr. and Mrs. Conversational Agent - Gender Stereotyping in Judge-Advisor Systems and the Role of Egocentric Bias

Current technological advancements of conversational agents (CAs) promise new potentials for human-computer collaborations. Yet, both practitioners and researchers face challenges in designing these information systems, such that CAs not only increase in intelligence but also in effectiveness. Drawing on social response theory as well as literature on trust and judge-advisor systems, we examine the roles of gender stereotyping and egocentric bias in cooperative CAs. Specifically, by conducting an online experiment with 87 participants, we investigate the effects of a CA’s gender and a user’s subjective knowledge in two stereotypical male knowledge fields. The results indicate (1) that female (vs. male) CAs and stereotypical female (vs. male) traits increase a user’s perceived competence of CAs and (2) that an increase in a user’s subjective knowledge decreases trusting intentions in CAs. Thus, our contributions provide new and counterintuitive insights that are crucial for the effectiveness of cooperative CAs.