Dual humanness and trust in conversational AI: A person-centered approach

Hu, Peng; Lu, Yaobin; Gong, Yeming

doi:10.1016/j.chb.2021.106727

Cited by 50 publications

(14 citation statements)

References 75 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…D2 suggests that for consumers to perceive an opportunity to interact, robots should have proximity to humans as well as a clear ability to interact. The findings of this study provide insights into the growing body of research on social interactions and voice‐based technologies (Hu et al, 2021; Pitardi et al, 2021; Whang & Im, 2020), revealing a preference for medium or high levels of opportunity for social contact as represented by robots who were nearby humans and potentially able to hold a conversation (D2), but no significant preference for active social interaction with robots in the form of conversation. Placing this finding within previous literature, it is proposed that within this study, some participants saw the robots as human‐like enough to merit ingroup status and the potential for interactions this can bring (Vaes et al, 2003, 2012), while some did not.…”

Section: Discussionmentioning

confidence: 71%

Robots should be seen and not heard…sometimes: Anthropomorphism and AI service robot interactions

Letheren

Jetten

Roberts

et al. 2021

Psychology and Marketing

View full text Add to dashboard Cite

There is a growing need to understand how consumers will interact with artificially intelligent (AI) domestic service robots, which are currently entering consumer homes at increasing rates, yet without a theoretical understanding of the consumer preferences influencing interaction roles such robots may play within the home. Guided by anthropomorphism theory, this study explores how different levels of robot humanness and social interaction opportunities affect consumers' liking for service robots. A review of the extant literature is conducted, yielding three hypotheses that are tested via 953 responses to an online scenario‐based experiment. Findings indicate that while consumers prefer higher levels of humanness and moderate‐to‐high levels of social interaction opportunity, only some participants liked robots more when dialogue (high‐interaction opportunity) was offered. Resulting from this study is the proposed Humanized‐AI Social Interactivity Framework. The framework extends previous studies in marketing and consumer behavior literature by offering an increased understanding of how households will choose to interact with service robots in domestic environments based on humanness and social interaction. Guidelines for practitioners and two overarching themes for future research emerge from this study. This paper contributes to an increased understanding of potential interactions with service robots in domestic environments.

show abstract

Section: Discussionmentioning

confidence: 71%

Robots should be seen and not heard…sometimes: Anthropomorphism and AI service robot interactions

Letheren

Jetten

Roberts

et al. 2021

Psychology and Marketing

View full text Add to dashboard Cite

show abstract

“…One of the most promising AI applications, namely, conversational agents ( Carter, 2018 ), are disruptive innovations attracting extensive attention from practitioners and researchers. Considerable research focuses on the representative form of conversational agents known as chatbots and examines user responses to their human-like attributes ( Hu et al, 2021 , Luo et al, 2019 , Schanke et al, 2021 , Sheehan et al, 2020 ). For instance, Schanke et al (2021) investigated how consumers who want to sell used clothes respond to customer service chatbots.…”

Section: Literature Reviewmentioning

confidence: 99%

“…Trust is frequently mentioned in mutual relationships ( Cheng et al, 2017 , Jiang and Lau, 2021 , Zhao et al, 2020 ), such as the relationship between buyers and sellers in e-commerce ( Lu et al, 2016 ), between community members ( Cheng et al, 2019b ), between demanders and suppliers in ridesharing ( Cheng et al, 2019a , Cheng et al, 2020 ), and between humans and AI agents in human-AI interaction ( Hu et al, 2021 ). It refers to a positive and confident expectation of others’ behaviors ( Mcknight et al, 2002 ).…”

Section: Literature Reviewmentioning

confidence: 99%

“…Trusted team collaboration is necessary for team members to feel confident and satisfied with the deliverables they have collectively agreed upon, dependent on trust in both the system and the collaboration process recommended by it. Trust has been frequently discussed in both human-AI interaction and CE contexts ( Cheng et al, 2022a , Cheng et al, 2016 , Hu et al, 2021 , Shin, 2020 ). It has an influence on user experience while interacting with AI applications and also affects the collaboration performance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

An investigation on trust in AI-enabled collaboration: Application of AI-Driven chatbot in accommodation-based sharing economy

Cheng

Zhang

Yang

et al. 2022

Electronic Commerce Research and Applications

View full text Add to dashboard Cite

“…

…”

mentioning

confidence: 99%

What Drives Competency-based and Integrity-based Trust in Voice Assistants?

Huang

Kim

Lennon³

2022

Breaking Boundaries

View full text Add to dashboard Cite

Voice commerce in which consumers make purchases using a voice assistant (VA) is on the rise. A VA is a digital agent that uses natural language to communicate with consumers via a human-like voice (Hu et al., 2021). Amazon's Alexa, Google Assistant, and/or Apple's Siri are popular VAs. According to eMarketer (2020), the number of U.S. VA users is estimated to top 135.6 million by 2022, which is 40.2% of the population. COVID-19 has accelerated the growth of voice commerce, as in-person service interactions have become more difficult and uncertain (Harvard Business Research, 2020).An industry report revealed lack of trust was a major barrier for shoppers who used VAs to continue that use in the future (PwC, 2019); however, the role of trust on interactions between consumers and VAs remains under-researched (Foehr & Germelmann, 2020). Therefore, an important empirical question is what drives the development of trust in consumers interactions with VAs. The objectives of this research are twofold: (1) to unpack the dimensionality of trust in VAs; and (2) to explore the antecedents and outcomes of trust in interactions between consumers and VAs. Empirical findings of the study will help fill a void in the literature by developing a model of trust in the context of VAs.Trust refers to "the willingness of a trustor (e.g., consumer) to be vulnerable to the actions of trustee (e.g., VAs) based on the expectation that the trustee will perform a particular action important to the trustor, irrespective of the ability to monitor or control the trusted agent" (Mayer et al., 1995, p. 712). Applied to VAs, trust can be conceptualized: (1) competence-based trust that refers to a VA's ability and skills to meet consumers' expectations; and (2) integritybased trust that refers to the degree to which a VA adheres to a set of sound principles and performs actions in a reliable way (Chattaraman et al., 2019).Role Theory (Solomon et al., 1985) provides a theoretical framework to conceptualize the relationship between consumers and VAs. According to the theory a variety of norms (e.g., functional, social, and/or cultural norms) are involved in dyadic interactions which structure how individuals are expected to behave in certain contexts. Such expectations include expectations for VAs' behavior. Fernandes and Oliveira (2021) identify two dimensions of role expectations for VAs as social-emotional (i.e., humanness and enjoyment) and functional (i.e., innovativeness, autonomy, ease of use, and usefulness). Humanness refers to the extent to which a VA has anthropomorphic qualities in form and behavior (Wirtz et al., 2018). Enjoyment refers to the degree to which a VA is perceived to be fun (Venkatesh, 2000). Innovativeness is defined as a VA's ability to adopt new ideas or skills (McLeay et al., 2021). Autonomy is defined as the ability to perform the tasks self-controllably (Hong & Williams, 2019). Ease of use refers to the way a VA is perceived to be used without any effort and usefulness refers to the degree of which a VA can bring bene...

show abstract

Dual humanness and trust in conversational AI: A person-centered approach

Cited by 50 publications

References 75 publications

Robots should be seen and not heard…sometimes: Anthropomorphism and AI service robot interactions

Robots should be seen and not heard…sometimes: Anthropomorphism and AI service robot interactions

An investigation on trust in AI-enabled collaboration: Application of AI-Driven chatbot in accommodation-based sharing economy

What Drives Competency-based and Integrity-based Trust in Voice Assistants?

Contact Info

Product

Resources

About