Embodied Conversational Agents

Abstract

It was analyzed whether an embodied conversational agent (ECA) has specific advantages when employed with privacy invading technologies such as a biometric security system. The study compares the effects of an ECA interface with the effects of conventional text-based and voice-based interfaces on user acceptance and usability. An additional variable was whether the biometric system falsely rejected the user twice or whether it directly accepted him/her. Results of the 2 × 3 between-subjects design indicated that, although overall the text interface is rated most positive, voice and ECA yield distinct social effects: They have more advantageous consequences when problems arise – i.e., when the user is rejected repeatedly. The implications for social psychology in terms of applicability of new research methods as well as insights concerning fundamental research are discussed.

Attachment Dynamics in a Virtual World

“…Most psychological theories lack the computational and conceptual details that would be necessary to implement them in an autonomous agent, and in some areas like nonverbal behavior or communication, a lot of psychological knowledge about specific processes simply is lacking. Many current implementations of autonomous agents therefore rely on ad hoc models that are barely based on empirical findings (Krämer, Bente, Eschenburg, & Troitzsch, ; Schönbrodt & Asendorpf, 2011). Clearly, future research is needed to further clarify how psychological models can be incorporated into agents and how humans react on them.…”

Frontline Service Technology infusion: conceptual archetypes and future research directions

“…In terms of appearance, a first topic of interest relates to whether or not an agent profits from having a physical/virtual embodiment? Initial academic findings demonstrate mixed results (Krämer et al , 2009), and have not yet established the conditions under which agent embodiment (dis)favorably impacts customers and/or FLEs. When agent embodiment is considered, multiple forms may appear.…”

Living up to the chatbot hype: The influence of anthropomorphic design cues and communicative agency framing on conversational agent and company perceptions

“…Embodied conversational agents (ECAs) have a (virtual) body or face, usually human-like. By being embodied, ECAs not only engage in a dialogue via language (text or speech), but are also able to use nonverbal communication cues (e.g., facial expressions, gaze, body movements, distance) in realtime interactions with users, raising questions regarding the effects of embodiment on user perceptions of and reactions with these agents (Kr€ amer, Bente, Eschenburg, & Troitzsch, 2009). The proposal here, however, is that chatbots on social media and messaging apps should be considered disembodied conversational agents (DCAs), as they communicate with users primarily via a messaging-based interface.…”