Talking with an IoT-CA: Effects of the Use of Internet of Things Conversational Agents on Face-to-Face Conversations

Abstract Internet of things conversational agents (IoT-CAs) are making human–computer interactions ubiquitous. In this study, we experimentally examined the effects of IoT-CA use on face-to-face conversations between close partners. A total of 136 participants (68 close relationship dyads) participated in the experiment. We prepared an IoT chat environment and provided chat topics for each dyad. The dyads were randomly assigned into one of two IoT-CA use pattern groups (joint use: two persons using an IoT-CA together; individual use: one person using an IoT-CA alone) and three interaction conditions (no IoT-CA use, conversation content-relevant IoT-CA use, conversation content-irrelevant IoT-CA use). The results showed that compared with no IoT-CA use, IoT-CA use did not have negative effects on conversation experiences but produced feelings of greater closeness to the IoT-CA in the partners. Furthermore, joint IoT-CA use in the content-relevant condition (IoT-CA made comments relevant to interpersonal interactions) helped increase interpersonal self-disclosure.