Work

Social and Contextual Effects on Human Interaction with a Digital Conversational Agent

Year: 2020
Type: report
Abstract: Smart home assistants, or digital conversational agents (CA), can improve people's lives by helping with reminders and schedules, as well as serving a social function.Research shows that users prefer some features of CAs, such as their adaptability to individuals, more simple speech, and human-like voice.This study aims to develop a CA for older people to support their cognitive and social needs.Before conducting research on older participants, in order to gain a basic understanding of our study design and effects of the interactions, we first conducted the study with young college students at the University of Minnesota, by asking them to interact with a prototype CA for multiple times.User experience data show that participants were more satisfied after being familiarized with the process and after the CA adapted to them.Sociolinguistic analyses of the transcribed content of participants' interactions revealed that the interactions appeared to be more natural when the participant was alone during their interaction with the CA than when an outside observer (research assistant) was also present during their dialogue with the CA.These findings have implications for understanding and researching human computer interactions.Obtaining accurate profiles of users' patterns of interactions with smart home assistants requires a longitudinal rather than cross-sectional approach, that also considers the broader social or interpersonal context of the assessment environment itself.
Author: Hannah Qu
Language: English
Open Access status: green
OpenAlex ID: W3048577632
Cites: 5
Cited by:
Related to: 9