Back to search

Michigan State University

Mental models and home virtual assistants (HVAs : a single turn-taking is not enough to have a human-like conversation

Abstract

dc:description

As virtual assistants become more intelligent, these applications are dispersed in our everyday life. However, virtual assistants are still used in very routine tasks (Wingfield, 2018). People are expected to speak to virtual assistants in a conversational manner; nevertheless, there has been little research looking at people's mental models for what kinds of interactions they think the devices are capable of. I conducted a qualitative study to describe people’s strategies to seek a wide range of information when interacting with Google Home, which is a type of Home Virtual Assistant (HVA), and how their understandings of HVAs may change their interaction. I found that people believed applying human-to-human communication skills may facilitate a conversation with Google Home (i.e., common-sense models). Also, people expected Google Home performs like Google search or an artificial intelligence system that is able to become more intelligent if people provide more information (i.e., machine-oriented models). I present implications for the design of HVAs to encourage inexperienced people to be involved in the interaction with HVAs based on these results.

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Cho, Janghee
Contributors dc:contributor
  • Rader, Emilee
  • Wyche, Susan
  • Park, Taiwoo

Subjects

dc:subject × 4

Rights

dc:rights
Statement dc:rights
  • Attribution-NonCommercial-ShareAlike 4.0 International
Language dc:language
English

Identifiers

dc:identifier.*
Identifier
etd:16412
isbn:9780355939415
isbn:035593941X
oclc:1047957225
umi:10812507
local:Cho_grad.msu_0128N_16019

Chain of custody

source
Harvested from
Michigan State University
Base URL
d.lib.msu.edu/oai
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Cho, Janghee. Mental models and home virtual assistants (HVAs : a single turn-taking is not enough to have a human-like conversation. 2018. https://doi.org/doi:10.25335/9qwy-8p18