The Impact of Multimodal Communication on a Shared Mental Model, Trust, and Commitment in Human–Intelligent Virtual Agent Teams

Abstract

There is an increasing interest in the use of intelligent virtual agents (IVAs) to work in teams with humans. To achieve successful outcomes for these heterogeneous teams, many of the aspects found in successful human teams will need to be supported. These aspects include behavioural (i.e., multimodal communication), cognitive (i.e., a shared mental model (SMM)), and social (trust and commitment). Novelly, this paper aims to investigate the impact of IVA’s multimodal communication on the development of a SMM between humans and IVAs. Moreover, this paper aims to explore the impact of the developed SMM on a human’s trust in an IVA’s decisions and a human’s commitment to honour his/her promises to an IVA. The results from two studies involving a collaborative activity showed a significant positive correlation between team multimodal communication (i.e., behavioural aspect) and a SMM between teammates (i.e., cognitive aspect). Moreover, the result showed that there is a significant positive correlation between the developed SMM and a human’s trust in the IVA’s decision and the human’s commitment to honour his/her promises (the establishment of the social aspect of teamwork). Additionally, the results showed a cumulative effect of all of these aspects on human–agent team performance. These results can guide the design of human–agent teamwork multimodal communication models.

Coordinating Mechanisms Are More Important Than Team Processes for Geographically Dispersed Emergency Dispatch and Paramedic Teams

“…The present study showed a high correlation between trust and SMM, which is in line with previous findings ( Hanna and Richards, 2018 ). While both SMM and CLC were related to team processes, trust was not.…”

Conceptual Designing of a Virtual Receptionist: Remote Desktop Walkthrough and Bodystorming in VR

“…The design of IVAs focuses on different user-centred aspects such as dialogue, emotion, appearance, presence and behaviour, but design decisions do not always reflect the design goals or the qualities valued by the users [16]. Research on cross-platform (multi-modal) IVA design or development tends to refer to systems/platforms that combine aspects of an IVA (e.g, [10,35]) or the integration of an IVA platform with another platform such as VR or augmented reality (e.g., [2,13]). We have in this design study presented a somewhat novel and interesting concept of a cross-platform IVA where the agent is able to adapt to different embodiments, be it physical (robot) or virtual (on screen), and provide seamless interaction.…”

Low-rank human-like agents are trusted more and blamed less in human-autonomy teaming

“…In a review and analysis of 24 studies that measured trust in HAT ( O’Neill et al, 2020 ), higher levels of agent autonomy ( Azhar and Sklar, 2017 ), and higher levels of transparency ( Boyce et al, 2015 ; Sharifheravi et al, 2020 ), had a positive effect on trust. When an autonomous agent’s personality ( Hanna and Richards, 2018 ), values ( Mehrotra et al, 2021 ), and work style are similar to participants, they trust agents more, even preferring to work with an agent over a human in low-risk situations ( You and Robert, 2018 ). Other factors influencing higher levels of trust in autonomous agents include experience with similar computers or computers that employ independent agents ( Chen et al, 2011 ), team training ( Walliser et al, 2019 ), and agent reputation ( Hafizoglu and Sen, 2018 ).…”