Is My Alexa Happy with Me? Attributions of Emotional Displays in Human-Computer Interactions
Abstract
Virtual assistants such as Amazon's Alexa can emulate a variety of emotions in their spoken feedback to user requests. The persuasive impact of these emotional displays depends on the inferences individuals make about these emotional displays. This dissertation investigates a class of inferences – attributions of cause to a virtual assistant's emotional display – which individuals use as social information about computer-interactants. The project hypothesized that three key factors influence individuals' attributions of cause to virtual assistants' emotional displays. The first is general tendencies in peoples' attributions of virtual assistants' emotional displays. Second is the target of a virtual assistant's emotional display – the communicated object of their displayed emotional state. The final component is borrowed from Kelley's covariance theory of attribution – the distinctiveness with which a specific emotional display repeatedly co-occurs with plausible explanations for the emotional display. Four attributional outcomes were predicted based on combinations of emotional display targets and the distinctiveness of the emotional display's covariance with potential causes for the emotional display. Findings suggest that individuals first and foremost attribute virtual assistant's emotional displays to the virtual assistant rather than situational causes, that emotional display targets can influence attributions, and that further work is needed to assess the role of distinctiveness in attributions of virtual assistant's emotional displays.
Degree
Ph.D.
Advisors
Reimer, Purdue University.
Subject Area
Computer science|Information Technology|Web Studies
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.