Voice Switching in Voice-Enabled Digital Assistants (VDAs)

Voice-enabled digital assistants (VDAs) provide users with options to switch the "out-of-the-box" or default voice interface. Numerous studies have investigated digital assistants. However, no studies have examined factors influencing user decisions to switch the default voice interface in VDAs. Informed by the similarity-attraction behavioral theory, we investigated whether perceived age, accent, gendered voice, and personality of the voice interface influence user decisions to switch the voice in VDAs. Guided by the status quo behavioral theory, we examined factors influencing user decisions to keep the default voice unchanged (status quo). We recruited thirty-one participants who took an online survey consisting of 42 questions, including 27 closed and 15 open-ended, collecting participants' demographic information, experience in and knowledge of how to switch the voice, voice switching behavior, and preferences, among others. We employed the Big Five Personality Traits to assess the participants' personality traits and the perceived personality of the voice in VDAs. We found that nearly 39% of the participants switched the voice interface in VDAs. Another finding is that the majority of male participants and all female participants (switchers and non-switchers) had a female-gendered voice in the VDAs. We detected a high correlation between the participants' own personality traits and the perceived personality traits of their VDAs. Factors such as perceived age, accent, and gender did not influence the decision of the majority of the participants to switch the voice interface. The findings have implications for designing VDAs with personalities that leverage the user experience (UX).