Intelligent Agents in Everyday Settings: Leveraging a Multi-Methods Approach

Conversational Agents (CAs) or Intelligent Personal Assistants (IPAs) (e.g., Apple’s Siri, Microsoft’s Cortana; Amazon’s Alexa and Google’s Google Assistant) are voice-based interfaces designed for tasks in everyday life including: retrieval of information (e.g., weather, traffic, news), streaming of music, online shopping, controlling of home appliances, and voicecalls within the home and automobiles. Continuous enhancements of their natural language processing abilities, seamless set up of miniaturized hardware, and large-scale cloud-based infrastructures render CAs as unobtrusive, artificially intelligent voice sensors. With CAs rapidly making their way into the home market, the social implications remain unclear. Some product companies have released open-source software platforms that allow third-party developers and the general public to contribute software towards the growth of CAs. However, research around userinteraction with CAs in social settings is still at a nascent stage. In this workshop paper, we unpack the methods used in our ongoing work on people’s social interactions with CAs in order to generate discussion around how the research community can leverage various methodologies using both qualitative and quantitative techniques.