Conversational Agents and AI

Emma, one of our chatbots

Artificial Intelligence (AI) and related technologies are reshaping our communication environment in an extremely fast pace. These changes encompass not only drastic changes to the media landscape, but especially to the partners with whom we communicate. More than a medium for human-to-human communication, computers and online environments now allow humans to communicate with machines, i.e., conversational agents in the shape of chatbots, socialbots, and virtual assistants.

Given the growing levels of sophistication and adoption of these agents, the Digital Communication Lab focuses on innovative research with (disembodied) conversational agents, including the content, causes and consequences of interactions with AI in the shape of conversational agents across different areas of communication. Methodologically, the lab also explores the potential of actual conversational agents as data collection tools.

Current research encompasses a set of experimental studies with actual conversational agents (chatbots) developed specifically for each study. These studies cover, among other topics, the antecedents and consequences of anthropomorphism and social presence perceptions about these agents in perceptions about the agent, the content of the conversation, and the organisation responsible for the agent.

Projects – 2019

Persuasive consequences of conversational agents

Digicomlab Researcher: Carolin Ischen

Conversational agents (e.g. in the form of text-based chatbots) are reshaping our communication environment and have the potential to inform and persuade in new and effective ways. In this light, this project investigates how chatbots’ communicative characteristics influence, over time, the human-chatbot interaction and persuasion outcomes. The theoretical focus of the project are intended as well as unintended persuasive consequences associated with chatbots, communicating in a conversational manner, using social and/or persuasive cues and displaying human-like behaviour. The project has an explicit emphasis on method development on how to integrate conversational agents in communication research.

Conversational Agents in Communication: Methodological Approaches

Digicomlab Researcher: Theo Araujo

This research program aims to extend and develop methodological approaches to research conversational agents such as chatbots, virtual assistants and other digital entities. This includes the development of frameworks to run experiments with chatbots investigating the content, causes and consequences of communication with these agents, and also studies to explore how these agents can be used as data collection modes – complementing surveys. For 2019, two main projects are ongoing: (1) exploring the affordances and limitations of chatbots as data collection modes, including ethical and privacy considerations, and (2) investigating how to create and integrate conversational agents into voice platforms (e.g., virtual assistants).

Presentations and Publications:

  • Ischen, C., Araujo, T., van Noort, G., Voorveld, H., and Smit, E. (2019). How important is agency? The persuasive consequences of interacting with a chatbot as a new entity. Paper presented at the Human-Machine Communication Preconference, ICA Conference, Washington, USA
  • Araujo, T., and Bol, N. (2019). From speaking like a person to being personal: The effects of personalised, regular interactions with conversational agents. Paper presented at the Human-Machine Communication Preconference, ICA Conference, Washington, USA
  • Araujo, T. (2018). Living up to the chatbot hype: The influence of anthropomorphic design cues and communicative agency framing on conversational agent and company perceptions. Computers in Human Behavior, 85, 183-189