In conversation with chatbots

On 11 February, Dr. Stefanie Klein successfully defended her dissertation on the interaction between humans and text-based chatbots. Her research provides valuable insights into how users perceive chatbots and how organisations can make the best use of these technologies.

In her thesis, the doctoral student from the Everyday Media lab investigated two key questions: Do people prefer human agents or chatbots? And how should chatbots be designed to improve the user experience? The results show that although people prefer to communicate with human agents for a task that requires social and emotional skills, chatbots are perceived as more entertaining. In addition, human-like cues increase user satisfaction. ‘We found that responsive verbal cues in particular, which convey active listening, and an apology from the chatbot if it has not understood a user request, ensure greater acceptance among users,’ explains Stefanie Klein.

Her results represent an important contribution to a better understanding of human-machine communication. At the same time, they provide valuable impetus for future studies. In follow-up work, Stefanie Klein would like to find out whether and how interactions with AI-based systems influence our behaviour and our relationships with other people.