Mind the Style: Impact of Communication Style on Human-Chatbot Interaction

A new research paper titled 'Mind the Style: Impact of Communication Style on Human-Chatbot Interaction' reveals that how chatbots communicate significantly affects both user experience and task performance. Published on arXiv by researchers including Erik Derner, Dalibor Kučera, and Nuria Oliver, the study conducted a between-subject experiment with 120 participants interacting with a chatbot called NAVI during a map-based navigation task.

The researchers created two identical versions of NAVI that differed only in communication style: one used friendly, supportive language while the other employed direct, task-focused communication. Results showed the friendly style increased subjective satisfaction across all users and, more notably, significantly improved task completion rates specifically among female participants. The performance improvement for female users was approximately 20% compared to the direct style condition, while no baseline differences between genders were observed in control conditions without the chatbot.

Interestingly, the study found little evidence of users mimicking the chatbot's communication style, suggesting limited linguistic accommodation occurs in these interactions. This challenges assumptions about how humans adapt to AI conversational patterns. The research highlights the importance of developing user- and task-sensitive conversational agents, particularly as AI systems increasingly mediate everyday digital interactions. These findings provide empirical support for communication style personalization as a method to enhance both interaction quality and practical performance outcomes in human-AI collaboration scenarios.