AI-powered Chatbots: Effective Communication Styles in Sustainability Conversations
Two pre-registered experimental studies demonstrate that AI-powered chatbots utilizing a Motivational Interviewing style are more effective than those using a Directing Style in fostering user engagement and interest regarding sustainability issues, though both styles yield comparable learning outcomes.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine a world where the path to a sustainable future depends less on lectures and more on conversation. For decades, scientists and policymakers have known that telling people what to do often fails to change how they act. Instead, the most effective way to shift behavior is to help people find their own reasons to change. This idea, known as motivational interviewing, is a counseling technique where a guide asks open questions and listens deeply, helping the other person explore their own values and resolve their mixed feelings about a difficult topic. It stands in contrast to a more traditional, directive style, where an expert simply provides instructions, facts, and a clear plan of action. While this approach has long been used in healthcare to help people quit smoking or manage weight, a new question has emerged: can a computer program, specifically an artificial intelligence chatbot, use these human-like conversation skills to help us care more about the environment?
Two researchers, Ennio Bilancini and Leonardo Boncinelli, along with their colleague Eugenio Vicario, set out to test this question. They wanted to see if an AI chatbot could effectively discuss the United Nations' Sustainable Development Goals—a global list of seventeen targets for a better world—and whether the way the bot spoke would change how much people engaged with the topic. They designed two distinct ways for the chatbot to talk to human participants. In one scenario, the bot acted as a supportive partner, using the motivational interviewing style to ask questions that encouraged the user to think deeply about why sustainability mattered to them. In the other scenario, the bot adopted a directing style, acting more like a teacher or a manager who provided clear information and tried to convince the user of the importance of the goals.
To find out which method worked better, the team ran two large online experiments involving nearly 1,600 people. Each participant was randomly assigned to talk to one of the two chatbot styles about sustainability issues. The conversations were short, lasting less than five minutes on average, and took place on a standard web browser. The researchers measured three things: how much the participants felt they had learned, how interested they felt in the topic afterward, and how much they actually wrote during the chat. The amount of writing served as a direct measure of engagement, showing how much effort the person was willing to put into the conversation.
The results revealed a clear difference in how people responded to the two styles. Those who spoke with the chatbot using the motivational interviewing style wrote significantly more words than those who spoke with the directing style. This suggests that the empathetic, question-based approach kept people more involved and willing to keep the conversation going. Furthermore, participants in the motivational interviewing group reported feeling a higher level of interest in sustainability topics after the chat compared to the directing group. They felt more curious and engaged with the subject matter.
However, the study also found a surprising limit to the power of conversation style. Despite the higher engagement and interest, the researchers found no measurable difference in how much people felt they had learned. Whether the chatbot was a supportive listener or a direct instructor, the participants' self-assessment of their new knowledge was essentially the same. This suggests that while a friendly, inquisitive tone is excellent for getting people to care and to participate, it does not necessarily make them feel they have absorbed more facts than a straightforward, instructional tone would. The study indicates that the "how" of the conversation matters deeply for motivation and attention, but perhaps not for the immediate feeling of having gained knowledge.
These findings offer a practical insight for anyone trying to promote environmental awareness. Using an AI chatbot that listens and asks questions rather than one that simply lectures appears to be a more effective way to spark interest and encourage people to spend time thinking about sustainability. The study does not claim that this method solves the climate crisis or that it creates permanent experts, but it does show that the tone of a digital interaction can significantly influence a person's willingness to engage. By letting people explore the topic on their own terms, even through a brief chat with a machine, we may be better able to reach the hearts and minds of those who need to be part of the solution.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.