A short summary of the project’s aims:

This research investigates political discourse in online spaces, with a focus on deliberative styles—specifically, justification and reciprocity—and their role in fostering reasoned dissent. These styles, integral to effective political discourse, guide users in engaging with opposing viewpoints through constructive, evidence-backed dialogue. The project leverages computational linguistic approaches applied to discussions on online forums, survey experiments and field experiments to assess how LLMs’ rhetorical strategies impact user engagement, argument quality, and the broader dynamics of political conversation. One key outcome of this research is the creation of a comprehensive dataset, offering deeper insights into the interaction between style, content, and persuasiveness, and chatbots providing practical implications for enhancing LLM-driven deliberation across various online platforms.