Argumentative Maps, GenAI and Multimodality: A Quasi-Experiment to Develop Argumentative Skills in Undergraduates

Argumentative skills are essential for processing complex information (CoI) and fostering critical thinking (CT), particularly in the era of AI-mediated information. Generative AI (GenAI) tools like ChatGPT raise concerns over uncritical use and misinformation, underscoring the need for conscious navigation of information abundance. Argument maps (AMs), traditionally applied to analog texts, are increasingly used to understand and elaborate multimodal and dynamic arguments. However, their interaction with GenAI to refine argumentative approach remains underexplored. This study investigates the combined impact of AMs and ChatGPT on developing argumentative skills, particularly through guided prompt-building techniques that align with AI literacy. A quasi-experimental design involved 27 students from a “Research Methods in Education” course divided into two groups: a control group (G1) using analog texts and an experimental group (G2) engaging with multimodal texts. Students analyzed argument structures to improve critical thinking, interacted with ChatGPT to verify and refine arguments and practiced creating effective prompt to better interact with AI and process complex information. Preliminary results show AMs enhanced CT and CoI in both groups, with G1 excelling in text comprehension and G2 adopting a more reflective approach despite multimodal cognitive load. Interaction data revealed increased confidence in using ChatGPT, with G1 focusing on confirmation and correction and G2 leveraging autonomous and critical strategies. These findings suggest that guided pathways for prompt creation can optimize AI interactions, supporting argumentative skills refinement and fostering critical engagement with intelligent agents. Future research will deepen strategies to enhance these interactions and their impact on students’ learning process.