Generative AI introduces complexity by offering potential support for human skills development, while also generating noise and false information in a postdigital context. Grounded in a Vygotskian perspective, this study explores the combined use of argument mapping (AM) and ChatGPT as mediational tools for developing argumentative skills as a proxy for critical thinking in higher education. These tools are conceptualized as socio-technical assemblages providing double stimulation within students’ Zone of Proximal Development (ZPD), addressing how multimodal texts and AI mediate comprehension of information (CoI) and critical thinking (CT). Adopting a case study approach with a quasi-experimental design, the research involved 17 female undergraduate students from the University of Padua, divided into three groups: G1 working with analog texts, G2 with multimodal texts, and a control group interacting only with ChatGPT (G3). Chatbot interactions were analyzed to explore its potential to support reflection and personal information re-elaboration. Results indicate improvements in comprehension and critical thinking, especially in the multimodal group (G2). G1 achieved weaker outcomes, possibly due to limited external stimulation. G3 outperformed G1 but showed stable yet comparatively lower results in advanced argumentative reworking despite a positive median. Overall, AM seems to support text comprehension and meaning reconstruction, while multimodality fosters the integration of multiple perspectives. Generative AI can further support critical engagement and understanding of AI systems when embedded in a structured pedagogical design rather than used “in the wild”.
Crudele et al. (Thu,) studied this question.