Maps and AI: a lab to support undergraduates' argumentative skills

Argumentative abilities are required both individually and professionally to digest complex information (CoI) related to the critical reconstruction of meaning (critical thinking - CT). This is still a particularly important objective, especially in the age of social media and artificial intelligence-mediated information. Recently, the advent of the generative artificial intelligence (GenAI), with the specific example of ChatGPT (OpenAI, 2022), has made it easier to obtain and share knowledge, and new tools are sorely needed to deal with the abundance of post-digital information without getting lost.

The research activity aimed to investigate the extent to which AMs, already used to develop argumentative skills and critical thinking, helped students increase their level of comprehension of multimodal information and their ability to rework their own opinion about it. Juxtaposed to this line of thinking was the idea of adopting the ChatGPT intelligent chatbot to extrapolate and analyze whether an initial communicative and argumentative interaction with an intelligent agent would help students reframe their arguments. 

Our study was based on a two-group quasi-experiment with 24 students in the Education Evaluation teaching, from the Bachelor of Science in Education and Training, Education and Human Resource Development curriculum, University of Padua (L-19). The control group (G1) conducted the information comprehension and reprocessing course in an analogue context, and the experimental group (G2) was exposed to multimodality with texts and tutorials bearing dynamic and digital elements. Both groups then interacted with ChatGPT, integrated within an interactive lecture in the middle of the course. 

Preliminary analyses showed that in both groups, AMs improved students' text comprehension and critical thinking. The group that used analog texts (G1) showed more uniform improvement than the group with multimodal texts (G2), which had greater variability in results. However, multimodality stimulated deeper and more integrative reflection. The use of ChatGPT showed potential to improve information reprocessing, with students recognizing the usefulness of AI and taking a critical approach to its interaction. Future research will aim to better integrate these tools to maximize learning.

This Zenodo record follows the full analysis process with R (https://cran.r-project.org/bin/windows/base/ ) composed of the following datasets and script:

1. Comprehension of Text and AMs Results - Arg_G1.xlsx & Arg_G2.xlsx

2. Opinion and Critical Thinking level - Opi_G1.xlsx & Opi_G2.xlsx

3. Data for Correlation and Regression - CorRegr_G1.xlsx & CorRegr_G2.xlsx

4. Descriptive and Inferential Statistics Comprehension and AMs Building - Comprehension&AM_VAL.R

5. Descriptive and Inferential Statistics Opinion and Critical Thinking level - Opinion&CriticalThinking_VAL.R

6. Correlation and Regression - Analysis_VAL.R

Any comments or improvements are welcome!