Integrating Generative AI into Critical and Creative Thinking: The CRAIL Model

This study investigates the development of Critical and Creative Thinking (CCT) skills among secondary school students, using a collaborative learning model supported by Generative Artificial Intelligence (GenAI), the CCT_GenAI Model. This research adopts a mixed methods approach, combining quantitative and qualitative methods to assess the initial and final levels of CCT in 23 students from a Year 12 class.

Quantitative data were collected through questionnaires that measured CCT skills and dispositions. The qualitative data collected included observations made over three months within the classroom context and an interdisciplinary framework, as well as group interviews and students’ self-assessments. The instruments and the results of the quantitative analysis, in conjunction with the qualitative data, served as the basis for creating a coding book. This coding book was defined and explored during the qualitative phase to code the collected data. In the inductive aspect of the study, ChatGPT was used in an initial and exploratory phase to identify categories (codes), and the webQDA software was employed to internally validate the results generated by ChatGPT. The coding book was also validated by external experts using the Delphi technique.

Using the coding book brought additional advantages, such as increased efficiency in the coding process and improved validity and reliability of the analysis by providing clear guidelines for the codes’ application. Furthermore, the robust coding structure contributed to a deeper understanding of learning dynamics and the factors influencing the development of CCT.

The quantitative results indicated a moderate improvement in CCT skills and a significant increase in critical thinking dispositions, particularly among students with initially lower scores. The qualitative data allowed for a deeper understanding of the quantitative results, concluding that a mixed methods approach, through a structured coding process, provided an effective means of investigating and promoting the development of CCT in GenAI-mediated educational contexts.