Abstract:

Generative AI (GenAI) has been widely discussed as a tool to stimulate students’ engagement in and learning of STEM content. The increasing societal importance of STEM education further highlights the need for research that supports teachers in developing the knowledge required to use GenAI in STEM teaching and learning in an ethical and pedagogically aligned way. Grounded in the Technological, Pedagogical and Content Knowledge framework (TPACK), this study aims to explore how contextualised interactions with GenAI based on co-created activities, can support in-service teachers’ design of classroom activities including GenAI to promote students’ learning.

The study is based on a co-creation workshop with 11 in-service upper secondary STEM teachers, in which participants, organised in groups by their primary teaching subjects, designed learning activities that included GenAI. The contents represented in this study were related to biology, physics, mathematics and technology education. During the workshop, participants were instructed to test-run the designed activities and refine them based on the interactions with GenAI. The collected data consisted of audio recordings of the co-creation workshop, chat logs from the interactions with GenAI, documentation of the co-created activities, and audio recordings of group discussions conducted after the workshop.

Findings indicate that test-running the co-created activities gave STEM teachers the opportunity to reflect on the outcomes of their interactions with GenAI, enabling them to consider the content-related difficulties students might encounter. This first-hand experience led to adapting the learning activities in relation to the technology’s affordances and limitations experienced. Alongside these adaptations, teachers also discussed guidelines for conducting the activities, including suitable content to address, assessment strategies, and specific support for students. The results also indicate how test-running GenAI-supported learning activities focusing on a specific subject supported the development of aspects of teachers’ TPACK related to GenAI. Examples of this include knowledge about prompting strategies to foster meaningful interaction with the technology in a learning context (Technological Knowledge, TK), awareness of the STEM content-related limitations of the technology (Technological Content Knowledge, TCK), and a better understanding of the knowledge required to support students’ engagement with content in GenAI-supported learning activities (Technological Pedagogical Knowledge, TPK).

One of the main differences between GenAI and other technologies frequently used in school practices is that GenAI is far less predictable; its outputs are not limited by preprogrammed functions that can be consistently applied across different content and subjects. For this reason, test-running learning activities that include GenAI can provide information of how the technology addresses certain content, and as such, might contribute to the development of aspects of teachers’ TPACK.

Keywords:

GenAI, STEM education, professional knowledge, TPACK.