Abstract:

The rapid advances in Large Language Models (LLMs) and AI chatbots as convenient user interfaces offer new possibilities in education. On the one hand, schools, universities and educational institutions are exploring how to use AI chatbots such as ChatGPT in the classroom, for instance, to provide personalized and interactive experiences to learners. On the other hand, teachers and educators should be aware of the potential of integrating these AI tools in their own daily work. In particular, AI tools could help with the time-consuming task of further developing teaching materials and in experimenting with different teaching methods.
The paper focuses on the teacher’s perspective and discusses how an AI chatbot can be used as an assistant to create inspiring teaching materials for science, technology, engineering and mathematics (STEM) lessons. We demonstrate how teachers can collaborate with AI chatbots in an efficient way to create exercises tailored to specific target groups. For instance, the AI chatbot could generate a motivational story for a particular age group around a certain task. If the task is to teach a statistical method, the AI chatbot can create an engaging narrative with a real-world scenario. Similarly, for app development, the AI chatbot can build a story around the requirements. By immersing students in a story-driven context, teachers can evoke creativity and encourage critical thinking. Instructions could be adapted to different levels of knowledge, providing more or less guidance for a given task, thus accommodating diverse learning styles and learner abilities.
The use cases for co-creation discussed in the paper focus on addressing girls and women. STEM teaching materials often reflect male-centered perspectives [1]. Female learners are rarely attracted by such exercises [2]. A female-responsive design of learning activities represents girls and women and takes their interests into account. This paper shows how to create tasks in different variations that might better match their interests. This could help integrate all genders into STEM education.
Using an AI chatbot as an assistant requires teachers to expand their knowledge about the capabilities and limitations of LLMs. The paper discusses several aspects that teachers need to consider in order to use AI chatbots in an ethically responsible way. First, teachers need to ensure that task descriptions, instructions etc. generated by the AI chatbot are consistent with educational goals and curricula. Second, teachers must ensure that task descriptions are age-appropriate. Thirdly, and this point is crucial when designing teaching materials for a particular gender, teachers must be aware of the biases that LLMs contain (because of the data they have been trained with), especially gender biases [3]. We will discuss how to be aware of biases and, at the same time, take advantage of them.

References:
[1] P. Medel, and V. Pournaghshband, “Eliminating gender bias in computer science education materials,” in Proceedings of the 2017 ACMSIGCSE Technical Symposium on Computer Science Education, pp. 411–416, 2017.
[2] A. Steffen et al., “An Exploration of Critical Incidents Impacting Female Students’ Attitude Towards STEM Subjects,” in Proceedings of the 6th International Conference on Gender Research, vol. 6, no. 1, 2023.
[3] A. Nadeem; B. Abedin and O. Marjanovic, "Gender Bias in AI: A Review of Contributing Factors and Mitigating Strategies", in ACIS 2020 Proceedings, 2020.

Keywords:

STEM education, AI chatbots, generative AI, teaching materials, women in STEM.