F2-STEM2-4 - Can you explain AI to me? Teachers' pre-concepts about Artificial Intelligence

3. Research Full Paper
Annabel Lindner¹ , Marc Berges^1
1 Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU)

Introduction
As a starting point for a professional development (PD) program for teachers, this Research Full Paper investigates which ideas and pre-concepts teachers have about artificial intelligence (AI).

Along with artificial intelligence's increasing relevance in society, it is also finding its way into curricula for K-12. However, in this context, it is not only the students who are confronted with a new topic, but also their teachers. The vast majority of CS teachers have not dealt with AI during their studies and are, therefore, facing a stiff challenge when it comes to teaching it. Consequently, there is an unquestionable need for professional development for teachers in the field of artificial intelligence.

Following the strategy of the Professional Development for CS Principles (PD4CS) initiative, a broad characterization of the content knowledge related to the respective topic is necessary as a first step. For this reason, semi-structured interviews were conducted with teachers with different levels of experience and varying previous knowledge of Computer Science to find out what pre-concepts about AI teachers would bring into a PD program. Based on this, together with a body of knowledge gathered from experts and reference books in artificial intelligence, the necessary AI content knowledge of the teachers is defined.

The semi-structured interviews were analyzed using a qualitative approach based on the text analysis of Mayring. To respond to low intercoder-agreement, we used an approach described by Campell and Quincy.

Investigations of previous knowledge in programming show that this knowledge has a significant impact on misconceptions found with the participants. Bonar and Soloway identified gaps in previous knowledge as one reason for this. Our findings point to a similar effect of previous knowledge in the field of AI. This poses a problem as our participants are teachers who will be AI instructors themselves and who might pass on certain misconceptions to their students. Consequently, the accurate identification and description of the teachers' knowledge gaps is an important starting point for providing a suitable and successful professional development program.

Keywords: Artificial Intelligence, K-12 CS Education, Professional Development for Teachers