Artificial intelligence literacy in primary education: An arts-based approach to overcoming age and gender barriers

Abstract

Artificial Intelligence (AI) literacy education was previously taught primarily at the university and secondary school levels but has recently started to be expanded to primary school settings. When available at the primary school level, AI literacy is often taught within computer science courses, which may potentially reinforce gender stereotypes and discourage female students' engagement. In AI literacy education, the predominant teaching methods are constructivist approaches, which, while effective in fostering active learning, heavily emphasize technical skills and are therefore limited in their pedagogical scope, as they underplay other important questions, such as disinformation, data justice, and AI's ethical and societal implications. The lack of a clear definition of AI literacy for primary education also raises questions about what to teach and how to teach it. Additionally, little attention has been devoted to understanding gender differences in learning outcomes within AI literacy primary education.This study advocates the use of an arts-based transdisciplinary approach for teaching AI literacy to 25 primary school students. A pilot study utilizing mixed methods was conducted to assess the effectiveness of this arts-based approach. Quantitative analysis through a paired t-test revealed a statistically significant improvement in AI literacy among students participating in knowledge tests. Moreover, the results of the Mann‒Whitney and Kruskal‒Wallis tests indicated that gender and age did not impact pre- and post-knowledge test scores. Qualitative analyses further revealed the pedagogical benefits of the arts-based approach, demonstrating that students enhance their conceptual understanding of AI literacy by reflecting on their artifacts. This study contributes to the literature by providing evidence with a small sample size that the arts-based approach can overcome age and gender barriers to accessing AI literacy education and can serve as a means to teach AI thinking.