Turning Students into Neurons — Evaluating an Embodied CS Unplugged Learning Activity on Neural Networks

Unplugged learning activities are a low-threshold method to foster AI literacy in K-12 education. “Seven Little Neurons”, an unplugged activity proposed in 2023, aims to teach high school students how feedforward neural networks operate and how they can be trained. The activity employs embodied learning, letting students play the role of neurons and representing what it means for neurons to “fire” or the network to “learn” through movement and physical actions. In this paper, we report on a user study (n=78) building on the activity to investigate students’ perception of the activity’s embodied learning aspects as well as its impact on students’ AI literacy and motivation to learn about AI. Results show a significant increase in students’ self-assessed competence to “create AI” and indicate that the embodied aspects motivate and engage students as well as facilitate their understanding of the discussed concepts, while also calling for further qualitative insights.