Different students move differently: Insights for the design of embodied learning technologies

• In Event: Learning and the Growing Body: Embodied Learning Research Across Different Age Groups

Abstract

We extensively use our bodies for learning, to communicate, represent concepts, alleviate cognitive load, or even regulate our emotions (Wilson 2002). As such, various approaches have been explored to design learning activities supporting embodied cognition (Tran et al. 2017). Concurrently, Human-Computer Interaction researchers have explored how physical context, and our bodies, must be leveraged to design meaningful interaction with digital objects (Dourish 2001, Mueller at al. 2018). However, previous work has shown that designing meaningful embodied interaction does not necessarily support learning, and that we lack an integrated framework encompassing both embodied cognition for learning and embodied interaction (Chatain et al. 2023, Chatain et al. 2022).
In this work, we take an exploratory approach to provide a first set of design considerations, addressing the following research questions: (1) How do students move when making sense of mathematics? (2) How can interaction design support and leverage such bodily actions?
We limit our scope to the topic of derivatives as it has vast applications, can be understood in various ways (slopes, speed, variation rates, etc), and has been widely explored in embodiment research.
Learners learn from two types of bodily actions: spontaneous bodily actions, such as gestures, spontaneously performed during sense-making, and directed bodily actions, performed as implicitly or explicitly requested by the learning activity (Walkington 2022). We captured both types of bodily actions through two user studies (Chatain et al. 2024).
For spontaneous bodily actions, we performed a qualitative study with 6 participants of various math ability, math affect, and experience in math teaching. We designed 4 tasks, emphasizing derivatives as slopes, speed, variation rates, and mathematical objects respectively. We analyzed speech and gestures using thematic analysis (Braun & Clarke 2006, Clarke & Braun 2021).
For directed bodily actions, we conducted a user study with 130 high-school participants, using a digital embodied learning activity on derivatives (Chatain et al. 2022) both in Virtual Reality (VR) and on tablet. We analyzed the movements and sequences of movements (Scholz 2016, Ching 2006) of learners with various math ability, math anxiety (Hopko 2003), and body awareness (Shields 1989).
From both studies, we aggregated the following design considerations: (1) Expand embodied interaction design beyond position and movement as learners also leverage muscle tension to represent mathematical concepts; (2) Consider embodied metaphors as learners spontaneously integrate metaphorical gestures; (3) Allow for coarse gesturing as it supports identification of deep features and is a sign of mastery; (4) Support and evaluate sense-making anchors by carefully considering which interactive elements should be visible and which should be imagined; (5) Integrate embodied in-VR assessments as these support math anxious and highly body aware learners, and allow identification of preliminary learning.
In future work, we want to explore how embodied design differs between different age groups as practical considerations and learning spaces vary. While classrooms for young children often include space and opportunities for movement, it is not true at university level. Similarly, as adult learners have limited availability, design for embodied learning must shift from exploration to value creation.

Authors

• Julia Chatain, Singapore-ETH centre

• Manu Kapur, ETH Zürich