Application of Embodied Cognition to Research on Child-Robot Interaction

• In Event: 28.026 - Designing Robots for Children: Multisensory Interactions and Multimodal Assessments

Abstract

This presentation will discuss the theory of embodied cognition as a theoretical framework for CRI research and design, addressing Q2. What are the cognitive and social theoretical perspectives that may guide research on child/robot interaction (CRI)? A recent view of cognition emphasizes that our perceptual experiences, such as seeing, hearing and feeling, are vital elements to reach conceptual understanding because our thoughts and knowledge emerge from dynamic interactions between our body and the physical world. The primary lines of research have discussed embodied cognition for many years. Cognitive linguistics, which explains the cognitive process of understanding linguistic expressions with metaphorical mapping process (e.g., Lakoff & Johnson, 1999), Cognitive psychology explains how people understand a new concept with perceptual simulation based on prior multisensory experiences (e.g., Barsalou, 2008). Based on this emerging paradigm of understanding human cognition created by merging these two lines of research, educational researchers have investigated the effectiveness of interventions that provide perceptual and embodied experiences in learning. Examples of embodied interventions include the use of gestures (Hu, Ginns, & Bobis, 2015) and bodily action (Sidhu & Pexman, 2016), which were often enhanced by technologies such as multimodal simulations including haptic sensations (Han & Black, 2011), touch screens (Agostinho et al., 2015), or augmented reality (Lindgren & Johnson-Glenberg, 2013). In addition to these technologies, educational robots have gained scholarly attention as emerging technology that has a potential to facilitate young children’s learning. While previous studies on embodied cognition can provide implications for designing and developing educational robots as well as child/robot interactions, little research has been done to connect these two evolving areas.

Young children’s interaction with robots using gestures and bodily actions is one area to be explored based on embodied cognition. Considering that the majority of recent studies have examined the effectiveness of embodiment in mathematics, science and language learning, educational robots can also be applied to conceptual learning. For example, the use of gestures has been researched as evidence that the body is involved in thinking (Alibali & Nathan, 2012). In mathematics learning, in particular, an educational intervention that fosters the use gestures that correspond to mental representations required to solve problems can facilitate creation and elaboration of new ideas (Goldin-Meadow, Cook, & Mitchell, 2009). Also, bodily actions and behavioral performances enhance learning in language acquisition (Lan, Fang, Legault, & Li, 2015) and science (Lindgren & Johnson-Glenberg, 2013). Thus, robots can be used as a learning companion to facilitate learners’ cognitive processes by modeling gestures/actions, evaluating learners' gestures/actions to assess their level of conceptual understanding, providing proper gestural feedback, or recording gestural data.

Considering young learners tend to learn better with multisensory experiences and perceptual/bodily activities and are motivated by the use of educational robots, research exploring the intersection between embodied cognition and robots will help us design and develop appropriate interventions for young learners.

Authors

• Insook Han, Temple University

• Michael Tscholl, University of Wisconsin - Madison