Computational Models as Multimodal Participants With Emerging Bilingual Students

• In Event: 31.031 - Multimodal STEM Learning With Emerging Bilingual Students

Abstract

Objectives. We aim to explore how two emerging bilingual students, Garrincha and Bruno, achieve and sustain productive and personally meaningful lines of inquiry with computational models. Using data from a 6th grade STEM classroom, we analyze how students manage and leverage social interactions with computational models and their peers. Our data suggest that interacting with computers and peers as participants within the social learning space of a classroom is important for disciplinary engagement in computational thinking.

Theoretical Framing. To consider the complex social dynamics that occur as students engage with computational models, we use Goodwin’s interactive footing (2007). This framework builds from Goffman’s deconstruction of speakers (1981), which parses the interplay between separate voices within an individual’s speech. Goodwin argues that participation frameworks should extend beyond an individual as a speaker to encompass the embodied actions of others who are present and who contribute to achieve communicative acts. Although computational models are not conscious participants in interaction, they encode and express disciplinary perspectives and intentions of their authors, exhibiting the phenomenon of co-action (Author, 2018).

Methods/Data. These data were collected during the implementation of a 9-week ecology unit. The unit is part of the SAIL+CTM project, which aims to develop a yearlong NGSS-aligned curriculum that integrates computational modeling with a focus on English Learners. Engaging in computational modeling can allow students to build deeper understandings of the causal mechanisms underlying phenomena and develop computational thinking skills (Author, 2013; Papert, 1980; Wilensky & Reisman, 2006; Wilensky & Resnick, 1999). But computational modeling can be challenging, for both syntactic and conceptual reasons. To mitigate these barriers, we use StarLogo Nova, a learning environment that combines visual programming with agent-based modeling.

Results and Scholarly Significance. Our results demonstrate how the computational model allows Garrincha and Bruno to personalize and generatively build on the classroom’s line of inquiry. As Garrincha changes the isopods’ shapes to Bs and Gs (their initials – Figure 1), the boys became newly invested in the isopods’ survival. They explore factors that could contribute to their survival, including size, shape, color, and setup configurations. The model acted as a participant in these interactions, challenging the boys’ assumptions about what mattered for the survival of agents in their models. The results for Garrincha and Bruno are mirrored in the findings for the rest of the class as students engage in efforts to leverage the computational models and their relationships and interactions with other students to support sensemaking.
Understanding the social context of students’ computational participation is critical for designing meaningful learning environments that support all students, including emerging bilingual students (Burke et al. 2016; Kafai and Burke 2013). While existing research carefully chronicles students’ learning opportunities in computational modeling environments, less is known about the ways that students actually participate in such environments in the context of a science classroom. This paper is a contribution because it explores how emerging bilingual students leverage interactions with their computational models and peers to support science learning.

Authors

• Ashlyn Pierson, Vanderbilt University

• Corey Brady, Vanderbilt University

• Doug Clark, University of Calgary