Design Considerations for Centering Emotion and Failure in Problem-Solving

• In Event: Inclusion in Computer Science Education: Documenting, Designing, and Supporting Transformative Learning in Computer Science Classrooms

Abstract

Better understanding the experience of learning to code is an area of interest for computer science (CS) researchers and practitioners (e.g., Sengupta et al., 2021). This interest stems, in part, from an acknowledgment that learning to code is a complex emotional endeavor (Kinnunen & Simon, 2010), often marked with explicit signals of success and failure (McCauley et al., 2008), especially during debugging. Perhaps more importantly, educators may not be able to design learning environments to fully curb difficult emotion; frustration and doubt creep in even during the protected space of play (Juul, 2013), peer dynamics in coding can be entrenched in complex power dynamics (Shah & Lewis, 2019), and embracing failure solely as a positive experience risks erasing systemic inequities (e.g., McGee & Stovall, 2016). If we accept the inherent heterogeneity of emotion during learning (e.g., Jaber & Hammer, 2016; Sengupta et al., 2021), how can we then design for CS learning experiences that not only acknowledge emotion and failure as part of CS, but support students in their process of learning to code?
This work takes a phenomenological perspective on the process of learning to code, and acknowledges that learning to code in general, and the debugging process in particular, is a social endeavor situated in talk, body movement, and computer activity in the classroom (Colleague & Author, 2018; Grover & Pea, 2013; Authors, 2010). We present design considerations drawing from the above literature and from a three-year design research project in which late elementary through early high school students learned to code while generating arts-based reflections in summer and weekend workshops. Our design recommendations derive from empirical analyses attending to (1) the narrative structure of students’ arts-based reflections (Authors, 2020); (2) three multi-year case studies of the intersection between identity, emotion, and problem solving (Authors, 2021); and (3) interaction and content analyses of how students position emotion as a debugging strategy and as a transformative outcome of art making (Authors and Colleagues, 2020).
We present what we learned as researcher-practitioners about designing for spaces that supported students in reflecting on their emotions during coding and debugging. Drawing across data sources (e.g., interviews, artists statements, video recordings of coding conversations), we abstracted features of our designs that nurtured reflection on the complex emotional experience of learning to code. Meaningful features of the designed reflection spaces included: a) centering emotion and storytelling as central to the problem solving process; b) cultivating dialogue about the experience of learning to code; c) supporting feelings of uncertainty and surprise through flexible goals; d) allowing materials and open-ended projects to guide activities; and e) creating open communication channels about failure. By designing and studying learning environments that aligned with these design considerations, we found that students offered descriptive accounts of their experiences of failure that transformed how they understood themselves in relation to their coding practice. We describe each of these design considerations and discuss implications for researchers and practitioners, including how to support reflection spaces through art and storytelling in CS learning environments.

Authors

• Maggie Dahn, University of California - Irvine

• David DeLiema, University of Minnesota