Online Platforms and Code/Space: Increasing Access and Equity in Computer Science Education

• In Event: Space, Place, and Learning: The Material Conditions of Education Equity

Abstract

Objectives
“Code” can be considered immaterial as it expresses ideas and data rather than material artifacts. The language of computing, however, involves addresses, loops, space, platforms and other terms that spatialize logical relations while also tapping learners’ spatial sensibilities. This paper reports on a study which explored online learning environments for how they encode spatial relationships.

Modes of Inquiry
We use a software and platform studies (Galloway, 2021; Sack, 2019; Bogost & Montfort, 2007) approach in our analysis. Software studies offers methods to considers technical/formal, historical, creative, social, and critical analysis of software. Platform studies require rigorous attention to the material aspects of a platform to reveal interactions between systems and the creative and cultural works produced on those systems. In Code/Space, Kitchin and Dodge (2011) argue that the production of social space is increasingly linked to code, while code is also written to produce space. Software studies, they argue, has largely ignored the role of space as a conceptual and analytic tool for understanding how and why software matters. Our layered-networks approach helps us attend to the flow of learning across physical and virtual spaces, and the movement of learners in both space, and time as their circumstances change.

Theoretical Framework
By looking at learners’ modes of inhabiting learning platforms, this study traces the way learning involves a co-construction of the platforms, as agents build digital artifacts, new knowledge, and produce the space itself. The ability to write and share code is a key aspect of classical learning theories in computer science (Papert & Harel, 1991), where the ability to collaborate on solutions with others, and to remix existing materials connects with a knowledge building approach to learning (Scardamalia & Bereiter, 2006).

Data & Insights
Three significant online platforms for learning computer science and computer programming were studied: Khan Academy, Kaggle, and CodeSandbox. Khan Academy features the most traditional approach, which can be recognized as an online analog of school-based learning; with lectures, problem sets, and quizzes, now augmented by GPT-4 driven interactive tutors. Kaggle focuses on problem based learning and learning-by-doing, featuring many real-world data sets for learners to explore and competitions to compete in. CodeSandbox does not consider itself a learning platform, but rather a collaborative, online software development platform, however it serves as a significant learning site through public code exemplars, and collaborative software development.

Scholarly Significance
Computer science has a “space” problem in duplicating and in fact worsening the spatial injustices of the world (Crawford, 2022), and failing to dismantle the barriers that have excluded Black, women, Latinx, and others from entering the field. Put simply, the code/spaces where computer science happens does not include many people from these excluded groups: in K-12 schools (Shapiro, 2021), colleges and universities (NSF, 2021), and the high tech workforce. As an attempt to open up conversation on these concerns, this study presents new ways of analyzing the code/space of common learning platforms and associated software, opening up the blackbox of computing, as part of a spatial justice project.

Author

• Matthew X. Curinga, Adelphi University