Bringing Computational Modeling into the Classroom with Custom Block-Based Programming Languages in StarLogo Nova

Abstract

It is possible to improve equity and accessibility in computer science education by incorporating computational thinking into science classrooms through agent-based computational modeling activities that use custom, task-specific programming languages in StarLogo Nova. However, StarLogo Nova’s block-based programming environment does not support extending the language with new task-specific blocks. This thesis resolves this issue by enabling programmers to add new custom blocks to a StarLogo Nova project. The technical contributions of this work resulted in a custom block system that allows StarLogo Nova programmers to create, edit, and view custom blocks, organize them into customizable drawers, and use them to build their models. It is now possible to create task-specific programming languages within StarLogo Nova for the purpose of making computational thinking concepts, such as abstraction, more approachable to learners with minimal programming experience. The conceptual contributions of this work resulted in a new design for a custom drawer interface and a system for sharing task-specific languages across StarLogo Nova projects. The original goal of the thesis was achieved, as custom blocks are now able to be used within StarLogo Nova as a new mode of abstraction within the language. Furthermore, directions for future work to increase the utility of custom blocks as a learning tool were identified and considered. As a result, custom blocks will enable curriculum designers working in the DC-Models project to create customized modeling projects for high school science learners.