The design space of visualization tools for data science education: Literature review and framework for future designs

Abstract

Data science education is a rapidly growing field, motivating the development of several tools aimed at engaging students in data-related activities across all educational levels. Within this landscape, new block-based programming (BBP) environments have been developed to integrate data visualization functionalities. However, the construction of visualizations entails distinct goals and forms of interaction from those typically used in conventional programming, which should be considered in designing those environments. Drawing inspiration from taxonomies within the field of Information Visualization, we derived a framework for the analysis of educational tools in this domain comprising four categories: supported goals, expressiveness, abstraction, and transparency. We situate six existing BBP environments within the design space of the framework and illustrate its application by describing the design of PlayData, a BBP environment developed to address a gap in data visualization tools for young audiences. Our goal is to contribute to a more theoretically grounded discussion on the topic by connecting well-established research on professional tools for data visualization and the design of related educational environments targeted at middle and high school students.