Drawing Inference From Data Visualisations

Abstract

The importance of statistical education coupled with the emergence of powerful visualisation tools - has led to some reconceptualization of teaching statistics (Ridgway, Nicholson, & McCusker, 2013). Students need to become familiar with reasoning about multiple variables, covariation between multiple variables, and the use of complex visualizations to represent quantities in new ways, using intuitive visual artefacts. However, we know little from empirical research about how students interpret the multivariate view of data within data visualization tools and even less about students' ability to meaningfully construct data visualisations that highlight important aspects of data. 43 students of grade 8-9 (2 classes) were asked to select appropriate quantities represented in the data to construct visual representations that would help them to make informed decisions about the following task: "Preparing to live in an unknown country in the future: Which is the best country to live and work abroad? Consider different variables that impact on your decision." Students were paired to discuss how to visually represent data sets as they engaged within the Gapminder data visualization tool that has built-in access to large global datasets regarding economy, education, energy, environment, health, infrastructure, population, society, and work. The results suggest that the collaborative task challenged students to construct visual structures that highlight relevant information for their analysis task.