The proliferation of digital technologies in education has led to an explosion of data being generated as a byproduct of teaching and learning activities. Everything from online learning platforms to intelligent tutoring systems capture huge amounts of data on how students interact with educational materials. However, this "big data" on its own does little to provide educators with actionable insights. The data is often unstructured, spanning multiple formats and contexts, and in volumes too large for a human to make sense of. Simply having lots of data does not necessarily make it useful.

To take advantage of these digital data trails, educators need ways to structure the data to ask meaningful questions that can inform teaching practices. Data visualisation provides tools and techniques for organising data and presenting it visually to reveal patterns, trends, and relationships. Rather than peering at spreadsheets or databases, data visualisations use charts, graphs, maps, and other graphics to highlight key aspects of complex data sets. Effective data visualisations make it easier for educators to grasp key takeaways and generate hypotheses to test.

However, there are important considerations in how data is interpreted and presented through visualisations. The choices involved in data cleaning, analysis, and visualisation methods can influence the story that emerges. Educators need to approach data visualisations thoughtfully, being aware of limitations and potential biases. Visualisations should be used to open up inquiry and discussion rather than suggest definitive conclusions. When designed well, data visualisations do not replace human judgement but rather provide a springboard for deeper understanding and improvement.

As digital learning proliferates, educators will need to harness data visualisation approaches. This will enable them to glean insights from learning analytics data to evaluate and enhance educational experiences. Data visualisations make the overwhelming volumes of data accessible and actionable. However, care must be taken to ensure visualisations enlighten rather than mislead.

This week's reading introduces the Live Data project at the University of Oxford, funded from 2015-2017, which aims to support researchers in creating interactive data visualisations. Visualisations are seen as vital for bridging the "data gap" between raw data, publications, and wider understanding. The project recruits case studies across disciplines and develops generalisable workflows for visualising data from repositories or uploads. The article outlines services explored including Plotly, Shiny, and Tableau Public. These allow embedding interactive charts, graphs, and maps into publications and websites. Benefits highlighted include communicating insights efficiently leveraging visual processing of the brain. But care is needed to avoid misrepresentation through visualisation choices. Prototypes covered include visualising letter networks in the humanities and cancer research collaborations at Oxford. Training materials and a showcase website are other outputs being developed. The goal is growing an ecosystem for open, interactive visualisations promoting data reuse and accessibility at Oxford and beyond.