How to share your data interactively using Datapages — Stanford

Data sharing is both a critical scientific need and, increasingly, a mandate by many research funders. The FAIR principles – that data should be findable, accessible, interoperable, and reusable – are a widespread guide to how data are shared. But sharing data in accordance with these principles presents a challenge for many researchers, especially those with limited technical expertise or resources. If datasets are shared, it's most often as static files, decreasing impact. Some high-profile datasets are shared on custom webpages with interactive tools, but these can be complicated and expensive to create and maintain. We bridge this gap by introducing a set of open-source tools and templates for easily creating Datapages, interactive websites to disseminate de-identified data for broad reuse. Datapages enables researchers and other data distributors to easily share versioned datasets along with interactive visualizations, rich documentation, and user-friendly access functionality. A Datapage is a web-based interface to a dataset – a way that users can learn more about it and access it as seamlessly as possible. This functionality is enabled by storing the data in Redivis, a data platform for academic research which provides datasets with versioning, flexible access controls, and multi-platform API access. The Datapage itself is a website (built using the Quarto publishing system) that is connected to one or more Redivis datasets and is automatically populated with interactive visualizations (built using Observable Plot), as well as an embedded data browser, documentation on programmatic access, and information on citation, DOI, metadata, and more. An automated deployment pipeline renders this website and publishes it as a static site, so it can be hosted on a serverless platform such as GitHub pages. As such, a user without much technical expertise can create a Datapage for their dataset without writing any code or setting up any deployment infrastructure. More advanced users can build out their Datapage with custom visualizations, theming, and any content they may wish to add. We've created a diverse array of Datapages for datasets from psychology, linguistics, economics, education, and atmospheric science (datapages.github.io). We plan to showcase these, demonstrate how to create a Datapage, and share key design insights, with the aim of promoting and improving data sharing.