When you’re analyzing data to find solutions, the problems are different but the procedures are similar. The NDP team uses this Use Case Framework to outline assignments and ensure a useful outcome.
First, Identify a Pressing Question. Working with our partners, we identify and articulate a pressing question they are trying to answer. Everything flows from a well-considered question.
Second, Aggregate & Curate Data. “Hyper-local” data is usually needed to answer a partner’s question, but the data generated by small businesses and local government is often unstructured and sometimes incomplete. We get reliable data organized for analysis while we also search for data generated by outside entities. This data usually needs to be cleaned and structured. Sometimes we match or merge data sets, and we often use mobility data to understand the origins and destinations of the island population.
Third, Undertake Analysis. The analytical team will usually include data scientists, demographers, a writer, and graphic artists to provide visualizations. The goal for each use case is to identify at least three key insights, although there are often many more.
Fourth, Create Visualizations. The data visualization techniques and programs we use include (but are not limited to) Tableau, D3, Python, and Matlab. We can prepare static visualizations, interactive visualizations, and data dashboards.
Fifth, Provide Support. We work collaboratively with our partners in an iterative, dynamic process. We frequently provide updates and require feedback. We write summary memos when needed to support identified and actionable insights.
Sixth, Enter a Subscription. We use a subscription model to keep data dashboards up to date.