Dashboarding Resources for COVID-19 Support
Using data viz as a practical tool for community response
It goes without saying that the COVID-19 pandemic has uprooted lives across the world, and everyone has acutely felt the effects in their respective communities. From my vantage in Indiana, nonprofits, community organizations, first responders, and many others have been racing across the state to provide aid to those most affected by the coronavirus outbreak. While various organizations scrambled to collaborate, the onus fell on us to compile resources to aid families that have been ailing. An open source project was quickly created to aggregate and curate the information from these various organizations, placing them in an easily accessible centralized location, ultimately hosted via a Tableau dashboard. I’d like to take a few minutes with this article to walk through how that dashboard of community resources came together.
Data Collection: Together, We Create
Together with my partner, Aubrey Xiong, we started by creating a Google Spreadsheet to keep track of community resources. First, we invited nonprofits to enter information about their own community initiatives.
Many nonprofits and community organizations had developed their own disparate resource pages. As creators of our Google Spreadsheet, we were in a position to structure the resource data and create a protocol for how data would be entered in a consistent manner that we could quickly structure in Tableau.
Next, we sent a resource link to our Google Spreadsheet to leaders of nonprofit organizations and government bodies. We invited them to add resources to our spreadsheet, as we strove to continue growing the data set. Even today, the list continues to grow as more resources are created, discovered and shared.
Aside from the nonprofits we tracked down, government websites had the most up to date information for many other types of resources. We began by looking at their individual municipality websites to find commonalities and categorization opportunities as well as any other items we may have missed.
There are several ways to organize and visualize a database of resources, however we focused on creating an intuitive dashboard that anyone could use, and one that would provide the most benefit with the least amount of extra work on the part of the user. We provided vital information about each resource, embedded in a state asset map. We realized that user interface optimization is heavily dependent on the amount of data being presented, the intent of the message, and the target audience. We chose a simpler interface to increase usability and accessibility across a wide audience.
We put ourselves in the shoes of a desperate citizen who urgently needed help, and consistently applied this mindset while developing to our data set. Our goal was to figure out what people needed most and how we could help deliver this information to them efficiently.
Data Transformation: Structure Determines Function
We love Tableau. We are confident in our ability to develop geographical components, dashboard actions, and display formats we desire in Tableau. Tableau Public is an incredibly accessible platform that allows us to share dashboards without worrying about licensing or platform limitations. In the past, Aubrey and I have also worked with Power BI and structured our data differently to accommodate its data modeling and transformation capabilities. Bottom line — different tools have various nuances and limitations, so be sure to carefully structure the data for the tool of your choice.
We chose the following fields for our dashboard. The fields are organized by categorical columns that function as filters/slicers, and detailed columns that provide the main information.
(1) Categorical Fields
- County
- Resource Type
- Accessibility
(2) Details
- Organization
- Organization ID (Primary Key)
- Org Group ID (for organizations with multiple locations/services)
- Address
- Additional Information
- Open Days
- Fees
- Service Days
- Detail
- Link
- Phone Number
Visualize function
We planned on incorporating a broad set of resources in the data set and wanted a simple drill down structure to navigate to detailed information for each item.
It was important for us to map the user journey on the dashboard alongside our own data exploration, as it allowed us to be mindful of the filters and interactions we wanted to provide. For instance, we noticed that all resources belonged to one of the following categories: Food Resources, Financial Assistance, Healthcare, Childcare, Volunteer Opportunities, and Technology Resources. We decided to make these our Key Performance Indicators (“KPI’s”) that also functioned as slicers for the rest of the dashboard. For each specific KPI we focused on, we spent a great deal of time planning how to best structure any data we aggregated.
Thinking about geography
It sounds basic, but it is imperative to understand the goals of your project, as that drives all of your design and data architecture choices. Are you interested in capturing the resources for a city, a county, and/or state? The type of dashboard and visualization that you create with it will vary depending on the scope of your geography.
For example, the City of Seattle Human Services Department employs a city map to show all available food resources in Seattle. Visualizations that fully utilize a Business Intelligence “BI” tool’s geo-mapping features require a specific list of geographical fields all the way up from country and region down to county and zip code. You may also need to expand your data set to include longitude and latitude to map facilities and locations.
This dashboard was developed for the state of Indiana, so we chose to highlight the entire state and then segment resources by county. In the event that we would map resources down to the street level, we chose to collect latitude and longitude for our resources just in case.
Designing the User Experience
As mentioned earlier, we wanted a simple interactive structure for the dashboard that used drop down filters even my grandmother could navigate. The breadth and detail of information displayed will always determine the optimal layout for any dashboard. We chose the layout below.
The two main visual focal points of the dashboard — our KPIs and the Indiana state map — are the primary drivers of how our users search for services they urgently need, and can physically access.
The middle tab displays high level information including target age demographic, and clicking on a resource from that table will display operational details about the organization in the container to the right. Hyperlink buttons will appear in that right container when a single resource is selected, allowing the user to call, email, or visit the website for more information.
A Community Call: Share Your Work
Once the dashboard was published we began to share this resource to community leaders, nonprofits and now our data visualization family. I hope that others in our data viz family also consider creating similar tools to bring communities together in their respective localities.
The Indiana Covid-19 Resource Dashboard is available for download on Tableau Public and the Google Spreadsheet is public as well, for anyone to contribute to, but also to help provide a template and inspiration for those looking to replicate our efforts in Indiana. Data visualization is a powerful tool for change, and as data viz practitioners, we can be the architects for bringing our communities together in this time of need!
Ingrid Arreola loves all things data visualization and is a data analyst who loves to learn, share her work, and increase data literacy. Connect with her on LinkedIn or follow her data viz work on Tableau Public and Twitter.
Ingrid’s partner on this project, Aubrey Xiong, is an artist turned data analyst who loves creating dashboards that focus on accessibility, storytelling and user experience.
