DataKind DC's Latest DataDive
May 20, 2016

This past April, 50 DataKind DC volunteers gathered for a weekend DataDive sponsored by Teradata to help four local DC nonprofits use data science to amplify their impact.

DataDives are marathon-style events where teams of volunteers rally together for a weekend to help social change organizations do initial data analysis, exploration, and prototyping. But DataDives are much bigger than weekends. There are weeks of work ahead to scope the projects and prep the data followed by months of impact that ripple out after the weekend is over.  

Read on to see what these teams accomplished and where these projects are headed next. A key theme for all the projects? Open data! 

Targeting efforts to prevent home fires across the U.S.

American Red Cross

DataDives are often the start of a beautiful friendship between DataKind and the organization. This is certainly the case here, as DataKind DC started working with the American Red Cross to help prevent home fires across the U.S. at their last DataDive. (And DataKind originally started working with the American Red Cross on this question all the way back at one of our first DataDives in Chicago!)

A volunteer team led by DC Chapter leader Andrew Brooks and Data Ambassador Hannah Recht built upon work completed at the last DataDive to create key features in the interactive mapping tool to help the Red Cross identify high risk regions for smoke alarm install campaigns. The volunteers created a proof of concept “print view” of the map since printing maps for volunteers to use is an important feature for the organization to have. They also created a table listing the most at-risk tracts or neighborhoods in the region and added in the ability to zoom in on one of those tracts.

These updates will help American Red Cross region managers coordinate volunteer efforts to efficiently identify communities that need smoke alarms installed, areas that are in greatest danger of having home fires, and areas where people are at higher risk of injury and death from fires. Volunteers from DataKind DC and Code for DC will be continuing to work with American Red Cross to improve the tool for their needs.

Check out the latest version of the mapping tool >

Providing better nutrition to DC-area families in need

Capital Area Food Bank

The Capital Area Food Bank (CAFB) is the largest organization in the Washington metro area working to solve hunger and its companion problems: chronic undernutrition, heart disease, and obesity. By partnering with 444 community organizations in DC, MD, and VA, as well as delivering food directly into hard to reach areas, the CAFB is helping 540,000 people each year get access to good, healthy food.

To do this effectively, CAFB needs to have an accurate, real-time understanding of its current stock of high nutrition or “wellness” food. Right now, food donated from individuals and community organizations do not get counted towards wellness foods for CAFB’s partner organizations since volunteers don’t have the time to sort into two nutrition levels - “wellness” or “not wellness.” To bridge this critical gap in their understanding of donated food, Data Ambassadors Elaine Ayo and Paul Forina and DC Chapter Leader Sid Kulkarni led a DataDive team to create a prototype of a barcode-scanning app for volunteers to more quickly classify bulk donated salvage packaged food into these nutrition levels. This will ensure that all food from the Food Bank is properly accounted for in terms of nutritional value and help organizations and CAFB clients make better, healthier choices.

CAFB also needs to know where to target its services to make the biggest impact. Using Census and CDC data, the team began building a model that can create Census tract-level risk scores for diabetes, heart disease, and obesity. This will aid CAFB in targeting wellness foods and other interventions in those areas.

The team is continuing to work with CAFB to create an alpha version of the app and a first version of the diet-related health map. Together, the app and analysis will empower CAFB to provide better nutrition to DC-area families and create a healthier population. Moreover, the above problems are not unique to CAFB, but are actually key problems faced by many food banks across the United States. As such, results from this Data Dive could help further similar solutions for food banks across the U.S.

Check out the github page for more detail >

Understanding Communities in Need

Stewards of Affordable Housing for the Future

Stewards of Affordable Housing for the Future (SAHF) is a network of eleven social enterprise nonprofits that aim to provide high quality, affordable rental homes to over 115,000 households in 49 states, the District of Columbia, Puerto Rico, and the Virgin Islands.

To better understand the communities they serve and how their residents are faring, they approached DataKind to help answer key strategic questions and highlight any unique insights about SAHF households and the neighborhoods they live in.

Using publicly available Census data as well as SAHF’s own outcomes and property data, Data Ambassadors Bright Small and Val Benson led a team to compare SAHF households to others in the neighborhood/community with respect to demographics, income, health insurance and housing stability. The team also explored relationships between health outcomes like hospitalizations and ER visits with other variables.

As just one example, the team explored a question around whether there were any indicators in SAHF’s dataset that might predict how long a resident will live at a property, and/or whether they are evicted. The team found that, out of all residents who have data for "move_out_reason" (n=17,617) 62% are female and 38% are male. Of the males, 5.9% were in households that were evicted or forced out while only 4.3% of females were. Out of all residents who have data for "move_out_reason"& "gender" (n=17,617), the group with the highest rate of eviction are those age 51 to 60 (12%) followed by age 21 to 30 (9%).  Lowest eviction rates were those over the age of 81.

The data exploration generated insights like these for many questions and also sparked new questions for SAHF to use when making strategic decisions to improve and expand their services and support families in need.

See full results and process in the hackpad >

Bringing Healthy Food to Food Deserts Efficiently

DC Central Kitchen

Access to healthy grocery stores is limited in Washington, DC. Of the city’s 43 full-service grocery stores, only three are located in Ward 8 and four in Ward 7, both low-income areas of the city. Through its Healthy Corners program, DC Central Kitchen (DCCK) partners with corner stores to deliver fresh produce and healthy snacks to 67 corner stores in DC’s ‘food deserts,’ where access to nutritious food options is limited. DCCK sells produce to corner stores at wholesale prices and in smaller quantities than a conventional distributor. The stores sell the produce at below-market prices, making it an affordable option for the consumer.

This program also helps generate revenue for the organization and DCCK wondered how they could leverage their own data to understand how their stores are performing so they can ultimately deliver services more efficiently to those in need.

Data Ambassadors Jonathan Bryan, Vivek Khatri, and Astrid Atienza led a team to look at how stores are performing in terms of sales, waste minimization, and product diversity. The team created a clean and comprehensive data file, merging data from four sources, to deliver two things for DCCK. First, they created a store scorecard to compare different outcomes for a store across multiple months. Secondly, they created a product item scorecard to to look at sales and waste data by store and product type to determine product efficiency.

Both scorecards will help DCCK understand how its stores are performing so they can improve efficiency across these areas in need. Going forward, DCCK will be working with DataKind volunteers to identify indicators of store performance and begin working on predicting which new stores will perform the best.


Get Involved

A huge thanks to everyone that volunteered their weekend, to these four nonprofits doing incredible work and to Social Tables for hosting. A special shout out to Teradata for making this event possible and for its long-time support of DataKind!  

In DC looking to use your skills to give back? Join DataKind DC’s Meetup and stay tuned for their next event!

Join DataKind DC >

Read more posts
January 11, 2022
Our Ethics + Responsible Data Science Practices at DataKind
At DataKind, we take an expansive definition of data ethics and responsible data science as broad terms that can be used to describe the appropriate handling of data...
Read full story
December 21, 2021
Lessons from DataKind San Francisco’s Launch of DataAdvisory Projects
From financial forecasting to targeted advertisements, advancements in data collection and analysis have benefited a myriad of for-profit organizations today.
Read full story
October 14, 2021
Celebrating DataKind’s CEO: An Interview with Lauren Woodman
We’re thrilled to welcome Lauren Woodman as the new CEO of DataKind. She brings to the role over 25 years of experience working at the intersection of technology, development, policy, and NGOs...
Read full story
December 20, 2021
Shining a Light on Community: Looking Back at DataKind’s Virtual DataDive® Event
We hosted a DataDive® event in fall 2021, and with it being the season of giving, we thought what better time to share some highlights and express our deepest gratitude to our partners, volunteers, and sponsors...
Read full story
Blog Archive