Dev sprint 5: Data & Downloads

Sprint 5 goals

• Continued API integration work, wiring up most offenses to support queries
• Enabled the downloading of queried data and bulk IBR data sets as a CSV
• Developed content and design for the About page
• Refined auto-generated content in the Explorer view to more clearly explain results
• Made UI/UX refinements to the explorer view and homepage based on user testing

Learning goals

• Are consumers able to understand the results of their query? What factors affect their ability to interpret it?
• What information is useful when downloading incident-data? Does the UI make sense?
• Is information about UCR participation valuable to consumers?
• What information do consumers expect to find on the About page? For those already familiar with the CDE, what are their current expectations for the product?

Participants

• 18F and CJIS team members participated in 3, 1hr remote interviews with a range of prospective users (UCR program manager, data reporter, data journalist / advocate) to better understand how they wish to interact with the CDE. Specifically, these tests focused on refining the Crime Data Explorer and designing the About page.

What we learned

Feedback from Research

• Consumers are enthusiastic about what they see. Maps and visualizations make the data more accessible.
• Our current approach for downloading bulk incident data appears to be working. Savvy users are familiar with working with large data sets in this way.
• Some users have misconceptions about the purpose and goals of the CDE. Specifically, a UCR program manager thought the CDE would show “real-time” data and leverage geo-tagging - neither is planned for the CDE.
• The stability of the app is impacting our ability to test with users. This is because the team has been focused on developing new features as quickly as possible. We need to prioritize some time over the next few sprints for improving stability and performance.
• As the data becomes more readily available through the CDE, we are likely to encounter more data quality issues & obvious differences between what is presented in the CDE & what exists in the CIUS publications.

Messaging for the CDE:

• The data we show is only as good as what we have. Also, it only consists of reported crimes only and is not definitive.
• Expect differences between what is in the CDE & what is published in the CIUS reports.
• Need to explain who participates in the program & how this affects results.
• There’s room to improve how we explain the purpose of UCR and the voluntary nature of the program

Downloads:

• For bulk downloads, the incident ID # is an important anchoring point for analysis
• NIBRS link should include resources: manual, definitions, map
• Need description about bulk downloads but not necessarily joining by pivot tables
• Include a data dictionary and common pitfalls interpreting the data

Next steps

Sprint 6 is about wrapping up core functionality & preparing the CDE for initial beta testing. As the last 2-weeks before the next scope of work starts, Sprint 6 will be followed by a “transitionary sprint” that helps us plan out a strategy for launch and prioritize features for the months ahead. Feature development will start again in Sprint 8.

Prototype tested