Over four weeks in January and February, 13 first and second year students participated in Rachel Abrams’ MTA Big Data workshop. Devised for and with the Metropolitan Transportation Authority (MTA), the workshop’s goal was twofold:
First, to introduce the public agency to the principles, methods, tools and activities of data visualization, so they can harvest greater value from the information that New York City Transit (and other parts of the MTA) generates about itself 24/7, year round.
Secondly, to let the students loose on content with real-world context and impact, to give them an insider’s view of a complex piece of New York’s public infra- and info-structure, and explore ways to make the transit information landscape more legible for its executive decision makers, municipal operators and passengers.
Two of the four workshop sessions included much-appreciated masterclasses by guests of the moderator: Jennifer Kilian (Facebook) on prototyping and Eddie Opara (Pentagram) on data visualization.
Five topics from the Subway
With preselected topic areas, five project groups gathered, cleaned, organized source data to represent stories in visual prototypes about each issue. Each group identified a general theme, hunches and questions to explore, aspects of the data to visualize and next steps to develop deeper inquiry into their given topic. From there, the whole team suggested key recommendations for the MTA, to highlight best practices the agency could adopt to make more of the data riding around New York produces every minute of every day.
Addressing these topics, students:
- Created a snapshot of Grand Central Terminal subway station by ridership, based on turnstile data
- Analyzed the reliability of the two major manufacturers’ train cars
- Investigated the relationship of track fires and garbage-clearing schedules
- Identified the key causes of delays during Monday morning rush hour
- Mapped all named sections of track to every line within the network
Students’ key recommendations for the MTA included:
- What makes data glorious is that it’s the raw material of evidence: Key to making cases that guide decisions
- Start with just enough data, clean it, structure it. Identify what decision-makers really need to know, and collect data to support just that.
- User-centered design means start with decision-makers’ and operators’ real needs and establish a common visual grammar and vocabulary everyone can use
- Limit jargon and acronyms: Use simple language regular people can understand
- If you can’t draw conclusions, spot outliers, tendencies and patterns that send the data analysts on a new hunt
- Engage users of visual data progressively: Start by establishing those visual basics, then build up layers of complexity from there to tell the full story.
- Consider scale – showing macro and micro views reveals details in a big picture.
- Scorecards and personas are an excellent way to summarize attributes of an issue and of typical users
- Getting stuck and identifying what’s missing can provide useful clues
- Standards for collecting, organizing and representing content are all critical for managing data
- Tell a story for people, not data for data’s sake
- Like gold, the value of data is not in mining it, but in smithing it.
In conclusion, both the students and the MTA agreed, we’d accomplished the first steps to demonstrate how to make the network legible and to make the users literate. It’s a massive undertaking to to embed user experience within functionally siloed, complex, always-on public agencies, but together, we’ve shown the potential design strategy and craft can offer in that process.