CNI Fall 2013 – Visualizing: A New Data Support Role For Duke University Libraries

Angela Voss – Data Visualization Coordinator, Duke Libraries

Data visualization can be typical types such as maps or tag clouds, or custom visualizations such as parallel axes plots. Helping people match their data to their needs, and what they want to get out of their data. Also help people think about cost/benefits of creating visualizations.

Why visualize?

  • Explore data, uncover hidden patterns. e.g. Anscombe’s Quartet.
  • Translate something typically invisible into the visible – makes the abstract easier to understand, increase engagement. Important to people performing research as well as reporting to others.
  • Communicate results, contextualize data, tell a story, or possible even mobilize action around a problem. (see Hans Rosling: The River of Myths). Important to build context around data, not just think that the numbers speak for themselves.

Visualization at Duke

  • No single centralized community, but plenty of distributed groups and projects.
  • Library was already offering GIS help.
  • Who could support visualization? Faculty/department? College/school? Campus-wide organization – was the only option with wide enough reach. There were several options – Duke created a position that reports jointly to Libraries and OIT.
  • Position started in June 2012 – Dual report to Data and GIS Services in the Libraries and Research Computing in OIT.
  • Objectives: instruction and outreach; consultation; develop new visualization services, spaces, programs.

After 18 months, what has been the most successful?

  • Visualization workshop series – software (Tableau (full time students get software free), d3 (Javascript library)), data processing (text analysis, network analysis), best practices (designing academic figures/posters, top 10 dos and don’ts for charts and graphs). The barrier is understanding data transformations to get data into software
  • Online instructional material
  • Just-in-time consulting – crucial to people getting started.
  • Ongoing visualization seminar series – this had been happening since 2002. Helped introduce the community.
  • Student data visualization contest

d3 monthly study group – Using GitHub to share sample code. Using Gist and to see the visualization right away. e.g.

Top 10 Dos and Don’ts for Charts and Graphs:

  • Simplify less important information
  • Don’t use 3D effects.
  • Don’t use rainbows for ordered, numerical variables. Use single hue, varying luminance.

Just in time consulting

  • Weekly walk-in consulting hours in the Data & GIS Services computer lab
  • Additional appointments outside of walk-in hours
  • Detailed support and troubleshooting via email

Weekly visualization seminars – Lunch provided, speakers from across campus and outside. Regularity helps. Live streaming and archived video.

Student data visualization contest

  • Goal: to advertise new services, take a survey of visualization at Duke – helped build relationships across the campus.
  • Open to Duke students, any type of visualization
  • Judged on insightfulness, narrative, aesthetics, technical merit, novelty
  • Awarded three finalists and two winners. Created posters of the winners to display in the lab, and run them on the monitor wall.

After 18 months, what are the challenges?

  • Marketing and outreach – easy to get overwhelmed by the people already using services at the expense of reaching new communities.
  • Staying current – every week there’s a new tool.
  • Project work, priorities – important to continue work as a visualizer on projects.
  • Disciplinary silos and conventions
  • Curriculum and skill gaps – there aren’t people teaching visualization at Duke as a separate topic. Common skill gaps: visualization types and tools; spreadsheet and/or database familiarity; scripting; robust data management practices; basic graphic design

Hopes for the future

  • Active student training program (courses, independent studies, student employment)
  • Additional physical and digital exhibit opportunities
  • Continued project and workshop development

What should a coordinator know?

  • Data transformations
  • Range of visualization types, tools
  • Range of teaching strategies
  • Marketing

What should a coordinator do?

  • Find access points to different communities
  • Use events to build community
  • Collaborate on research projects
  • Stockpile interesting datasets
  • Beware of unmanaged screens
  • Block out plenty of quiet time for the above

How should an organization establish a new visualization support program?

  • Identify potential early adopters
  • Budget for a few events, materials, etc
  • Involve othe service points
  • Provide a support system for the coordinator
  • Expect high demand

Working primarily with staff and grad students, this quarter a lot of undergrads due to a few courses.

Angela’s background is in communication for the most part. There’s a IEEE visualization conference.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: