In Information Architecture, Modeling
Summary: Kaarin Hoff and Daniel O'Neil's give solid practical tips for listening to data. Data visualization starts with a conversation-based, data-driven strategy.

Daniel O’Neil and I had the pleasure of giving the keynote at this year’s A2 Data Dive. Our talk, Listening to Data, focused on how data visualizations must be conversation-based and data-driven.

Conversation based
Picture of the room
By keeping this main tenant in mind, the core principles of visualizations suddenly seems obvious. Of course we must be Clear, Useful, Ethical, and Credible when having a conversation. When you talk to a friend you don’t mumble useless things, lie, or exaggerate—so don’t do it in your visualizations. It is easy to treat visualizations as a conversation input, as if you will always be there to narrate your visualization and correct any misconceptions. But why would you want to? You can make a visualization good enough to allow people to have a conversation with your visualization alone. This will allow you to have efficient meetings where you aren’t saying “no, what is meant here is…” and to put it on your website confident that it will convey your story to the world.

We challenged attendees to “Hear the Who” like Horton. He listened to the data and exposed it to the world.

We discussed a lot more than what is mentioned above including how visualizations are all rhetorical and abstract – and should provide a the multiple punch of emotional, cognitive, analytical, and visual impact.

Twitter Notes:
– “You’ve been doing data viz your whole life” @KaarinH. Ever made a bar chart? You’ve visualized data
– “Your data has a point of view and wishes to start a conversation.”
– Visualizations should be “clear, useful, ethical, credible”
– Try not to find data to support your argument. Find the data and then formulate an argument
– “The bell curve’s not boring!” – Kaarin Hoff // This says something important about how we think the world works.

Share on
Recommended Posts