The conference on Neural Information Processing Systems (NIPS) has conducted a fascinating experiment: split the program committee into two and get 10% of submissions reviewed by both. The article I’m linking to above has a great analysis of what they found (and it’s not encouraging). [Read more…] about Link: The NIPS Experiment
These postings are primarily links to other things (and their titles link directly to their destination, rather than the short descriptions). Click on the header image if you want to go to the posting itself.
Jon Schwabish is running a blog and podcast called PolicyViz. In both, he talks about communicating data and how to deal with numbers for the general public. He recently had two interesting guests back-to-back on his podcast: Nigel Holmes and Edward Tufte. Both episodes are well worth listening to.
Ben Shneiderman has put together a series of postings about the Pioneers of Human-Computer Interaction (HCI). Each includes a brief biography, some personal notes by Ben, as well as Ben’s photos of them. The latter are particularly remarkable, often going back to the 1980s – like the gem at the top of this page, showing the man himself in 1986.
People from Georgia Tech, INRIA, University of Stuttgart, and other institutions have put together a comprehensive dataset of all papers presented at Vis/VisWeek/VIS since 1990. This was first collected for a set of visualizations last year, but has been updated with the 2014 data. They intend on keeping it up to date. [Read more…] about Link: Visualization Publication Data Collection
In his piece Disinformation Visualization: How to lie with datavis, Mushon Zer-Aviv makes some interesting points about how framing the same data differently in visualization can make a big difference. Using the example of the abortion debate, he shows the usual chart tricks, cherry-picking, subsetting, etc., that is done to make the data support a particular story. [Read more…] about Link: Disinformation Visualization
Wayne Lytle created this video about the Viz-O-Matic that provides lots of tools to make visualization glitzier. It’s a nice little spoof, and a throwback to the computer graphics of the early 1990s (it was made for SIGGRAPH 1993). This video was brought up in a discussion about storytelling at CHI last week, though I don’t think that its lessons are very deep on that subject.
Lena Groeger (of ProPublica) has written a beautiful piece about the Power of Wee Things. She talks about using small things, multiples, and units to display data and get people interested. The article goes through many, many examples covering many different areas and ideas. She also gave a great talk on the topic at OpenVis 2014.
On a somewhat related note, Jake Harris wrote about the importance of individual items in data journalism and visualization, and how to connect with them. The two pieces work very well together to illustrate a way of visualizing data that is often overlooked.
Fernanda Viégas and Martin Wattenberg have written a wonderful piece titled Design and Redesign in Data Visualization about criticism in data visualization. They thoughtfully analyze the practice and point out some of the issues when people create redesigns, including intellectual honesty and perfect hindsight.
They then go on to define some “rules of engagement” for a more reasonable approach to redesign. They argue for a kinder, more respectful, and more balanced process. Their ideas are informed by the critique in design and certainly make a lot of sense for visualization. [Read more…] about Link: Design and Redesign in Data Visualization
Giorgia Lupi and Stefanie Posavec are collaborating on a clever and beautiful new project they call Dear Data (Twitter account). Every week, they are sending post cards to each other with hand-drawn visualizations of data they have gathered: public transportation, ways they communicate, etc.
Scott Klein of ProPublica has written a great story about an early use of data in journalism, and Horace Greeley, the colorful journalist behind it. Greeley found an issue and then gathered the data to show the extent of the problem. This is not unlike today.
In Greeley’s case, the issue was how much money members of Congress were paid for their travels to their home states, despite modern conveniences like railroads that made those journeys much faster than they had been in the past.
The story is very well written and represents an important piece of history and context for today’s practice of data journalism.