Fuzzy charts are things like pie charts or Wordles that are visually appealing but actually do not convey data properly or provide a good medium for analysis.
"Some of the most confusing new visualizations are the popular network diagrams, which are intended to show connections between nodes and invite inferences about the forces that govern the connections. Numerous groups have produced maps of social networks, internet traffic, and other complicated phenomena, but the impression one gets is merely of connectivity, rather than of any of the patterns the visualization purports to convey. Few obey the principles of perception-informed design or Edward Tufte’s rules for graphical integrity, which state that graphics should make viewers think about the subject matter, not design."
--Greenwood from SEED magazine in the article "Getting Past the Pie Chart", February 18th, 2009.
This is odd, as making networks is what how our brain functions and to make those networks spatially appealing is how we discovered the brain worked in the first place. The brain has 100 billion neurons with over a trillion connections. It has been said that emotion and thought are born here [the neurons of the brain], ie. the neuron connections in the brain is what makes us human. However, saying that displaying data similar to how our brain physically works is a jump in logic that is quite flawed. The reasoning mind is simply a set of binary code or switches. Feelings we value are another story. But our mind must interpret visual data in a hierarchical (computational) format. That is why we value bar graphs and lists to break apart or analyze data. The more we can stick to the latter types of data visualizations the less our brains will have to work on transforming the data from a confusing pie chart to a hierarchical list, thereby leaving time for emotion and creative thought and not get tripped up or worried that we are reading the data wrong.
-More on Edward Tufte's rules for graphical integrity in a later post
-More on Psychology (study of ourselves(mind)) and creativity in a later post
Back to emotions and who we are bring us to a man I very much admire, Alan Turing. The Turing test is build a computer that could chat to a human and have the human not know that you were in fact talking to a computer. Alan thought we have this accomplished by the year 2000. He was wrong. The turing test shows that are brains are not entirely a set of binary code and no computer can match our intelligence. Either because something else is a driving force behind what we do and/or we have emotion (ie. we are irrational).
-More on B.F. Skinner and Behavioralism in a later post.
-More on Alan Turing (esp. the chaos theory) in a later post.
Overall, in terms of data visualization the simplest and most hierarchical are the best. When it comes time to analyze that data in your writing, it is best to leave emotion (as defined in this post) to the discussion. Following this logic, the IMRDC format is actually quite spot on. Something I have grappled with for years.
Quick notes for data visualizations:
NO
-pie charts
-networks
-wordles, or idea clouds
-colors unless needed to recognize a specific pattern
-unnecessary arrangements
YES
-simplicity
-interactiveness
-few colors
No comments:
Post a Comment