In this chapter Kieran Healy has grouped bad visualization figures mainly into three categories: aesthetic (visual), substantive (contents) and perceptual (psychological perception).
He points out that for most visually bad visualizations, it’s not as efficient as other because of duplicated labels and pointless effects. I strongly agree with this opinion, if viewers can be easily distracted by unnecessary design features and having a hard time to interpret the data, then visualization itself has lost its initial intention already. Jon Maeda once said in his book Laws of Simplicity that ‘the simplest way to achieve simplicity is through thoughtful reduction’. It seems to be applicable in any type of design, in the Figure 1.4 do we really need the shadow and three dimensional effects here? Does it enforce the data presentation and makes it clearer or better? If we ask ourselves more questions like these maybe then we can have better decision on what’s necessary or not.
However it’s hard to find the balance between ‘too much’ and ‘too little’. Just like the example Kieran Healy gave us: the minimalist version from Tufte’s own work proved to be the most cognitively difficult for viewers to interpret. We can’t determine if one design is appropriate or not just by itself, for most circumstances we have to place it into its context. `Monstrous Costs’ by Nigel Holmes might be a little ‘too comic’ in some ways but it would fit into a tabloid than a formal news report and vice versa.
About substantively misrepresented visualizations, different angles of looking at the same dataset would produce various interpretations, such as the choice of which type of graph we are using, what’s plotting along x and y axises etc all would affect the ultimate information. How can we prevent ‘taking a part for the whole’ to happen?