Tufte emphasizes the various types of visual display and the functioning of each visual display. Tufte argues that the purpose of data graphic is to draw viewer’s attention on data and meaning of data. This claim is straightforward and points out that one principle of data graphic is to be precise and concise. Then Tufte introduces data-ink, a un-erasable core of a graphic, including dots, lines and labels. Then Tufte moves to introduce data-ink related concepts, and the first one is maximizing the share of data-ink. That is to say, maximize the data-ink ratio, within reason. The second data-ink related concept is two erasing principles, that is, erase non-data-ink, within reason, and erase redundant data-ink, within reason. The two erasing principles is not hard to understand. Since data-ink plays an important role in drawing viewer’s attention on data, and in order to reinforce the function of data graphic, we would like to decrease the number of non-data-ink and to eliminate redundant data-ink. The next part of the article is about applications of the principles in editing and redesign. Personally, I would like to say that this part may be even more important than theory part. Because we can represent one group data via several data graphics, and the most basic but also the most important ability we should have is the ability to determine the efficiency of each data graphic and use the most efficient one. Tufte offers a bunch of applications to show the importance of previous principles. We can refer to the example on page 15 and 16. We can easily get the conclusion that the third graph is more readable, more accessible and more friendly than the first two, since the third one incorporates more guidance, erase non-data-ink and erase redundant data-ink. The left part of the article talks about vibrations, grids and ducks, which are all considered as chart-junk.
The picture I used here is a dot chart with guidance. From this chart, we can easily read the tendency of those messy dots and we can draw our conclusion upon the guide line. But, if there is no guidance going through those messy dots, the readability and accessibility of the dot chart decreases dramatically. This chart follows the principles proposed by Tufte, including maximizing data-ink, erasing non-data-ink and erasing redundant data-ink. And this chart offers guidance to make it more readable.
My question: Is there any other principle to increase the readability and accessibility of a over complicated chart? Is instruction or explanation helpful?