a plan.  Representing Industry Information Using Graphs

47 
3. There are sound principles for the display of quantitative infor-
mation. Perhaps the greatest theorist in the quantitative arena is Edward 
Tufte. He has written a number of books and articles on the subject and 
draws huge audiences worldwide when he speaks. Tufte virtually 
founded the field of analytical design—the field that studies how best to 
represent information—especially quantitative information. He has de-
veloped a number of principles over the years.
4. However, in his latest book, Beautiful Evidence, published in 
2006, he organized the principles under six major headings: 
a. Show comparisons, contrasts, differences. Comparisons inform 
and invite reflection by the reader. For example, showing the growth 
rate of platforms using the Mac iOS vs. other smartphone operating 
systems helps show if the market is growing or shrinking. 
b. Show causality, mechanism, explanation, systematic structure. 
Sometimes the data clearly suggest a cause or lack of a cause. For ex-
ample, many predicted a drop in iPhone 4 sales after news that the an-
tenna dropped calls when held in a certain way. But the predicted sales 
drop was not borne out by the data and the product launch was wildly 
successful. 
c. Show multivariate data. That is show more than one or two vari-
ables. The more variables graphed, the greater the chance of providing 
a clear causal explanation. For example, better to show sales of paid 
apps and free apps per platform. 
d. Completely integrate words, numbers, images, diagrams. Tufte 
sometimes calls this “whatever it takes.” Annotate your graph if that 
helps explain the data. For example, on a time series you should label 
key events. 
e. Thoroughly describe the evidence. Provide a detailed title, indi-
cate the authors and sponsors, document the data sources, show com-
plete measurement scales, point out relevant issues. For example, al-
ways show your data sources and list your name as a author. If the data 
only holds under certain conditions, then state what those are. 

48 
f. Analytical presentations ultimately stand or fall depending on 
the quality, relevance, and integrity of their content. This may be the 
most important principle of all. If your content is bad then nothing will 
save it. Try to tell the truth at all costs. 
5. Most of the graph design guidelines come from Edward Tufte’s 
principles of analytical design. These principles require refining the 
graph after Excel has applied its default settings. Though the design of 
a graph is important, the content is even more crucial to the delivery of 
information. Graphs can have good design, but if the data or content is 
flawed, the graph has no purpose. The data on the X and Y axis of a 
graph should always have some correlation, or some relationship that 
can be demonstrated. 

жүктеу/скачать 1.33 Mb.

Достарыңызбен бөлісу: