Hi all,
First, full disclosure - I only read the first half of The Visual Display of Quantitative Information by Edward Tufte. Trust me, it was more than enough. All this book did was describe ways to turn data sets into visually informative representations. The tools involved - graphs, tables, charts - are all old hat for any business bro. I recommend this book to anyone who consistently works with these tools in their day-to-day job function, though, and might suggest it as a refresher for those seeking a quick review of the basics.
What I'm here to do today is share a couple nuggets of information. No thanks needed, reader, I'm happy to oblige.
One common question a report designer will encounter is whether to use a simple table layout or opt for a more visually compelling representation. A good rule of thumb is presented in this work - a table is best for a small data set while a pictorial representation works better as the size of the data set increases. This acknowledges how most people are unable to comprehend differences in large numbers with the same skill they employ in analyzing small figures - therefore, use numbers for small totals and pretty little pictures for larger sums.
There is also a useful note of warning targeted at those involved in the design step - beware of showing design variation, not data variation. A simple example highlights this point. Imagine that you wish to represent the doubling of some annual total - perhaps the color preference for cars purchased at a certain dealership. You might think it is a good idea to use images of different colored cars to represent changes year on year - this year's totals in blue, last year's in green, and so on.
Design this carefully, however. A car one inch tall and one inch wide intuitively 'doubles' to a car two inches tall and two inches wide. But the area of the car on the graph will quadruple!
(One inch wide) x (one inch tall) = 1 square inchThis design feature might mislead a viewer to conclude that the popularity of a car color increased four-fold. To avoid such misrepresentation, you may opt for a bar graph with a fixed width.
(Two inches wide) x (two inches tall) = 4 square inches
Until next time,
The Business Bro
Footnotes / for those who carry Monopoly money around
0. Hint: just make sure the axis doesn't increase proportionally...
Of course, reader, you may wish to mislead your reader! It's a dog eat dog world out there, I've heard. So perhaps you want to know all of these design tricks to help your Machiavellian cause! Well, in this case, give this book a full read, deceptive reader, and don't forget to send your favorite business bro a little tip - in cash, of course - for this helpful recommendation.
