Pie charts
Pie charts can make data really difficult to interpret, particularly when 2 sets of data need to be compared, when shown as a 3D pie, or when too many figures need to be shown.
This image below is a good example of when not to use pie charts. It needs a key, and is difficult to compare the data from both years.
The data relates to the number of programs produced by each TV network over 2 years, and whether they were male or female, white or non-white.
Style and tone
Questions to answer before designing:
1 - What is the information trying to show? (factual)
Highlight the vast differences between men and women, white and non-white directors for the US TV networks.
2 - What is it trying to say? (emotive)
That these differences are wrong, there should be much more of a balance. It is trying to make people agree with the author's point of view and make them feel angry or appalled.
3 - Is there a hook?
I decided to use the heading "Racial and gender inequality in US TV networks". Words highlighting the emotive subject of gender and race will automatically draw attention.
4 - How else can you visualise the data?
I wanted to make the actual figures virtually unnecessary, and make it easier to compare the figures at a glance.
Choosing colours to represent white and non-white people is quite emotive in itself. I didn't want to choose white and black, as not everyone who is non-white is black. I thought about using grey for non-white, however this could infer that non-white people are grey/dull in personality, which I wanted to avoid. I therefore decided to use shades of grey for both categories. I wanted the inequality of both sex and ethnicity to be the defining feature of the infographic, so used flat bright colours with lots of contrast.
The typeface choice was easy - Helvetica. I selected this ubiquitous font as it did not want anything to distract from the shocking data.
I wanted to represent the data with icons, so created them in Illustrator with shape, shape builder and pathfinder.
I created the first set of icons with the height being mathematically proportional to the figures, by creating the icons and changing the height in pixels to the number of directors. When all were created, I then selected them all and scaled them to size as one, thus maintaining their proportions.
When I began creating the next section, the absolute numbers were smaller. I could have used percentages, but that would have distorted the figures, making the small networks appears as it they had the same numbers of directors as the smaller ones, which I wanted to avoid. I therefore decided to represent the directors with an icon per director.
For the final section, I used area so used the square root of the number of directors, creating blocks for each one, before scaling them all together, to retain proportionality.
Final complete infographic is shown below.