Icons, small multiples, and color can convey a deep dive of information at a glance. They can also hide info in plain view.
Final segment of the Women in Numbers infographic review. I’d like to reiterate this review is the result of an attempt to defend infographics, not an attempt to attack them. This review is not meant to illustrate how lame Fast Company is. I don’t think Fast Company is lame.
There are either two or three parts to this section in the lower right — there are two sets of icons and a group of circles with numbers inside them. The circles seem to be related to the icons with flags above them — there are 20 icons and one of the numbers is 20 — but it is not clear that the circles are related to either set of icons. Upon further inspection, it seems that the circles are related to the icons above them, but not the icons to the left of them.
But let’s get to the bigger problems and solutions.
Problem: Icons of women in power lacks context. There are 20 icons presented, each with a flag. Presumably the flag informs us as to which country the icon represents. Unfortunately, most of us do not know what the flag for Kyrgyzstan or even Slovakia look like, so the authors added the name of each country. Adding the name is much more informative and eliminates the need for the flags. But the flags didn’t go away, they remain. In fact, the flags cloud the information trying to be delivered — they are the same size as the person icon, so they draw our eyes to them where we wonder why they are so prominent.. Really, the flags are chart junk.
Moreover, the icons are colored differently — some colors much warmer and more saturated than others. The warm ones pop, and the cool ones fade. Is Argentina less important than Switzerland?
Taking a slightly wider view, the three columns of unequal amounts makes this feel like a column chart — it seems to imply that the third column is less important. And the ordering seems haphazard — why are Thailand and Trinidad and Tobago at the top of the tallest columns? Are they the best?
Solution: A map with a list may be more effective and may also tell a broader story. It may illustrate that there are clusters of women-led countries, clusters of female exclusion, or that women in power is widespread. Color could represent continents, which could contextualize some information quickly. If the icons must be used, make them all the same color and value, put them in four columns of five, and arrange them alphabetically.
Problem: Numbers in circles are confusing. Rather than the size of the circle being used to illustrate relative quantity, they seem to do the opposite — sort of. The biggest number (196) has the smallest circle, and the two other numbers, 33 and 20, have the same sized circle. The two smaller numbers differ by 35%, and the largest of them represents only about 17% of the biggest number.
Solution: Use charts to show relative share of women-led countries now and until now. Try a single pie chart with 33 and 20 superimposed to make the point — current is a subset of all time, so this could tell the story compactly.
Better yet, explore what a line chart could do. I suspect that the ‘all-time’ number could be more interesting (and informative) to look over time. Are there any interesting spikes? is is a steady rise, or an exponential increase? is it going up or down?
Problem: “45%: Swedish Government Officials Who Are Women” lacks context— Does Sweden have the highest percentage of government officials in the world? If so, who has the lowest? Is Sweden an outlier — do they have more than twice as many as the second most country — or are they barely leading the pack? What is the median number, which countries represent the median, and who has the lowest percentage of women in power?
The worst part is that 9 icons out of 19 icons equals 47%, not 45%. The ratio is inaccurate. When your little stuff is wrong, no one believes your big stuff.
Solutions: A bar chart and a short list. Illustrate high, low, and median with three bars. This would show the interesting part — distribution. Also, I suspect that the top five and bottom five countries would be an interesting subset of data to think about. Are the bottom five and top five clustered? Do the clusters illustrate female-friendly or female-hostile continents?
The best solution to inaccurate numbers (47% does not equal 45%) is to use accurate numbers. If 9 out of 19 does not yield the percentage that you need, do not use 19 icons to illustrate your point.
Other parts of this review: