Skip to content

Infographic Dissection

So here I am dissecting another of the infographics that Jay Brooks has posted. I really do like most of them, so I feel a little bad that I have found another that angered me so much I felt the need to call out the designers.

Jay points out that the “standard unit” is problematic all by itself, but that said, one could still come up with some pretty interesting information by using it as a base. For one thing, we could learn how successful the British Ministry of Health has been at getting people to know the standard unit is and to get people to start using it to alter their behavior. The answer to that questions seems to be: “not very successful,” which is probably for the best since, as Jay himself has pointed out, the units are ridiculous.

What I don’t like are the color and icon choices. Although I made the argument last week that women and men really are different in a lot of ways, it simply isn’t so stark that the icon for men should be a beer bottle and the icon for women a wine glass. Wine is delicious and men drink it too! The icon choice reinforces stereotyped gender expectations which is insulting at best.

But what I really dislike here are the color choices. The colors bounce back and forth between a dark red and a bold shade of pink. In some instances the color choice seems to be solely for visual separation as it is in the left hand side of the “average drinking” section, the bar graph—each bar alternates colors and no additional meaning seems to be attached to the color choice. However, given that just to the right the man is pink and the woman is red, it seems that we should carry an additional gender meaning to the bar graph. This confusion is compounded by the fact that wine is red, as is the woman; and the man is pink as is the beer bottle. But in the partial pie charts below, “beer” is red and “wine” is pink. This makes it clear that colors are arbitrary—but why should they be? In the cute little gender graphs under the man and woman icons, the colors are just confusing. Which part is the 31% and which part is the 25%. Because the icons both have a lot of negative space, volume of color is harder to assess. With study it appears I should be concentrating on the pink part. But why am I studying it so hard? The entire purpose of an infographic is to make it easy to get information and turn it into knowledge! I’m not supposed to have to try and figure out what you mean. Colors continue to be confusing as one goes down the chart. In the next section, colors are definitely not …or definitely are related to gender, it is literally impossible to tell. Either women or beer drinkers are drinking over 21 or 35 units at some hard to see rates. Since red has been used to mean “beer drinkers” at one point and “women” in another, it is impossible to determine with certainty. In the section below that, the gender/color relationships are reversed explicitly from what they were before. Gah!

Although this is not related to the two previous critiques, I want to add two critiques of the final two charts under the heading “Pupils at secondary school aged between 11 and 15 were also questioned about their drinking habits.”

First of all, the questions are clearly not mutually exclusive questions of the type where the different responses will total 100%. A single student could answer yes to “drinks once per week,” “drank last week,” AND “have drunk before.” And where’s the “never drinks” and “drinks less than once per week” categories? Whatever. The point is that variations of pie charts are inappropriate for displaying non-mutually exclusive data like this.

Besides, (critique 2) the purpose of this section seems to tell a story and that story is “young people’s drinking habits are getting healthier” (provided you think that young people should be drinking less or not at all, which I do). But because the designer chose a pie chart, that story does not come across–in fact, the opposite is true. For each question 11-15 year-olds showed some improvement: less respondents “drank once a week,” (a decrease of 13%); less respondents “drank last week,” (a decrease of 14%); and less respondents “have drunk before,” (a decrease in a stellar 16%!). The ratio of the first graph is 61:26:20 or approximately 3: 1.3: 1. The ratio of the second chart is 45:12:7 or approximately 6.5: 1.7: 1. In other words between the first and second charts the red (“have drunk before”) appears to have grown, which means that to the rapid observer the story looks much, much, worse. That is bad infographic design.


Share and Enjoy:
  • Print
  • Digg
  • StumbleUpon
  • Facebook
  • Yahoo! Buzz
  • Twitter
  • Google Bookmarks
  • Add to favorites
  • Google Buzz
  • LinkedIn
  • Tumblr

Posted in The Drinking Class.