What a shame. ‘Time’ magazine used to produce great news infographics less than a decade ago.

Alberto,

It’s interesting, isn’t it, how news graphics have managed to skip the strict editorial standards that have traditionally been applied to news stories. Have news organizations typically lacked the expertise that’s required to edit graphics? Do they typically pay little attention to graphics? Has the demand for infographics grown during a time when news organizations were distracted with other concerns, such as with the effects of the Web on news generation and distribution? What’s your take on how this happened?

Hi Stephen,
I would be in the group that argues that visual interest and engagement are critical for transferring information visually (there’s no-one asking people to read this, they must want to). Your approach is simple and shows all the same metrics in a very clear manner, however, I believe it is significantly less engaging (I’d even say a bit ugly), and still lacks clarity (the numbers on the right are not really that clear, and don’t show the different scales these things operate on) and a story.

I guess in the end, as Adil and yourself pointed out – there could be significant improvements to both the original (for clarity and a degree of simplification) and yours (for engagement and story telling).

Stephen, at the risk of being a little facetious, you have fundamentally misunderstood the role of the original graphic. Its primary role is not to communicate information in a clear and easy to interpret manner. Its primary role is to gain Twitter shares and facebook likes.

Your redesign is accurate and effective but it’s just a graph and who wants to share an Excel chart?

Times’ infographic is really complex and about an important topic, so it must be worth sharing.

I wonder…. what proprtion of “good” infographics are produced by women as compared to men?

Dimitar,

Your question, “How about just accepting…ornamental illustration [that isn’t really meant to effectively inform people about the news?]” assumes that this is the intention of infographics such as this one produced by Time. To begin, I actually don’t think this was Time’s intention. I suspect that the infographic designer who produced this, if asked, would say that he (or she) designed it to tell the story about the ongoing lack of female equality in government. In other words, the fact that it doesn’t tell the story well was not by plan but was due to a lack of skill.

However, for the sake of discussion, let’s say that your assumption is correct: there is a new and intentional category of ornamental illustration that is playing a role in news publications. Why not just accept it? First, because it is neither journalism nor worthwhile. Does a news organization deserve to exist if it cannot get people to read its news stories to be informed? What is the purpose of getting more visits to your news site if people aren’t getting news content as a result of those visits? Visits and clicks in and of themselves don’t justify the existence of a news organization. Second, because it is deceitful. If a news organization is merely trying to boost its “engagement KPIs”, why do this by producing infographics, which are supposed to inform us about news events? Why not just put pictures of cute kittens on the site? Third, you are assuming that ineffective but eye-catching inforgraphics produce greater engagement than well-designed and informative infographics, but there is no evidence that this is true. Your assumption is based on a low opinion of the people who read publications such as Time. People who read news publications usually want to be informed. They don’t want to wade through a series of silly, ill-chosen, and mismatched graphics to get the story. They would prefer an infographic that engages them in the story, not the visual features of its design.

Asking me, “Why not just accept this?” is like asking a high school teacher, why not just allow your students to spend their class time playing with their smartphones? The reason is obvious, because they are there to learn and its my job to teach them.

I think their 64% pie chart is even more confusing — it’s the only metric that does not refer to women/(women+men) but rather (voting women)/(all women), but it’s presented in the same color scheme as all the others.

It’s interesting that the Time infographic doesn’t identify the designer(s) as an normal article would identify the journalist who wrote it. This seems to confirm that infographics are thought of as different from regular news articles and therefore don’t need to adhere to the same guidelines for disclosure, just as this example suggests that they don’t apply the same standards for quality.

Hi everyone,

It would take very long to write my opinions about this discussion. Long story short, I agree with Steve, but I also understand why some people are asking for making this specific infographic (meant to be shared in social media, I guess) a bit more fun. So I recorded a video and made my own redesign. You can see it all here:

http://www.thefunctionalart.com/2015/09/stephen-few-asked-me-what-i-thoughts.html

Best,

Alberto

Great analysis as always Stephen.

Any reason why you avoided Pink as the color for Women, when are stereotypes useful and when are they less so?

I get the “less engaging” criticism all the time when presenting clear concise bar charts, comments like “it looks like Excel, we want something more flashy”.

I share the “Save the Pies for Desert” article regularly, yet still I get 3D pie charts in my inbox from so called Data Professionals.

It is disheartening. However, I am glad I can visit this wonderful blog to once again feel that I have right on my side.

Keep up the good work.

As I was scrolling down the comments I wondered if using something similar to the Supreme Court chart for all of the data would be the best compromise. While I understand (or hope I do) the arguments for both images, the plain chart could perhaps come across as a bit sterile while the ‘picturesque’ graphic can obscure the message but engage people on an emotional level more readily than the plain chart.

Love the image created by Alberto Cairo

—-Alberto, are your “graphical elements” just classic example of Tufte’s “Chart Junk” reducing your data to “ink ratio”?—–

No. As long as the icons don’t interfere with the chart, they aren’t ‘chart junk’, a term that I’ve always disliked.

And I feel compelled to add this: If applied uncritically or taken to the extreme, Tufte’s ‘data-ink-ratio’ rule leads to charts that lack elegance.

Duncan,

I forgot to respond to your question about the use of pink for women. I have used pink and blue in charts to distinguish data about males vs. females many times, which I think is fine for most audiences. The meanings associated with these colors are instantly recognized. On rare occasions objections have been raised about my use of pink being stereotypical and demeaning. I happen to love pink and was one of those boys who, during kindergarten, preferred to play house with the girls, dolls and all, rather than the rough and tumble games that took place in the schoolyard. This probably had something to do with the fact that I was the smallest boy in class. Pink doesn’t carry negative associations for me, so when I use it, I certainly don’t do so with ill intent. So, why did I choose not to use pink in this particular case? Because I didn’t want the discussion about the chart’s merits to take a detour into color stereotypes. Had the original infographic used pink, I would have stuck with it, but I was concerned that a change from red to pink might derail the discussion.

Stephen,
My apologies for not taking more time to articulate my arguments better (life is surprisingly busy!), I admittedly only browse here on occasion, and it was Alberto’s tweet that brought me over yesterday, I posted on a whim but I wasn’t ready for your level of engagement, but congrats on it!.

I’m going to try to respond to a couple of points, but so many arguments to both sides of the issue…

When I said your design lacked clarity perhaps this wasn’t quite the right word. Your chart simplifies the information and presents it plainly, and in a way in which it’s easy to compare across the categories. This simplicity and ease of interpreting the information is to be applauded, and in my science/engineering past, this is exactly what I tried to do (and with insight gained from your blog/comments I might add). My priority was to simplify data and inform in an honest, logical, and informative way.

Since joining the business world, I’ve found people much less willing to look at data, but the need to understand it remains. My challenge in reporting & Viz is that I need to engage my audience first, and inform them second. I feel that given the Times primary purpose is to make money, I’d suggest their priorities are similar – engage first, inform second. As such, design is very important (much more so than science & eng, where more people will look past poor design to the underlying information).

On visiting your site I saw the infographic standing alone (ie. Like I saw it on your site), and I felt that it was intended to engage someone to investigate further, a teaser so to speak. As such, I wanted your chart to engage, explain, and intrigue me. Your chart did not. I think this was mostly due to the design elements, rather than the choice of chart. Alberto’s chart takes things in a direction that I like, and Erik Jacobsen’s take on the chart (to me) is a massive success (it made me happy). Erik takes your bars, changes some design elements, and critically for me, splits them into manageable chunks, while still maintaining the ability to compare across categories.

In Erik’s chart the flow from the “At the federal level” reading left to right, and then the note around the % of US voters that were women, really sets the background for the story of interpreting the %’s that come after. By breaking it into smaller chunks I find that I look at the individual lines, rather than trying to interpret them all as a ‘batch’.

One further improvement I’d make to Eriks chart, is to note that the percentages on the bars (right hand side) themselves are redundant, and I think he could add value by having the percentage women outside on the left (as he does), but putting the absolute numbers in each section of the bar (the numbers you put on the right of your bars, but split, and inside the bars).

From a providing evidence / facts to back-up my opinion, I feel that it’s important to remember that it is only an opinion – it’s my opinion, it’s not true, it’s not based in facts, nor is it right. It’s only an opinion.

I hope I’ve explained my position, but I know there are still lots points we could disagree on and discuss further.

Cheers,
Rowan.

@Erik, I totally agree with your comments. In terms of information transfer, the original does transfer, but not cohesively. Yours takes the best of Stephens and chunks it, so I both took the information in, and had some comparative context. Well done!

Duncan and Stephen,

See page 59 of The Visual Display of Quantitative Information to see why Alberto’s embellishments are not a form of chart junk. Tufte himself said, “Sometimes decoration can help editorialize about the substance of the graphic.” He goes on to say that it shouldn’t distort the data measures.

I discuss this issue in Creating More Effective Graphs starting on page 360. Alberto’s suggestion is consistent with the recommendations I make there.

Rowan,

Thanks for the thoughtful response. Like you, I appreciate the useful embellishments and organization that Alberto and Erik have added. In a sense, a good process for infographic design has been enacted during this discussion. I proposed an appropriate type of chart for communicating the telling the story clearly. Alberto and Erik then added embellishments that complemented and improved that chart without taking anything away from the clarity. This is a good design process for infographics that must not only inform but also grab the reader’s attention. Start with and appropriate type of chart and design it to inform clearly and accurately, then add other features that will incite interest without decreasing the charts ability to inform.

One thing that you said concerned me a bit. You said that, now that you’re in the business world, you feel the need to add embellishments to charts to grab people’s interest. I’ve worked in the world of business for my entire career. I know it well. The notion that data presentation within businesses must be dressed up to get people to look at the data is disturbing. This isn’t a viable business model because the skill and time that’s required to design eye-catching and sometimes beautiful infographics will never be broadly distributed in businesses. If a business has one person who possesses this skill set, it’s fortunate. It’s also shouldn’t be necessary. When business people rely on data to do their jobs, clarity and accuracy should be design goal, not entertainment. I don’t doubt for a second that you’ve felt pressure to dress the data up to get business people to look at it, but giving into this pressure is not a winning game–at least not for the business. This situation reveal two fundamental problems: 1) as I’ve already mentioned, relatively few people are capable of adding visual interest to the design of a chart without compromising its ability to inform, and 2) if business people don’t naturally care about the data that they rely on to do their jobs, they aren’t good employees. The first of these problems leads to horrible design atrocities, such as ridiculous 3-D charts with lighting effects and the second leads to low productivity. Business people who rely on data must value that data intrinsically for the useful information that it provides. Dressing the data up for them will usually not only undermine its ability to inform but will also encourage people to spend their time looking for entertained, which isn’t productive. People who care about the data for the useful knowledge that it provides desire clarity, accuracy, and efficiency of use in data displays. Giving into business people’s desire to be entertained rather than informed does not support the business.

Sorry Alberto (and Stephen) but I don’t like your redesign. Stephen’s was a good start and you improved upon the actual data display. But the pictorial is, in fact, chart junk. I find it definitely detracts from the actual data and makes it harder to find the real story in the data.

But I wish to make a different point. I may be too generous to the TIME authors – but this is self-serving, as I wish to make a point about what I do for a living – teach. Perhaps the TIME author has read a little bit about displays of percentages – Gerd Gigerenzer eloquently describes how probability information is more easily understood if displays that show, for example, 3 figures of people out of 100 will have side effects than if we say the side effect rate is 3%. So, perhaps the TIME author read some of that and thought they were doing something good. This is not to defend their poor display (poor for all the reasons Stephen mentions). This is what teachers are for – to help people really understand and appreciate the context and the big picture. The Gigerenzer improvement is a good idea – when used in the right context. In this particular case, it is counterproductive, unnecessary, and confusing.

And, I don’t think the bar chart (particularly with Alberto’s improvements) is at all boring or fails to engage. In fact, if we were to add a few more categories (% of women on boards of directors in Europe, % of women politicians in Europe, % of women doctors in Saudi Arabia, etc. etc.) the display would be even more compelling. I’d rather add more information than make the limited information more “engaging.”

Dale,

Gigerenzer’s work is wonderful. A few months ago I reviewed his book “Risk Savvy” with extreme favor. His knowledge of data visualization, however, is limited. I hope to someday show how some of his improved information displays could be further improved. For example, his point that many people to whom doctors much communicate understand frequencies (counts of things) better than percentages does not mean that the information must be displayed by showing large numbers of icons that people must count. Information can be expressed as frequencies rather than percentages without forcing people to count individual objects, such as in bar graphs with quantitative scales that show frequencies (e.g., 0 to 100) rather than percentages (e.g., 0% to 100%).

Hi,
I’m an engineer keen on financial presentation, i.e. quarterly earnings (“Show me the numbers” :) ). I’m not an infographics lover at all. Surprisingly, this Time’s infographic is the first I almost* like.

Flow. Think for a while about this infographic alternatively as PP presentation with 10 or so slides and with a presenter in person saying respectively i.e. #1 Did you know, that only 19% of House of Representatives are women? #2 Hard to believe, but only 20% of Senate are women #3 Do you realize that only 12% of Governors are women? #4 Why is that no more than 18% of Mayors of Cities More than 30K People are women? And so on.

One picture/chart – one statement. Ten slides, ten charts. But … in such case the same chart type on each slide would be rather boring and uninteresting. And, right, different chart type on each slide would be weird. I mean this “ribbon” Time’s infographic is like presentation “One picture for one information” model with differentiated chart types. And everyone is capable to manage all these pictures, as we have just two values on each (part to whole relation).

Gathering all that information in one chart leads us to something like this: “Women are misrepresented in Congress, Senate, as Mayors and Governors … . Compare, where are they more misrepresented”. I think Time’s infographic was not indented for comparison. It rather shouts: “Here, in Congress, only … % are women, and here, in Senate there are only 19% women, and only 12% Governors are female…”. It presents each category more separately. It is possible, that half of Time’s readers will stop little longer for each category when presented separately. They may pay more attention.

That’s just my impression.

*off-topic: someone inquiring, reasonable or unfavorable would simply ask, why women prefer to vote for men than for women ;)

Though I don’t want to derail the topic, I’m bothered by something else. The metrics provided don’t really support the underlying proposition: namely, that Women’s equality day is *needed* to bring these lopsided results into balance. That’s not to say these metrics don’t reflect alarming systemic inequality because they very much do. But Women’s Equality day commemorates a turning point in our nation’s history. In its memory we should strive to have more equal representation. But achieving equality itself would not obviate its commemoration. We would “still need Women’s Equality Day” long after true equality has been achieved.

Thought we can draw a straight line between this data and the importance of addressing inequality, they don’t actually address the underlying argument. The designer picked a catchy headline, but didn’t take the time to see if the evidence presented truly makes their case. It’s not enough to simply dump relevant metrics and hope your reader agrees with you. Too often designers believe the presentation of data itself is all that’s required to make a case.

Mark,

Fortunately, as Alberto has shown, infographics needn’t be superficial, even though they often are.

Leila,

Your take on this is common and it seems reasonable in theory, but it breaks down in practice. No matter how much someone likes glitter, that preference won’t enable a decorated visualization to communicate its message. All the glitter can do is attract attention, but once someone’s attention is caught, the chart must then inform. To inform, it must be designed in a way that enables the human brain to extract meaning. The design of the original infographic undermined the reader’s ability to extract meaning. There is no excuse for this. As Alberto so expertly demonstrated, you can design a chart in a way that attracts attention, when that’s necessary, but do so in a way that does not undermine its ability to inform.

People’s preferences (e.g., “I like pretty charts that spin”) do not necessarily reflect what works for their brains. Think of preferences as desires and what actually works for their brains as needs. If your children desire candy rather than a nourishing meal, would you feed them candy? Not likely. Until people’s desires become more aligned with their needs, it is up to us to nudge them toward nourishment. Information is food for the mind. Let’s not feed the public crap simply because they haven’t yet developed an appreciation for good information. Let’s patiently teach them how to appreciate finer thinking.

Your analytical vs. creative dichotomy isn’t real. Analysts must be creative and people who create must be analytical. Analytical and creative are not two ends of a continuum. The world isn’t this simple.

Yep. Surrounded by a circus. News isn’t what it used to be. Ratings being so important, it’s easier to appeal to everyday people by (apparently) sophisticated looking 3D graphics when showing any data. It all started so innocently with more and more glitzy weather report graphics. Now we have seamless transitions between news and finance analysis (not actually part of news), read by a monkey dressed in a suit.

I am only a student in graphic design, and infographic design is something that I am just getting started with. So I apologize if I am less acknowledgeable about this subject than most of you. However, I have been researching this subject a lot, and I believe that I have a unique perspective.

My problem with many infographics is the lack of insight between visuals and chart design. Often, many infographics will be one way or another. The first is the visually heavy infographic (like the original Times infographic). These infographics engage the reader, but they often misrepresent the data or make it more confusing than it needs to be, or there is no data at all and just informational bullet points. A lot of these infographics also give the reader information overload–there are too many images and facts in this tight layout to really give the reader a visual break so they can understand the information. Then, there are the data visualizations (like your fixed version of the Times infographic) where the data is well presented, but there are no explanatory visuals (like illustrations). To me, this shows the data perfectly, cleanly, but does not explain the story behind the information.

The visualization of a story is the biggest division between infographics and data visualizations (for me at least). A data visualization just collects, organizes, and visualizes the data in an informative matter. But an infographic does all of that and more; it tells a story, and gives the reader an experience. I think that if you create the infographic correctly, and with care, you can create a compelling story and experience that explains both the story, and interprets the data correctly.
This is my most recent infographic: https://mir-s3-cdn-cf.behance.net/project_modules/max_1200/2d66a031320471.564b66e8daec0.jpg —- This is just a small example of how I try to stay true to the data, but create a story using visual elements.

However, I completely agree that something needs to be done with infographics, and data visualizations within the news. There is so much misrepresented data, and people are really affected by those facts. Many people do not have the sense to doubt data, which causes a lot of information to be really skewed. This is definitely something that needs to be changed.