I refuse to believe that a moderately clever person makes a 3D chart to get better insights from the data. The real reasons behind this behavior have nothing to do with data visualization, and impression management is one of them.

Maybe this can be confirmed by research on basic needs, like eating. Wired published this interesting article discussing Why Do People Eat Tool Much? The researchers argue that:

The act of choosing a specific size within a set of hierarchically arranged [food] options is one avenue by which individuals signal to others their relative rank in a social hierarchy. As a consequence, larger options would be selected by consumers, not merely out of a functional need for hunger but due to a desire to signal status.

The article author concludes that:

Needless to say, this paper captures a tragic dynamic behind overeating. It appears that one of the factors causing us to consume too much food is a lack of social status, as we try to elevate ourselves by supersizing meals. Unfortunately, this only leads to rampant weight gain which, as the researchers note, “jeopardizes future rank through the accompanying stigma of being overweight.” In other words, it’s a sad feedback loop of obesity, a downward spiral of bigger serving sizes that diminish the very status we’re trying to increase.

Source: Wiki Commons

If you discuss obesity as a consequence of overeating you must add “what people eat” to the equation, and not only “how much”; unlike 100 years ago, fat today is a sign of poverty. The fat cat is the poor cat. We must update the cartoons. But I digress.

So this is an oversimplification, but can we use the same model to partially explain why people make bad charts? Let’s rewrite the sentences above:

The act of choosing a specific chart format within a set of hierarchically arranged chart formatting options is one avenue by which individuals signal to others their relative rank in a social hierarchy. As a consequence, bad defaults and formatting options would be selected by the chart maker, not merely out of a functional need for making sense of data but due to a desire to signal status.

And the second one:

Needless to say, this paper captures a tragic dynamic behind making bad charts. It appears that one of the factors causing us to make bad charts is a lack of social status, as we try to elevate ourselves by over-formatting charts. Unfortunately, this only leads to rampant ignorance about the data and the reality behind it, which, as the researchers note, “jeopardizes future rank through the accompanying stigma of not knowing the reality.” In other words, it’s a sad feedback loop of ignorance, a downward spiral of bad formatting choices that diminish the very status we’re trying to increase.

Hey, not bad, not bad at all! I’m getting addicted to this. Let’s try another one. First, from the original article:

When powerless participants were first told that smaller hors d’oeuvres were served at more prestigious events, the psychologists were able to reverse the effect. Although powerless participants initially consumed 30 more calories than others – they tried to compensate for their lack of status by eating more – learning about the prestige of small appetizers led them to eat 25 less calories than those primed with power.

Now the data visualization version:

When powerless participants were first told that cleaner and simpler charts were presented at more prestigious meetings, the psychologists were able to reverse the effect. Although powerless participants initially created more useless charts than others – they tried to compensate for their lack of status by adding more effects and useless formatting options – learning about the prestige of clean and simplified charts led them to make more effective charts than those primed with power.

Again, the roots of bad eating habits and bad chart making are more complex than this, but if there is at least a partial overlap between a “desire to signal status” and impression management this could help us finding better strategies to deal with bad charts.

So, is a 3D pie chart a double burger in disguise? What to you think?