To Win Facebook Debates, You Need a Terrible Font

With the dumpster fire known as the 2016 election still burning, social media is now just a stream of Facebook debates and Twitter rants. And maybe you’re thinking, “I can just show the clear evidence for my position, make a sound argument, and people will have to agree with it.” But have you met people?

It turns out that the path to making a convincing case might not be clarity, but obscurity. When we make arguments, we often underestimate the power of bias. The easiest path to dismissing contrary evidence is the one easiest to read.

For another project I’m working on, I looked at cases of debiasing or bias disruption. I’m curious about methods used to work around the backfire effect (see, “Why Fact-Checking Fake News Backfires”)—the idea that when faced with contrary evidence, individuals are even more resistant to facts. Human beings are confirmation bias machines, meaning we seek out that which confirms what we’ve already determined to be real. We also actively engage in disconfirmation bias, or the dismissal of evidence that does not help our case.

These deeply embedded biases make any sound argument or even solid fact difficult to accept. It is possible to disrupt bias, but it is not easy to do. It requires forcing the human brain to slow down and actually read the words in front of it.

Researchers at University of Illinois at Urbana-Champaign (summary), for example, looked at the polarization of liberals and conservatives on hot-button issues (see full article, “Disfluency Disrupts the Confirmation Bias“). “Changing beliefs takes time and effort,” they write, “and it is often easier to disregard alternative perspectives rather than to adapt existing beliefs.” They point to a study in which high school students whose class materials were in a difficult to read font (Comic Sans italicized) “scored higher on their examinations than when the material was presented” in an easy-to-read font.

This set the stage for their own study.

They decided to make it more difficult to get individuals to disregard alternative perspectives using a difficult-to-read font, a process known as disfluency. A difficult font forced a more “analytical and critical mindset” and even reflection. In this case, two fonts were used: a “light gray bold and italicized Haettenschweiler font, used in previous research to induce processing difficulty” and a Times Roman (12 point) “photocopied recursively three times on the lowest contrast setting until the text was significantly degraded, but still readable.”

Researchers tested their position in two studies. In the first study, they discovered that liberals and conservatives who read a political argument (on capital punishment) in a difficult font tended to become more moderate afterwards—as the disfluent font forced them to consider the counter argument, while those who read a political position in an easy font remained polarized.

In a second study, the same researchers formed a mock court. They gave individuals information that portrayed the defendant in either a positive or critical way. This was followed by ambiguous evidence. Those who read the critical version were more likely to convict than those who read the positive account. But when they introduced a difficult to read summary of evidence, individuals began to change their perspective. They concluded that:

“participants with prior attitudes on an issue became less extreme after reading an argument on the issues in a disfluent format. The change occurred for both naturally occurring attitudes (i.e. political ideology) and experimentally assigned attitudes (i.e. positivity toward a court defendant). Importantly, disfluency did not reduce confirmation biases when participants were under cognitive load, suggesting that cognitive resources are necessary to overcome these biases. Overall, these results suggest that changing the style of an argument’s presentation can lead to attitude change by promoting more comprehensive consideration of opposing views.”

In other words, to overcome our biases we need the conditions forcing us to actually read the counter-argument. Facts and sound arguments might work better on social media when they are presented disfluently. I mean, who wouldn’t want to read a Facebook post in Comic Sans or that was printed, photocopied multiple times and then posted?

Problem solved. You’re welcome, America.

Photo: Marco Djallo CC0