One well-known tenet in psychology is that of confirmation bias, which states that people seek out information that confirms what they already know. It certainly explains contemporary American politics and the 2012 presidential election, in which people seek out news channels and websites that fall into line with their already-held beliefs.

That idea of confirmation bias is why it is so difficult to persuade anyone, whether it is voters, consumers, a jury, or your parents, of anything that they do not already think. Now, researchers have found that if you want to convince anyone of something with the use of a proposal or another written document, using funny, frustrating, or otherwise unusual fonts will give you a way in.

The studies were conducted by Ivan Hernandez and Jesse Lee Preston from the University of Illinois, Urbana-Champaign. They found that people's polarized attitudes were likely to melt away if they were forced to read something with a font that was difficult to read. Because they had to slow down in order to comprehend what they were reading, it made it more likely that they would approach the material with an open mind.

The first experiment looked at 133 undergraduate students. The study began by asking participants to fill out questionnaires that identified how politically liberal or conservative they were. Then they had to read a short article in favor of capital punishment. Half of the group had to read it in standard 12-point Times New Roman font (high fluency), and the other half read it in light gray, bold Italicized Haettenschweiler font (low fluency). Previous research had found the font was difficult to process. Then all participants had to answer questions on the article, about how intelligent they found it and how they thought that the facts were presented.

They found that when the issue was presented fluently, liberals were strongly against the article and conservatives were strongly for it. But when the statement was difficult to read, those differences disappeared.

The second study was conducted on a wider group of people. The online panel of 398 people was tasked with reading about a man charged with robbery. Half were given positive traits about him, and were conditioned to be positively biased toward him. The other half were given negative traits, and were biased against him. Then they were given an unbiased account of the case. Half read it in 16-point Times New Roman font; the other half were given a copy that had been written in 12-point Times New Roman font and then photocopied three times, so that it was just barely legible.

When researchers asked the panel to give the man a verdict, those biased against the man and who had read the case in a clear font rushed to declare him guilty. Those who were biased against him and who had read the photocopied version were more on the fence.

But there is an exception: when people were rushed into a decision or needed to rely on memory, they fell onto their biases, no matter what the font looked like.

So, after the election, maybe all of Congress' bills should be written in Mistral. That way, Congress members will be more likely to actually reflect on the issues rather than adhering to their own biases.

The research was published in the Journal of Experimental Social Psychology.