The two dead comments below are right. The premise of the article is that people who scored higher on IQ tests tended to be more likely to support free speech. That’s got to be the laziest argument I’ve ever seen.
It wasn't even an IQ test. According to the paper, they gave participants 10 words, with each word being presented with 5 other words. They were instructed to match each of the 10 target words with the closest of the 5 words. The claim is that vocabulary knowledge is highly correlated with general intelligence, and so this simple test is a proxy for cognitive ability.