Nothing like a good piece of confirmation bias.
Do you remember that paper from a few years back that reported religious children were meaner than their secular peers? It was shared by outlets including the Los Angeles Times, Slate, and the Guardian, none of whom, at the time of this writing, have issued updates to their breathless recitations of the conclusions of this now-debunked paper.
Slate concluded, piously, “Perhaps the thing to take away from this is that studies that tout the moral authority of a certain community should be taken with a grain of salt. Or, more biblically: Judge not, lest ye be judged.”
It was clear all along that something was not right with the paper, published in 2015 by Current Biology. First, it measured altruism through subjective metrics such as how mean children thought actions such as pushing were and how harshly they’d punish the perpetrators. Another one of the tests measured how many stickers they were willing to share with others. On average, nonreligious children gave away four stickers, while religious children only gave away a mere — three.
Some writers expressed an initial skepticism about the findings, but it didn’t take long before they were debunked entirely. Azim Shariff, a social psychologist who studies the intersection of morality and religion, found the results particularly odd. They didn’t match up to the findings he’d discovered in his own work. His study had indicated that religious participation typically increased generosity.
Upon looking more closely at the data, Shariff found something wrong. Due to a coding error, the paper’s authors factored in country of origin in a totally nonsensical way. Once that coding error was fixed, the correlation disappeared.
This all happened in 2016, but the journal that published the paper didn’t formally retract it until just last month. Now that the paper has a giant red “RETRACTED” stamped on it, hopefully it won’t continue to be cited as fact, as it was in another article in August.
Some news outlets, though, still haven’t changed their reporting. The science journal made the (very belated) right call to formally retract the article, but it’s also important that those who disseminated it correct the widely shared and false findings. The outlets that published the results of the paper ought to update their articles, even if that’s painful for their confirmation bias.
