A dating website has implied that atheists are more educated or intelligent than religious people. But applying the rationality that is supposed to accompany atheist thought highlights some major caveats.

It’s not often you can release a set of statistics comparing the reading levels of people of different religions and it be among the least contentious sections of a report, but that’s what’s happened with dating site OKcupid. As we’ve covered on a couple of occasions, the people behind the site make a regular habit of digging into the database of user profiles and uncovering data trends.

Their latest effort involved scanning the profile essays written by 526,000 users. Their report on the results concentrates on the statistical distinctions between users of different racial backgrounds. The resulting lists aren’t the words used most often by each group, but rather the words which appeared most for that group in comparison to people as a whole.

If you are currently trying to rid your mind of prejudiced assumptions that white people like golf and hockey, black people like basketball and Kanye West, Latinos like dancing and fighting, and Asians like software development and muay thai, don’t worry: the results suggest you are exactly right. Still, this section was completely worthwhile for OKCupid’s Christian Rudder noting:

” If you’re trying to figure out if white dudes like something, put ‘fucking’ in the middle, and say it out loud. If it sounds totally badass, white dudes probably love it.”

The report also looked at the writing style used by people of different backgrounds and used a readability index to assess how sophisticated the writing was. On the racial side, if your assumption is that Asian and Indian people are on top (for international readers, the US definition of “Asian” refers to the far East) and that black and Latino people are on the bottom, you are again correct.

The site also listed a breakdown by religious background, as shown above.

At first glance this is great news for avid atheists who believe their view shows a greater degree of reason. I’m not going to discuss that belief (I’ll leave that to the comment section), but it is important to note that if reason is the issue, there are some objective points that mean we just can’t be sure how firm a conclusion can be drawn from this study:

* The differences really aren’t that significant: from top to bottom is perhaps a little more than one grade. While its debatable as to what the correct baseline is for reading ages by grades (and reading grades don’t represent a proportional increase), if you start the y axis at “zero”, the differences look less spectacular:

* The reading level at which you write is not necessarily a sign of education or intelligence. While it’s likely that the brighter you are, the more sophisticated or complex the level you are capable of writing will be, there comes a point where some writers will consider it more effective to write in a clearer manner that, while doing a better job of communication, scores lower on automated scales. The results could even indicate that people of certain religions (or non-religions) might be more likely to write in a style designed primarily to make them look more intelligent.

* There’s a flaw with the sample group. While 526,000 users means we can be confident it very accurately reflects the population, that population is users of dating sites rather than the general public. If you want to create a hypothesis to dismiss the results as significant, you could argue that those of a more religious background are more likely to have wanted to get married earlier in life — and the smart ones were snapped up first, leaving the less bright ones looking for a mate alongside clever atheists who find it more efficient to look for love online than at church dances.

Of course, there’s no evidence that any of my suggestions have any validity whatsoever. But anyone who believes in rational thought (regardless of their personal faith) should realize the study doesn’t give anywhere near enough evidence to back Rudder’s comment “Is there a Comic Sans version of the Bible? There really should be.”