So I contacted Andreas Krebs, the study lead, and got hold of a copy of the original report (it's available online here - in German). You probably won't be surprised to hear that the results don't exactly support the news headlines, but they're all the more interesting for that.
The study basically records the athleticism of first-grade school kids (7-8 years old) in the small Swiss city of Winterthur. The battery of tests included things like running, jumping, stretching and tapping (yeah, beats me too).
They broke down the results in a few different ways - religion is just an incidental part of the report, near the end.
The figure shows the results: boys at the top, girls at the bottom. The separate bars show the results for 4 sequential years.
The non-religious kids are in the middle. To the left are the Muslim kids, to the right are Protestants and then Catholics. The size of the bars shows the z-score - basically how far above or below average that sub-group is, relative to the group as a whole.
So, non-religious boys actually have average athletic ability, while non-religious girls are above average.
Why should that be? Well, I think the clue is in the particularly low scores for the Muslim girls (and other minor religions). These girls are often restricted by their culture from fully participating in sport.
Does the same apply to Christians, but to a lesser extent? If so, then that would explain the results.
It could simply be that religious parents - both Christian and Muslim, have sexist attitudes towards kids and sport. Muslims tend to discourage kids from sport, but especially girls. Christians tend to encourage kids, but especially boys.
Non-religious parents, on the other hand, seem to be not particularly sports-minded, but are completely non-sexist when it comes to encouraging their kids. As a result, their boys have only average ability, but their girls are way above average.
This article by Tom Rees was first published on Epiphenom. It is licensed under Creative Commons.
Maybe it would be more fair to say that nonreligious parents are LESS sexist than religious ones, rather than saying they are completely non-sexist. Since there's no real absolute zero in the data.
ReplyDeleteSo would this explain the tremendous athleticism of high school males in Texas, where football is a religion?
ReplyDeleteFun ! This goes back to the point of headlines that use the word 'religion' whole-cloth are often misleading, but it is a common catchy way to get readers -- even if it is misleading.
ReplyDeleteGood! Someone else (Syphax) got there before me:
ReplyDelete"Non-religious parents, on the other hand, seem to be not particularly sports-minded, but are completely non-sexist when it comes to encouraging their kids" is surely a fairly indefensible statement: there must (alas) be plenty of sexist atheists . . . although, without the corrupting factor of religion, almost certainly fewer of them per capita than among their devout counterparts.
(Hm. Open ID seems not to be working today.)
--JG
Yeah, you're right. Non-religious parents are less sexist, but not completely non-sexist. Well, on average anyway.
ReplyDeleteCris, interesting thought. There are a lot of sociodemographic factors bundled in here too though. The Protestants and atheists tend to come from the wealthiest families.
ReplyDeleteMaybe one should add that the region of Winterthur (near Zurich) is traditionally mainstream protestant. Catholics are either from other parts of Switzerland, or from Italy, but in the latter case, it's the parents or grandparents who immigrated (i.e. the kids are second or third generation immigrants). Non denominational families may be either. These three groups have a more or less similar cultural background, with the notable exception the study indicates.
ReplyDeleteMany muslims, in contrast, are probably relatively fresh immigrants, with much different cultural attitutes. It would be interesting to redo this study in 30 years time and see what happens with the kids of those muslims who grew up in Switzerland!
I just checked the study - they should get a linguist, I understand that they wanted to sort languages according to geographical and cultural aspects, but you can't call Turkish an Arabic language, you have to find a geographic term like Middle East. And "Czechoslowakian languages" wtf?? Can't they just write "Czech, Slowakian"? Or just Eastern European, and lump them together with Polish...?
ReplyDeleteYou infer a lot from a study that is limited to one town in one European country. I don't see any indication of standard deviation, but judging from the difference in performance between grades of similarly religious groups, it looks like it could be so large as to make the difference between non-religious and protestant girls statistically insignificant. This study doesn't make a strong case for half of what you give it credit.
ReplyDelete