Gale & Graham's comments are correct but they don't contradict the claims of the study. There's a big difference between median and average. (I do think people are lying though.)
Sure, it's always easy to play the purist-formalist and jump at the neck of the NY Times editors. But in our imperfect world, seldom does a non-constructive purist or formalist opinion move things forward...
What should we choose -- science articles which are not formally conclusive but at least make people think, or let people form an stupid opinion and feeling good about it?
BTW, since we're wondering about tails, here's another excerpt from the study's conclusion: 29 percent of American men report having 15 or more female sexual partners in a lifetime, while only 9 percent of women report having sex with 15 or more men.
The british study, in which men reported an average of 12.7 partners, and women 6.5, presumably -was- dealing in means, not medians, and so the article makes a valid point, although perhaps should not have mentioned the median statistic from the american study.
Note the subtle difference between mean and median...
ReplyDeleteIn more detail:
http://justoneminute.typepad.com/main/2007/08/they-cant-mean-.html
- grad student, addled from grading, whose non-mathy friends pointed this out to him
Gale & Graham's comments are correct but they don't contradict the claims of the study. There's a big difference between median and average. (I do think people are lying though.)
ReplyDeleteSure, it's always easy to play the purist-formalist and jump at the neck of the NY Times editors. But in our imperfect world, seldom does a non-constructive purist or formalist opinion move things forward...
ReplyDeleteWhat should we choose -- science articles which are not formally conclusive but at least make people think, or let people form an stupid opinion and feeling good about it?
BTW, since we're wondering about tails, here's another excerpt from the study's conclusion: 29 percent of American men report having 15 or more female sexual partners in a lifetime, while only 9 percent of women report having sex with 15 or more men.
The british study, in which men reported an average of 12.7 partners, and women 6.5, presumably -was- dealing in means, not medians, and so the article makes a valid point, although perhaps should not have mentioned the median statistic from the american study.
ReplyDelete