Curtis Reads the News

Jul 17, 2006 22:48

I'm starting to think I should make this a regular column. I could call it, "Correlation is Not Causation ( Read more... )

Leave a comment

Comments 3

silverhawk July 18 2006, 05:46:21 UTC
While correlation may not be causation, I do remember the family meals I had growing up, and remember them fondly. I also plan to have real meals with my children (if I ever have children). Forget this having to be eight or nine to join the adults. As soon as they can sit, they sit with us.

Reply

iainuki July 19 2006, 21:52:08 UTC
I'm certainly not criticizing family meals. I enjoy eating with my family. (Well, most of the time.)

However, I strongly suspect the arrow of causation runs a different direction: stable, happy families are more likely to eat together, simply because they like each other, and the children in said families are less likely to be self-destructive for more or less the same reasons. However, that's not just speculation, and I'm not stating it as fact. To do otherwise, as they did in the article, is irresponsible journalism, and irresponsible science.

Reply


sithjawa July 18 2006, 19:12:41 UTC
Sympathetic magic. If desirable thing A is often found together with thing B, people will try to achieve thing A by achieving thing B. I think this is probably a tendency acquired by the correlation between correlation and causation. If you can't figure out the cause of thing A, but you try thing B for all (or many) thing Bs associated with thing A, you might hit on the cause. So if you can't figure out the cause, looking at correlations is favorable.

*is random*

Reply


Leave a comment

Up