[lbo-talk] OT: Need Intellectual Refresher

Doug Henwood dhenwood at panix.com
Fri May 13 07:11:25 PDT 2005


BklynMagus wrote:


>When I posted a while back about causes of
>being queer, I was warned not to confuse
>causes with correlations (I think). Can
>someone refresh my mind and set mine straight
>(metaphorically that is) about this issue.

The standard phrase you learn early in statistics is "correlation is not the same as causation." That is, if A and B are correlated, you don't know whether it's A causing B, or B causing A, or some unknown C is causing them both. There are fancier statistical tests than correlation that purport to tell whether A->B or B->A.

On psychosocial stuff, some people think that if you find some neurochemical phenom associated with behavior or affect, you've identified the underlying cause. But it could be that experience affects the neurochemistry, and it's really an effect.

Doug



More information about the lbo-talk mailing list