Hacker News new | ask | show | jobs
by m0skit0 2375 days ago
Correlation is not causation.
3 comments

Maybe not, but please don't post unsubstantive comments here.
Is it even possible to prove objective causation in a complex social sciences study, since there can be no control group?
Yeah, by the same logic you can't blame sexism or racism for any differences in outcomes, since there's no control group there either.
It would seem any univariate analysis of social dynamics is by default erroneous
No, it could be a third variable or it could even be that women avoiding STEM somehow increases gender equality.

But regardless of the cause, if there is an inverse correlation then pursuing gender parity in STEM is at best a huge waste of effort and at worst counterproductive for all parties involved.

The graph in the article doesn't look too far removed from a cloud though, so I question the strength of the correlation.