[awesome_dude]

Personally I think that reporting of correlations should be dropped altogether - they're very prevalent (reports on correlations) in every day news, and I think that they're damaging because they imply, or outright claim, that the discovery is that some causative effect has been observed.

It's really clickbait territory sometimes (IMO)

[kashunstva]

> Personally I think that reporting of correlations should be dropped altogether…

It really should be a shared responsibility to report and understand the meaning of statistically significant correlation. Unfortunately, few journalists seem to have much interest in understanding it. And given that their readership likely has about average (i.e poor) numeracy and iffy understanding of probability, it’s a bad combination. The widespread misunderstanding about the iterative way that science converges on truth also contributes to this problem.

That said, I would rather know about interesting findings such as this if for no other reason than to start digging for the original paper.

[bigfudge]

I mostly agree. Although for many years the only evidence for harm from smoking was correlational.

[awesome_dude]

As far as I know, it still is. I'm happy to be shown otherwise.

[jeltz]

A lot of modern medicine is based on just correlations and an more or less educated guess. I do not think ignoring them makes sense, it is just that the reporting needs to be more clear and less sensationalist.

[awesome_dude]

That is really because we don't actually understand the human body, and how it reacts to various stimuli.

We used to think stress caused ulcers, based on a correlation. We now know the actual cause is a bacterium.

[jeltz]

Yes, sometimes we are right, sometimes we are wrong.

[awesome_dude]

Which is more damaging to society?

Having things that are wrong, but we don't know any better, or having things that are right, but we don't have the skills to prove it?

We often laugh how people in times before thought "crazy" things about health maintenance, but we're no better.