News for hackers. Stuff that matters.
Re:Massive 15-year Study finds no link between cell phones, ... by jerf on Friday October 21, 2011 @07:35PM (#3142205)
Lack of correlation isn't "proof" of lack of causation, but it's damned strong evidence. Yes, in theory a precisely calibrated confounding variable could be precisely canceling the real effect, but until you've got some sort of actual evidence of such a factor you're in blatent violation of Occam's Razor.

I'm really starting to loathe the thoughtless chants of "correlations doesn't imply causation"; it used to signal a deep understanding of science, now it tends to signal quite the opposite.

Re:Massive 15-year Study finds no link between cell phones, ... by wtvanhest on Friday October 21, 2011 @09:31PM (#3142441)
No, it is not strong evidence. It is only evidence of a correlation which is meaningless.

You shouldn't loath those thoughtless chants, you should embrace them. The first thing someone should say when presented with "linkage" is think, that is meaningless on its own.

The real problem is that there are so many scientists (people with the ability to conduct statistical analysis) and so much data, that every possible correlation is being discovered. Even correlations that have nothing to do with anything.

The best examples come from Finance, where the data has been more attainable, and where the rewards for finding real causality have been available longer and allowed for advanced computations on super computers before we all had excel (or put more useful statistical tool here).

One of the go to papers on Motley Fool's completely loosing stock selection strategy is here:

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=158409

There are lots of papers and examples in finance where people tried to find causality, but really only found correlation. Doing so is the best way to loose a ton of money.

It is also the best way to get people to do stuff you do/don't want them to do and it is the best way to stop progress.

So, don't make assumptions about causality, you should provide real evidence of causality beyond correlation if you write the word linkage.

Re:Massive 15-year Study finds no link between cell phones, ... by jerf on Saturday October 22, 2011 @06:30PM (#3144480)
Lack of correlation, wtvanfest. Important word! Lack of correlation is evidence of lack of causation. It isn't absolute proof but it is good evidence. After all, "lack of correlation is evidence of lack of causation" is the contrapositive(ish) of "Causation is evidence of correlation". If that's not true then all hope is lost of ever learning anything.

Sorry, but that's exactly the sort of thoughtless repetition I'm talking about. You're in such a hurry to chant the mantra and wave your scienciness around you didn't even take the time to notice it was the precise opposite of the conversation at hand. Which is precisely what my complaint was in the first place.

Re:Massive 15-year Study finds no link between cell phones, ... by gregwebs on Monday October 24, 2011 @05:01PM (#3151552)
Here is an article explaining why lack of correlation does not imply lack of causation: http://blog.cholesterol-and-health.com/2011/01/lack-of-corre...

Ignorance is excusable, but the insulting tone you take on is not. Please stop ruining Hacker News.