Correlation and causation...
Correlation and causation...
Spotted a few posts with the fact that correlation doesn't imply causation and then did a quick search on the forum for this link without success, so I thought I'd share it.
What's your favourite?
https://www.tylervigen.com/spurious-correlations
How's about:
"Bicyclists killed in collision with two- or three-wheeled motor vehicle" correlates very well with "Deaths caused by X-ray contrast media"...
http://tylervigen.com/view_correlation?id=239
What's your favourite?
https://www.tylervigen.com/spurious-correlations
How's about:
"Bicyclists killed in collision with two- or three-wheeled motor vehicle" correlates very well with "Deaths caused by X-ray contrast media"...
http://tylervigen.com/view_correlation?id=239
Last edited by [XAP]Bob on 26 Nov 2021, 1:41pm, edited 1 time in total.
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.
There are two kinds of people in this world: those can extrapolate from incomplete data.
Re: Correlation and causation...
Very good.
There used to be a famous one about the number of Church of England vicars and something... but I can't remember what. : - (
Jonathan
There used to be a famous one about the number of Church of England vicars and something... but I can't remember what. : - (
Jonathan
-
- Posts: 36778
- Joined: 9 Jan 2007, 2:44pm
Re: Correlation and causation...
Perhaps the skill involves recognising when it's not spurious eg when both are related and share a common cause. The value of that skill lies in then being able to try to influence the effects.
Re: Correlation and causation...
The famous example being the suggestion that yellow teeth caused lung cancer.
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.
There are two kinds of people in this world: those can extrapolate from incomplete data.
-
- Posts: 105
- Joined: 19 Mar 2013, 8:26pm
Re: Correlation and causation...
In a past life, I taught this stuff. A favourite session to invoke the causation conversation was to correlate sales of chewing gum with number of teenage pregnancies....this was in the early 80's
Strong positive correlation, and obviously pregnancy causes one to chew gum. Or vice versa
Strong positive correlation, and obviously pregnancy causes one to chew gum. Or vice versa
Re: Correlation and causation...
It reminds me of an A level General Studies question about the geographic correlation between the incidence of potato blight and neural tube defects, i.e. spina bifida etc.
-
- Posts: 1666
- Joined: 17 Jan 2011, 1:07pm
Re: Correlation and causation...
There was (apocryphally) a strong positive correlation between the wages of Scottish Presbyterian Church Ministers and the number of registered prostitutes in Rio-de-Janeiro.
Re: Correlation and causation...
Our standard example at uni used to be the strong negative correlation between annual UK alcohol consumption and infant mortality: drink more to save babies!
MJR, mostly pedalling 3-speed roadsters. KL+West Norfolk BUG incl social easy rides http://www.klwnbug.co.uk
All the above is CC-By-SA and no other implied copyright license to Cycle magazine.
All the above is CC-By-SA and no other implied copyright license to Cycle magazine.
-
- Posts: 7898
- Joined: 7 Mar 2009, 3:31pm
Re: Correlation and causation...
I notice that two of the three killed by trees in Storm Arwen were in cars.
A police car at one of the scenes was also crushed.
A police car at one of the scenes was also crushed.
It's the same the whole world over
It's the poor what gets the blame
It's the rich what gets the pleasure
Isn't it a blooming shame?
It's the poor what gets the blame
It's the rich what gets the pleasure
Isn't it a blooming shame?
Re: Correlation and causation...
That doesn't surprise me all that much - I could reasonably expect a causative explanation for that.
- Heavy storm, so people not likely to be out walking
- Therefore people are generally either in a car or in a house
- Houses are better protection, and much less likely to be under a big tree
Out of interest - where was that third person?
- Heavy storm, so people not likely to be out walking
- Therefore people are generally either in a car or in a house
- Houses are better protection, and much less likely to be under a big tree
Out of interest - where was that third person?
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.
There are two kinds of people in this world: those can extrapolate from incomplete data.
-
- Posts: 7898
- Joined: 7 Mar 2009, 3:31pm
Re: Correlation and causation...
Yes, indeed. That had occurred to me![XAP]Bob wrote: ↑2 Dec 2021, 3:39pm That doesn't surprise me all that much - I could reasonably expect a causative explanation for that.
- Heavy storm, so people not likely to be out walking
- Therefore people are generally either in a car or in a house
- Houses are better protection, and much less likely to be under a big tree
Out of interest - where was that third person?
A headteacher in Northern Ireland died after a tree fell on his car, another man was hit by a falling tree in Cumbria, and a third died after his car was hit in Aberdeenshire.
Last edited by Mike Sales on 2 Dec 2021, 3:46pm, edited 1 time in total.
It's the same the whole world over
It's the poor what gets the blame
It's the rich what gets the pleasure
Isn't it a blooming shame?
It's the poor what gets the blame
It's the rich what gets the pleasure
Isn't it a blooming shame?
Re: Correlation and causation...
Cumbria - well that strictly answers the question I asked
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.
There are two kinds of people in this world: those can extrapolate from incomplete data.
-
- Posts: 4659
- Joined: 2 Aug 2015, 4:40pm
Re: Correlation and causation...
Does that mean if we cut down all trees (as sometimes suggested on here) that storm fatalities could be reduced to almost zero?
Re: Correlation and causation...
Yep.
Trees are dangerous things in the wind. I won't go into our woods when it's windy.
I read that due to Ash Dieback, they are giving away trees for free.
All very laudable, but where are they planting them?
https://www.bbc.co.uk/news/uk-england-devon-59503256
Trees are dangerous things in the wind. I won't go into our woods when it's windy.
I read that due to Ash Dieback, they are giving away trees for free.
All very laudable, but where are they planting them?
https://www.bbc.co.uk/news/uk-england-devon-59503256
Mick F. Cornwall
Re: Correlation and causation...
Saw this on twitter recently
https://twitter.com/yoguruso/status/146 ... ANYBA&s=19
https://twitter.com/yoguruso/status/146 ... ANYBA&s=19
Former member of the Cult of the Polystyrene Head Carbuncle.