Bizarre Correlations

[Guest]

I've happened upon this and it reminds me of how statistics and relationships can be manipulated by anyone wanting to prove almost anything!!

http://www.tylervigen.com/

[ferneyhoughgeliebte]

People actually die "by getting tangled in their bedsheets"??!! How does that happen?

This never occurred to me as a possibility before - now I'll be paranoid about getting into bed! Go on; somebody please tell me this is a spoof

[Beef Oven!]

People actually die "by getting tangled in their bedsheets"??!! How does that happen?

This never occurred to me as a possibility before - now I'll be paranoid about getting into bed! Go on; somebody please tell me this is a spoof

Lol! Exactly what I was thinking, but didn't want to tempt providence and post!

Note to self: Only use a light duvet this summer. Save money, don't buy margarine.

P.S. Great post, Krystal

[Krystal]

It probably is a spoof; I don't know. But it shows that you CAN draw conclusions, if you want to, from any bit of data. On one level it's a bit sobering.

Regarding death by bedsheets; remember, sexual congress was regarded by Shakespeare in his plays as equated with 'dying' (yes, true!).

[Bryn]

It probably is a spoof; I don't know. But it shows that you CAN draw conclusions, if you want to, from any bit of data. On one level it's a bit sobering.

Regarding death by bedsheets; remember, sexual congress was regarded by Shakespeare in his plays as equated with 'dying' (yes, true!).

Ah, what Frenchie(s) call(s) le petit mort.

[Dave2002]

It probably is a spoof; I don't know. But it shows that you CAN draw conclusions, if you want to, from any bit of data. On one level it's a bit sobering.

Regarding death by bedsheets; remember, sexual congress was regarded by Shakespeare in his plays as equated with 'dying' (yes, true!).

No - I don't think it's a spoof. I believe it's an exploitation of the use of big data techniques to spot spurious correlations. There does not have to be any causal relationship.

Re the bedsheet deaths, I can't comment, but I think that Bill Bryson refers to people being hospitalised by problems with (their?) underpants!

[Roehre]

No - I don't think it's a spoof. I believe it's an exploitation of the use of big data techniques to spot spurious correlations. There does not have to be any causal relationship.

For far the best part a casual relationship is unlikely.
There are some however which might show a correlation due to some kind of casual relationship, especially the ones where population growth is one of the data sets used.

Brilliant post, Krystal

[ferneyhoughgeliebte]

Re the bedsheet deaths, I can't comment, but I think that Bill Bryson refers to people being hospitalised by problems with (their?) underpants!

Well, yes; who hasn't?

[Nick Armstrong]

people being hospitalised by problems with (their?) underpants!

Well, yes; who hasn't?

A brief stay, I imagine ferns?

...

Well, not 'imagine'... 'guess', perhaps...

[Pabmusic]

It probably is a spoof; I don't know. But it shows that you CAN draw conclusions, if you want to, from any bit of data. On one level it's a bit sobering...

It's actually highlighting a common misunderstanding caused by not appreciating the difference between correlation and causation. It isn't enough that two phenomena show a correlation, the correlation has to be causal. Take these:

If sales of hot chocolate go up, street crime drops. (Does this correlation mean that hot chocolate prevents crime ? No, it perhaps means that fewer people are on the streets when the weather is cold.)

The bigger a child's shoe size, the better the child's handwriting. (Having big feet makes it easier to write ? No, it probably means the child is older.)

It's this misunderstanding that underpins much superstition - particularly "lucky" talismans and things like buying a lottery ticket between 11.15 an 11.34 on a Wednesday (because that's when you were lucky once before).

There are well-known experiments with birds and animals in Skinner boxes, where the release of food is dependant on pushing the right buttons. The link between buttons and food delivery is then turned off, so that delivery no longer has anything to do with pushing buttons, and you notice superstitious individual behaviours. For instance, one pigeon will turn circles before pushing buttons, another will turn to one corner, etc. It's what they did once before food appeared. They've identified a false correlation between an action and an event.

[richardfinegold]

A brief stay, I imagine ferns?

...

Well, not 'imagine'... 'guess', perhaps...

Those underwear accidents tend to run in streaks...

[MrGongGong]

Some of you might find this interesting

http://books.google.co.uk/books/about/The_Secret_Power_of_Music.html?id=HcZ6QgAACAAJ&redir_esc=y

Confusing correlation and causation is what keeps the homeopaths et al in business

[Serial_Apologist]

Those underwear accidents tend to run in streaks...

But not on streakers, obviously...

[Bryn]

Some of you might find this interesting

http://books.google.co.uk/books/about/The_Secret_Power_of_Music.html?id=HcZ6QgAACAAJ&redir_esc=y

Confusing correlation and causation is what keeps the homeopaths et al in business

And there was I thinking it was human gullibility (wishful thinking) and Royal patronage per se.

[Flosshilde]

For instance, one pigeon will turn circles before pushing buttons, another will turn to one corner, etc. It's what they did once before food appeared. They've identified a false correlation between an action and an event.

It's good to know that my prejudices about pigeons are wrong, and that despite appearances they can be as intelligent as people.