Skepticism, Science & Pseudoscience

Warpy

(113,130 posts) Thu May 15, 2014, 06:31 PM May 2014

Correlation does not equal causation

I found a handy dandy list to bookmark to reference when antivaxers rear their heads. Example, the correlation between cheese consumption and dying by getting tangled in bedding:

They're all good, only I thought the margarine/divorce one might hae something to it. Margarine is nasty stuff.

http://www.tylervigen.com/

5 replies

= new reply since forum marked as read

Highlight:

Correlation does not equal causation (Original Post) Warpy May 2014 OP

Rachel Maddow a couple days ago showed some graphs like this. Archae May 2014 #1

I honestly had no idea it was possible to die like that. trotsky May 2014 #2

It doesn't take much for infants and frail elderly Warpy May 2014 #3

This is a flawed chart edhopper May 2014 #4

Correlation and Causation - Quack Watch EnvironmentalVK Jun 2014 #5

Archae

(46,798 posts)

1. Rachel Maddow a couple days ago showed some graphs like this.

Reply to Warpy (Original post)

Thu May 15, 2014, 07:30 PM

May 2014

Some were kind of funny.

trotsky

(49,533 posts)

2. I honestly had no idea it was possible to die like that.

Reply to Warpy (Original post)

Thu May 15, 2014, 08:38 PM

May 2014

Weird.

Warpy

(113,130 posts)

3. It doesn't take much for infants and frail elderly

Reply to trotsky (Reply #2)

Thu May 15, 2014, 10:42 PM

May 2014

Scary.

edhopper

(34,775 posts)

4. This is a flawed chart

Reply to Warpy (Original post)

Mon May 26, 2014, 08:54 AM

May 2014

Cheese consumption rose by about 9%, while the deaths almost doubled. The two vertical axis seem to be arbitrary and don't correlate in to each other's amounts. To be equal the Cheese axis should go from 29 to 97, not 34. So bogus charts. I also question the rise in deaths, I would guess it's more about reportage than actual incidents.
Too many of their charts do not have the two vertical axis match in percentage.
If they want to make a point about the math being used, they should get their math right in the first place. This kind of mistake negates any argument they make.
There is no true correlation here and it gives a false impression to the real problems of correlation/causation in many studies.

EnvironmentalVK

(6 posts)

5. Correlation and Causation - Quack Watch

Reply to Warpy (Original post)

Thu Jun 19, 2014, 03:18 PM

Jun 2014

The alternative medicine salesman does not want you to know about "correlation causation" and would give you charts.

I suggest to examine Quack Watch Steven Barrett Content.
http://www.quackwatch.com/03HealthPromotion/lowfat.html

Reply to this discussion