Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Quick Correlation Problem

  1. Aug 24, 2005 #1
    Hello All,

    I am trying to look at some correlation numbers and while I am aware that correlation does not always indicate causality can a correlation coeffiecent that is very near to zero be used to indicate that it is not a cause?

    Thanks

    C
     
  2. jcsd
  3. Aug 24, 2005 #2

    mathman

    User Avatar
    Science Advisor
    Gold Member

    Elementary logic should give you the answer.

    Causality => correlation.

    Therefore: not correlation => not causality.
     
  4. Aug 24, 2005 #3

    EnumaElish

    User Avatar
    Science Advisor
    Homework Helper

    Let's be careful.

    Causality does not always imply correlation because the relationship may be nonlinear. Correlation is a linear relationship.

    If two variables are dependent, their relationship can be linear or nonlinear. If it's linear, their correlation will be significant. If their corr. is not significant, then you can say that the variables are not LINEARLY dependent on each other. But, they can still be non-linearly dependent.

    So the answer is technically "No, a zero correlation does not mean no causality." A more roundabout way of putting it is: "A zero correlation would imply no causality only if a non-linear relationship can be excluded on other grounds."
     
    Last edited: Aug 24, 2005
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Quick Correlation Problem
  1. Cross correlation (Replies: 2)

  2. Two quick problems: (Replies: 0)

  3. Interclass correlation (Replies: 3)

  4. Correlation coefficient (Replies: 15)

Loading...