Correlation does not equal causation
Sometimes your analysis could be lead you to wrong direction . You could think there is relation between your varibles but its not always what it seems like.
In the above you can see the different types of correlation but is this correlation always right?
Let’s look at some definitions
Correlation: The degree to which two measurements vary together.
Causation:The act of one things being result or effect of another thing.
Coincidence: A striking accurence of two or more events at one time apparently by mere chance.
Spurious Relationship: Correlations due to chance or unmeasured variables but no direct casual relaitonship.
Fortunately when you stick up two different decision or act, there are something that you can do. It’s called work experiments, generally its is useful for testing reliability of your decision or analysis. Maybe you already heard about A/B Test which is great for testing.
You can face with correlation while you are working on huge volume data sets. For example, it can be sales counts. You should provide insight from your data analysis. In this case, correlation is useful to do that.