Correlation and Covariance -
The Difference Between the Two
A common question in statistics is what is the
difference between correlation (the Spearman correlation) and covariance
analysis.
First, let's discuss the commonality between correlation and covariance.
Correlation and Covariance both describe relationships between 2 variables.
The Spearman Correlation is considered to be a standardized form of
covariance.
Here are the differences between the Spearman Correlation and Covariance
Analysis:
Correlation values fall within the range of -1 to +1. Covariance values can
be outside of that range.
In the covariance calculation, covariance values depend on the units of
measure of X and Y. Covariance values of data sets using different sets of
measure are not comparable. The Spearman correlation coefficient is not
influenced by the units of measure when calculating correlation. The
Spearman correlation can be used to compare the similarities of multiple
data sets that use different units of measure or scale. The Spearman
correlation solves the units-of-measure problem by normalizing the
covariance to the product of the standard deviations of all variables being
compared. The dimensions or units of measure are then taken out of the
equation.
The Spearman correlation tells you how how close or far two variables are
from being independent from each other. It must be remembered that high
Spearman correlation interpretation does not imply causality between the two
variables. The covariance calculation and covariance analysis tells you how much two variables tend to change
together.
Copyright 2013