The coefficient of correlation r(X , Y) between two variables X and Y, due to Kearl Pearson defined by
r(X , Y) = Covariance (X , Y)/{SD(X).SD(Y)} ,
measures the degree of linearity of both way dependence of X and Y (i.e. the degree of linear correlation between X and Y).
(Here, SD means standard deviation.)
Now, Covariance (X , Y) is not based on the deviations between all pairs of observations on each of the two variables X and Y.
Therefore, logically it can not be regarded as the proper measure of co-variation of X and Y..
Similar is the case of SD(X) and of SD(Y).
Thus it is point to think of whether he coefficient of correlation r(X , Y) between X and Y due to Kearl Pearson, as defined above, can give the actual value of the degree of linearity of both way dependence (i.e. correlation) of X and Y.
Thus one question is
" Can the coefficient of correlation defined by Kearl Pearson be a proper measure of the degree of linear correlation between two variables ? "