| (1) | \rho_S(x,y) = \rho_p(rg(x),rg(y)) = 1 - \frac{6}{n(n^2-1)} \sum_i \big( rg(x_i)-rg(y_i) \big)^2 |
where rg(x_i) and rg(x_i) are the ranks of elements x_i and y_i.
The Spearmen correlation coefficient is showing the accordance in data set variations: whether they grow and decline accordingly with no regard of the actual values of the variations.
See also
Formal science / Mathematics / Statistics / Statistical correlation
[ Statistical correlation metrics @ review ] [ Pearson correlation ] [ Kendall correlation ] [ Fehner correlation ]