The rank correlation coefficient differs from the Pearsonian correlation coefficient in several ways. First, the rank correlation coefficient is derived from the simple correlation coefficient where individual values are replaced by ranks. It measures the linear association between ranks assigned to units, not their actual values. Due to this, all the information concerning the data is not utilised, making it generally less accurate than the ordinary method.

Second, the rank correlation coefficient is not affected by extreme values, whereas the Pearsonian coefficient is sensitive to them. This makes Spearman's coefficient very useful when data contains extreme values. Third, rank correlation can be used in cases where the relation is non-linear but the direction is clear, while Karl Pearson's coefficient should only be used for linear relationships. If the first differences of values were constant, both coefficients would give identical results. Generally, the rank correlation coefficient is less than or equal to the Pearsonian coefficient.