Spearman's Rank Correlation

Spearman's rank correlation, developed by the British psychologist C. E. Spearman, measures the linear association between the ranks assigned to items rather than their actual values. It is useful in four situations:

1. When variables cannot be measured but can be ranked — e.g. estimating the correlation between the heights and weights of students in a remote village with no measuring rod or weighing scale; the students can still be ranked by height and by weight.
2. When the quality is inherently a matter of ranking — attributes such as fairness, honesty or beauty cannot be measured like income or weight, but people can be ranked relatively. If at least one variable is of this kind, Spearman's coefficient must be used.
3. When the relation is non-linear but its direction is clear, as in the curved scatters of figures 6.6 and 6.7.
4. When the data contain extreme values — Spearman's coefficient is not affected by extreme values, an advantage over Karl Pearson's coefficient. …