When is the Spearman rank correlation coefficient preferred over the Pearson correlation coefficient?

The Spearman rank correlation coefficient is preferred over the Pearson correlation coefficient in the following situations:

1. Non-Linear Relationships: Spearman's correlation is based on ranks and therefore it is a non-parametric correlation measure. It can capture non-linear relationships between variables, while Pearson's correlation can only capture linear relationships.

2. Outliers: Spearman's correlation is less sensitive to outliers. Because it's based on ranks, extreme values have less influence on the Spearman correlation than on the Pearson correlation.

3. Ordinal Data: When the data is ordinal or when the assumptions of normality are not met, Spearman correlation is preferred. Ordinal data is a type of categorical data with an order (like movie ratings from 1-5). Pearson correlation assumes that the data is interval or ratio (like height or weight measurements).

4. Monotonic Relationships: Spearman correlation is used when the relationship between the variables is monotonic, which means the variables tend to change together, but not necessarily at a constant rate. The Pearson correlation can be used when the relationship is linear, meaning it remains constant.

In summary, the Spearman rank correlation coefficient is preferred when dealing with non-linear relationships, outliers, ordinal data, and monotonic relationships.