Kendall correlation is used to measure the correlation between two ordinal measures. Ordinal measures show the rank of something, e.g. placings in a race.

A Kendall Rank Coefficient close to 1 means that two measures are highly similar. When one measure has a high rank then so does the other one.

A Kendall Rank Coefficient close to -1 means that the two measures are highly dissimilar. When one measure has a high rank then the other has a low rank.

A Kendall Rank Coefficient close to 0 means that there is no correlation between the two measures. The rank in one measure does not relate to the rank in the other measure.

## When To Use Kendall Correlation

Kendall Correlation is useful for calculating the correlation between ordinal (aka ranked) measures. The correlation can be used to figure out if changes in one measure relate to changes in another measure.