Partial correlation, or conditional correlation, is a way to test for correlation between two variables, X and Y, given a third, Z. This technique is often useful in causal research when trying to discover confounding relationships.