Ignoring a Common Cause

Assuming that correlations within data show that one variable causes another, and ignoring a possible underlying variable that is responsible for variables to correlate.