Regression

A statistical application to correlated values that involves building an equation or model that describes the nature of the correlation or relationship. It will typically look like "Y-Response = Constant + Coefficient(X-Predictor)". A correlation measures the linear association between variables as represented by r, but the quality of the regression is measured by squaring the correlation coefficient, or r². It measures the proportion of variation explained by the regression model.
Related StatStuff Video: 
   -Section 6-Analyze Phase, Lesson #26 - Hypothesis Testing: Relationships (Overview)