Determines the amount of error we expect when we use a relationship to predict Y scores. It’s known as .

Formula:

\begin{equation} S_{Y’} = (S_Y)(\sqrt{1-r^2}) \end{equation}

where is the standard deviation of all Y scores.