How can you use the standard error of the estimate to assess the usefulness of a new predictor variable in a regression model?

Compare the SEE of the model *with* the new predictor variable to the SEE of the model *without* it. If the SEE is significantly smaller when the new variable is included, it suggests that the variable is useful in improving the model's predictive accuracy. A small change in SEE might indicate the variable is not significantly contributing.