In the Presence of Multicollinearity, the Predicted Values Of $\hat{y}$ Are Actually Quite Good for Values of X Far Outside

True/False

In the presence of multicollinearity, the predicted values of $\hat{y}$ are actually quite good for values of x far outside the range of the sampled values of x.