The Root Mean Squared Error is best described as a measure of what?

Correct. As the name implies, the root mean squared error is a measure of the "average" prediction error for all points in the sample. Average is in quotes because it is in fact, a measure of the variability or dispersion of prediction errors, but it is commonly interpreted as a average.

correct answerThe "average" amount of error expected in predictions

The number of cases that have a non-zero residual

The strength of the correlation between X and Y

The sum of prediction error for all points in the sample

The sum of residuals for all points in the sample