Answer
(D), the correlation coefficient.
Work Step by Step
Going through each choice:
(A): The y-intercept gives an idea of the value of the best-fit line at x = 0, with no indication of fit strength
(B): number of points below the line of best fit tells us which points are overestimated, but not how strong the fit is
(C): Slope represents the relation between y and x, but not the fit strength.
(D): A high absolute value of correlation coefficient means a strong model, r is the square root of $r^2$, which is the coefficient of determination, the percentage of points successfully predicted by the model. So this is a great measure of fit strength.
