For the second case, adjusting the scale of the data points will not affect the correlation coefficient. Indeed, suppose all the x’s are scaled larger by a factor of 1000. Then, the mean and the standard deviation of x-values will also be 1000 times larger (convince yourself of this) and the covariance will be 1000 times larger. Indeed, for covariance