We need to scale variable values before performing the clustering analysis. This is because variables that have the *most variation* in its values would dominate the distance measure. From output of `summary(DATA_HW6)`, which one variable will dominate the distance measure if we do not scale them?