looking at the scatterplot of pages versus amazon price, it appears that the data might be clustered around two separate regression lines. what is the cause of this split in the data?