7. The high-low method of analyzing mixed costs uses only two observation points: the high and the low points of activity. Are these always the best points for prediction purposes? Why or why not?