what are 2 other possible sources of error for this experiment? how would each of these affect the calculated specific heat of the metal?