multiple regression analysis a produces measures of probable error b establishes a cause and effect relationship c involves the use of independent variables only d is not a sampling technique