Assumptions of multiple regression: Correcting two misconceptions