Truth and myth of independent software testing: a controlled human experiment