On the continuing problems of media effects research