Two studies investigate whether the placement and/or wording (either positively or negatively) of diagnostic rating scale items influenced student responses to the global items in the evaluation of a course of instruction. The Instructor and Course Evaluation System (ICES) developed at the University of Illinois, Urbana-Champaign was used to conduct end-of-semester course evaluations. Thirty diagnostic items were selected from a catalog containing approximately 500 items. Half of the global items were about the course and half were about the instructor. Twenty of the 30 items were rewritten to create positively and negatively worded versions of each item. Three negative wording conditions were repeated on scales with the two global items appearing either before or after the 30 diagnostic items. In two studies, 455 undergraduates were randomly administered one of six evaluation forms. Results of a 2 x 3 analysis of variance with repeated measures indicated that the instructor was significantly higher rated than was the course. In neither study were the overall ratings of the instructor or course affected by the negative or positive wording of the diagnostic items. (RL)