This study explored the effect of directions on the Least Preferred Coworker (LPC) scale; specifically, this study tested whether thinking of a generalized least preferred coworker (General LPC) would yield lower scores compared to thinking of a specific least preferred coworker (Specific LPC). The data supported this hypothesis as responses to the General LPC yielded more critical LPC scores than did responses to the Specific LPC. The hypothesis that thinking of a generalized least preferred coworker would yield more stable result than would thinking of a specific least preferred coworker was not supported. Finally, the hypothesis that LPC scores would shift categories (e.g., shifting from task-oriented to relations-oriented) more when thinking of a specific least preferred coworker than when thinking of a general least preferred coworker was not supported. This study provides supportive evidence of the importance of using the original test directions during test administrations.