This study was designed to explore two questions: do a listener's reactions differ by media; and do variables such as credibility, saliency, and recall function independently in realistic rhetorical situations? The study was designed to investigate listening variables as they function in field communication situations through the use of multivariate techniques. Of the four groups of subjects, one group listened to a public speech, a second received a television reproduction of the same speech, a third received a tape recorded reproduction, and the final group read a written transcript. Five general findings of the study are discussed, and it is concluded that the listener does not tend to separate messages from the human source that produces them. (Author/RB)