Four Perspectives on . . .
10
demonstrated that individuals who experienced anxiety that was abruptly terminated were more
likely to agree to a request than are persons who were not made fearful. Their work also carefully
ruled out a variety of alternative explanations.
In sum, there are sound reasons to have reservations about the extent to which prior
research has definitively rejected the fear reduction hypothesis. And, the fear-then-relief studies
clearly imply that the deceleration perspective may yet have merit. Thus, we proposed the
following hypothesis:
H6: Fear deceleration is positively associated with persuasion.
A Message-Component Approach to the Study of Emotion and Persuasion
It is evident that testing most of the hypotheses outlined above requires a research design
in which fear is measured at multiple points in time. What may be less apparent is the degree to
which fear appeals lend themselves to clear judgments as to when and how often those
observations should be made.
It is widely held that fear appeals are comprised of two parts (Author Withheld;
Leventhal, 1971; Rogers & Prentice-Dunn, 1997; Witte, 1992). The first is a threat component,
which is comprised of information concerning the nature of the danger and the susceptibility of
members of the target audience to that danger. The second part is an action component that
details the efficacy of the recommended response and the ability of the members of the target
audience to carry out that response. Fear appeals are most effective when the threat component
precedes the recommendation component (Cohen, 1957; Gleicher & Petty, 1992; but see
Leventhal & Singer, 1966). Consequently, it seems clear that fear should be assessed prior to the
Convention