Four Perspectives on . . .
21
both acceleration and velocity appear to positively influence persuasion. Put differently, the data
show support for the theoretical positions that argue for increases in emotion as a determinant of
persuasion (e.g., Carver & Scheier, 1999) and for those that contend that emotional intensity is a
valid explanation for change in attitude (e.g., Author Withheld; Nabi, 1999). But, this conclusion
must be qualified in one notable way: It was not possible to clearly discriminate the effects of
acceleration and velocity in our data.
We argued earlier in this paper that the two measures should show some independence
because not all time values would be zero (due to naturally-occurring BIS arousal). This
expectation proved true. On a scale that ranged from 0 to 4, the mean pre-message fear score was
.30 with a range of 0 to 3.67. However, the correlation between acceleration and velocity was
.87, a degree of collinearity that was simply too great for regression analysis to distinguish the
unique effects of the two variables.
Clearly, one avenue for future research would be to remedy this limitation in our design.
A study that manipulates pre-message fear by exposing participants to frightening stimuli in
advance of the appeal is needed to clarify the effect of acceleration. However, this suggestion is
not so straightforward as it might appear. One issue to consider is the relevance of the pre-
message stimuli to the content of the threat appeal. Some existing research suggests that pre-
message fear may enhance the impact of a persuasive appeal to the extent that the arguments are
subjectively compelling and the threat is pertinent to the topic of the message (Baron, Logan,
Lilly, Inman, & Brennan, 1994; Gleicher & Petty, 1992). However, other studies indicate that
fear induced by concerns that are irrelevant to the message topic also enhances persuasion
(Hendrick & Borden, 1970; Lundy, Simonson, & Lander, 1967; Sigall & Helmreich, 1969;
Convention