Four Perspectives on . . .
16
comparison were t (359) = 7.25, p < .0001, r = .36 (M
hi
= 1.70, SD = 1.14 vs. M
lo
= .91, SD =
.93). These findings show an effect of threat on fear almost identical to the average effect of r =
.35 reported in Mongeau’s (2000) meta-analysis. The corresponding values for post-
recommendation were t (359) = 1.69, p = .09, r = .09 (M
hi
= .61, SD = .85 vs. M
lo
= .47, SD =
.74). This finding indicates that persons who read the high threat message remained slightly more
fearful than members of the low threat group even after both had read the recommendation
component of the message. In total, these results suggest that the messages aroused and reduced
fear, and that the high threat message did both to a greater degree than the low threat message.
Hence, the manipulation successfully produced the variation in fear responses that was necessary
for testing the hypotheses.
...FIGURE 1 ABOUT HERE...
H1-H3: The Reactivity Perspective
H1 predicted that the BIS is positively associated with premessage fear. The first entry in
the leftmost column of Table 2 shows a correlation coefficient of .25, p < .01. This result offers
support for the first hypothesis.
...TABLE 2 ABOUT HERE...
H2 anticipated a series of associations between BIS and the three dynamics aspects of
fear arousal. The correlations of .12, p < .05, for acceleration and .26, p < .05, for velocity are
both in line with expectations. However, the sign of the association between BIS and
deceleration ran counter to prediction: r = .18, p < .05. Thus, support for H2 was mixed.
H3 posited a lack of association between BAS and the fear indices. The pertinent results
appear in second column two of Table 2. The coefficients are -.04, .09, .07, and .08 for
Convention