Four Perspectives on . . .
23
connections between threat and recommendation. Given these many specific and potentially
important differences, we are reluctant to infer that the results of the two studies contradict one
another. Still, at the broadest level, one investigation suggests that deceleration has relatively
powerful persuasive effects, while the other concludes just the opposite. Future inquiry directed
at identifying the scope conditions for these findings might well find that both hold under
different conditions.
Emotion Dynamics and Experimental Design
Most research on persuasion and affect treats the message as if it were one indivisible
unit. Such an approach may be warranted in many instances. But, evaluating the message as a
whole obscures the fact that persuasive advocacies are constructed from a variety of components
and that each of these components might have a unique impact on emotion. In the current
investigation, measurement of affect (a) prior to the message, (b) after the threat component, and
(c) again after the recommendations component permitted us to create indices of acceleration,
velocity, and deceleration. However, one risk inherent to designs of this sort is that the act of
successive measurement will itself influence the outcome of the investigation. Happily, the data
showed no effect of measurement on any of the variables assessed in this study. Whether or not
message component designs are uniformly nonreactive is a topic for future research.
Of the many alternative means by which one might evaluate the emotional effects of
message components, and by extension, emotional dynamics, the research design employed in
this study was the simplest. Adopting a message component approach becomes more challenging
when one considers suasory appeals that depart from standard methods of organization such as
the problem-solution format. But, methods developed for the study of conversational cognition
Convention