Theories of threat appeals have been rightly concerned with the form of the relationship between fear and persuasion:
Linear or curvilinear. They have not, however, clearly distinguished the question as a between- or within-persons
phenomenon. In fact, the literature often treats these two perspectives as if they were interchangeable. We show that
between- versus within-person questions about functional form are distinct from one another. Previous research,
which is the product of between-persons designs, shows a linear relationship between fear and persuasion. Betweenpersons
studies cannot address the question of how changes in fear over time produce persuasion. Consequently, a
major piece of the fear appeals-persuasion puzzle may have been overlooked. Reanalysis of an existing data set shows
curvilinearity of fear in within-persons data and demonstrates that the curve predicts persuasion. Audience segmentation
reveals different curves for different groups as well as differential associations between those curves and persuasion.
Overall, the argument and the empirical results suggest that a great deal less is known about fear appeals
than it is currently believed.