SEARCH

SEARCH BY CITATION

Abstract

Previous studies of fear appeals in advertising have relied on a procedure that measures only the overall level of fear produced by the ad. That procedure cannot validly test the effect of drive reduction, which is the central causal mechanism in Hovland, Janis, and Kelley's (1953) original fear-drive model. To overcome this limitation, the present research identifies the fear pattern of the ad, based on moment-to-moment ratings of fear-to-relief taken for its duration. In Study 1, which examines ratings of seven antispeeding TV commercials, it is shown that a postexposure overall rating of fear is in fact measuring the maximum level of fear experienced, not the average level, and that this static rating of fear cannot distinguish very different patterns, such as the pattern of rising fear with no relief, the “shock” pattern of sudden fear with no relief (both representing positive punishment), and the classic fear-relief pattern (the drivereduction pattern). In Study 2, which is a laboratory experiment using antispeeding road-safety TV commercials and a dependent variable of speed choice on a realistic simulated driving test, it is demonstrated that the fear-relief pattern reduces young drivers' speed choice not only initially but after heavy repetition of the ad, whereas fear without relief increases speed choice initially, then, after heavy repetition, it decreases speed choice, although not down to the driving speed produced by fear-relief. The reason why a shock ad starts to work following heavy repetition is that viewers begin to anticipate the shock and the ad becomes, in effect, a fear-relief ad. The overall results strongly support the fear-drive model. © 2004 Wiley Periodicals, Inc.