Communication

Theory

Four: Four

James Price Dillard

November 1994 I'amtb:

29s-323

Rethinkin the Study of Fear Appeals: n Emotional Perspective

1

~

A selective review of theories of fear appeals is carried out which concludes that these theories have gradually become divorced from the notion offear. The forces that motivated that separation are examined and a set of arguments are advanced favoring a reunion of fear with a theory of fear appeals. As a means o/ achieving that union, emotion theory is reviewed and a set of axioms extracted. The axiomatic framework is then used to demonstrate how an emotional perspective on fear appeals might shed new light on this area of persuasion. Recommendations are offered f o r future studies of fear appeals.

Traditionally, fear appeals have been thought of as messages that attempt to achieve opinion change by establishing the negative consequences of failing to agree with the advocated position. Because the use of fear appeals is so widespread in both ancient and contemporary history, examples abound. Aristotle offered counsel on the use of fear as a suasory device (Cooper, 1932) and the Reverend Jonathan Edwards encouraged adherence to biblical edicts by describing the fate of sinners: " . . . the devil is waiting for them, hell is gaping for them . . . (Faust & Johnson, 1935, p. 161). Under the stewardship of Everett Koop, the Surgeon General's office of the United States became surprisingly candid about the possible consequences of sex without benefit of a latex barrier. And, at some time or another in our childhoods, fear was fostered by reference to the bogeymen that would visit us should we refuse to eat our peas. Independent of the social or historical context in which fear appeals occur, the theory implicit in use seems to be that persuasion will follow from a state of fright. Social scientific accounts of the operation of fear appeals also began with that premise. Although social scientific accounts James Price Dillard is Professor, Department of Communication Arts and The Center For Communication Research, University of Wisconsin-Madison, Madison, WI. 53706 I would like to thank Austin Babrow, Terry Kinney, Dan O'Keefe, David Mortensen, and Kim Witte for their valuable comments on an earlier version of this manuscript. 295

Communication Theory

have moved away from that premise in the previous two or three decades, they have not forsaken it altogether (e.g., Soames Job, 1988, cf., Sutton, in press). The first portion of this paper offers an historical analysis of theories of fear appeals and finds them wanting. The second portion suggests a new perspective for inquiry and presents some recommendations regarding how to implement that perspective.

The Study of Fear Appeals: A Selective Hlstory For present purposes, it is sufficient to arrange previous work into just

three theoretical families; drive theories, the parallel response model, and subjective expected utility theories. This partitioning of the literature reflects the chronological order in which these positions developed as well as the answers to three substantive questions: 1. What is the nature of fear? Or, more specifically, how is fear conceptualized within each perspective? 2. What role does fear play in the theoretical process outlined by the perspective? Stated in a more applied fashion, what is the hypothesized effect of fear on attitude change, behavioral intention, and behavior? 3. What implication does each theory hold for the analysis and/or construction of fear appeals?’ Drive Theorles

Several authors have advanced fear-as-a-drive accounts (e.g., Janis & Feshbach, 1953; McGuire, 1968; Miller, 1963) and, consequently, there is some diversity among those accounts. Nonetheless, they hold the notion of drive as an unpleasant, energy state that the organism strives to reduce in common. Put differently, drives are “bodily states that initiate tendencies to general activity. These states are frequently experienced as feelings of tension or restlessness” (Newcomb, Turner, & Converse, 1965, p. 23). Drives vary both in type and in intensity (Dollard & Miller, 1950). For example, hunger and fear are both drives, but they are qualitatively different; one is innate, the other is learned. Each of them may assume values indicative of their intensity that range from zero, such as when one has just eaten, to some larger value, such as when one has gone all day on only a bowl of cereal. The fear-as-a-drive models drew from the then-contemporary versions of learning theory for their conceptual parentage (Janis & Feshbach, 1953). In that framework, drives were only one of the four fundamental variables affecting learning (Dollard & Miller, 1950); cues, responses, and reinforcement were also essential parts of learning. Whereas the drive provides the impetus for action, for learning to occur the organism must also produce one or more responses that reduce the intensity of the 296

Rethinking the Study of Fear Appeals

drive. Cues helped to determine the choice of response and responses that resulted in drive reduction were thought to be reinforcing. In the typical fear appeal experiment, fear-as-a-drive was aroused by exposure to a message that detailed the gruesome consequences of failure to behave in accordance with the advocacy. Cues were provided in the message as to the “appropriate” response. The message recipient mentally rehearsed that response and, if a reduction in fear occurred, then that response was reinforced. In those cases in which cognitive rehearsal of the response was not adequate to the task of reducing the fear, then the individual presumably moved on to other responses, such as defensive avoidance, until an effective means of mitigating the drive was hit upon. One important aspect of this formulation deserves emphasis. It is drive reduction, not drive arousal, that is thought to mediate the effects of the fear appeal on attitude (see Figure 1).This is essential because, as Dollard and Miller (1950) note, “ . . . reinforcement is impossible in the absence of drive because the strength of stimulation cannot be reduced when it is already at zero . . . ”(p. 40). In other words, drive reduction is only possible to the extent that drive arousal has already occurred. By implication, a fear appeal should have two components: a part of the message that instills fear and another that assuages it. To conduct a convincing test of the theory, it would be necessary to measure fear at least twice, once at the end of the arousal phase and once at the end of the abatement phase. However, even today, most studies assess fear only once, usually immediately after the message. As data accumulated, drive theorists attempted to accommodate mixed findings. The broadest effort was Janis’ family of curves hypothesis which suggested that there was a point of optimum effectiveness in a curve relating fear and attitude change. The location of that point was determined by a host of potential moderator variables. All in all, a considerable amount of empirical work was carried out under the fear-as-a-drive banner. Leventhal’s (1970)critique of this body of work faulted the drive model on several grounds (see p. 168) including inadequate delineation of constructs, inaccurate specification of moderator variables, and nonfalsifiability of the curvilinear model. These short-

1-

lcssllocCQltmt

I

> I I F e u I I I AmraalI 1I I I 1 I

Figure 1 The drive model This line is not meant to indicate a causal relation, but rather a limiting condition fear reduction IS constrained by fear arousal

I8 1 ->I-

I I I I

I I 1

I -ti=

I I->AttiNdc/ I Bavvior

==v=

297

Communication Theory

comings prompted him to propose an alternative account of fear appeals and their effects (Leventhal, 1970,1971). The Parallel Response Model

This model assumes that fear appeals instigate two independent, but contemporaneous, processes. First, and similar to the drive theories, the parallel response model specifies that fear appeals will cause fear. In addition to making that emotional response to the message, the message recipient can be expected to engage in activities that are adaptive to the threatening circumstances described in the appeal. That is, fear appeals also activate a desire to counter the danger depicted in the message. As Leventhal (1970, p. 124) points out, this dual process feature of the model raises the interesting question of whether similar or different features of the situation elicit both types of responses. Whatever the answer, the individual must manage both response systems. Hence, the message recipient engages in fear control and danger control. Danger control involves the selection and execution of responses aimed at averting the threat. Because the threat exists in the environment, the individual must alter her or his relationship with the environment in order to reduce the danger. Consequently, the danger control process depends heavily upon external cues. In contrast, the purpose of fear control is to reduce or eliminate the unpleasant experience of fear. The sensation of fear arises from “internal bodily disturbances (which are labeled as affective or emotional)” (Leventhal, 1970, p. 172). A variety of methods of fear reduction are possible. These include, but are not limited to, avoidance reactions, distraction, denial, and management of one’s own physiology, perhaps through substances such as alcohol. Figure 2 provides an illustration of these processes. Despite the fact that fear control and danger control operate as separate systems, they may interact with one another in facilitative or disruptive ways. One instance of a facilitative effect is when efforts at danger control have the added benefit of reducing fear. But, when fear control predominates the individual may behave in ways that lessen the effectiveness of the danger control system. For example, refusal to personalize the possibility of acquiring AIDS as a result of maintaining multiple

Figure 2. Leventhal’s ( 1970) parallel reswnse model.

Y

298

Rethinking the Study of Fear Appeals

sexual partners (Freimuth, Edgar, & Hammond, 1987), lessens the fear of the virus, but not the danger. Although the nature of fear is somewhat vague in the parallel response model- it is an experiential state that apparently follows from somatic perception-its status as a causal agent in the persuasive process is more certain. Adaptive behavior results most directly from the danger control process, not from fear or the desire for fear reduction. Fear control may play a part in the process, but whatever impact it has must be mediated by danger control. In retrospect, the major contribution of the parallel response model was to separate the emotional and cognitive aspects of the processing of fear appeals. As noted above, this move raises the interesting possibility that different aspects of the message may instigate different aspects of the two response processes. Unfortunately, the model stops short of specifying which aspects of the stimulus bring about the parallel responses. Because of this ambiguity, the only real implication for message construction is the emphasis that this line of research places on the action plan. In part, it was just this ambiguity that prompted Rogers (1975, p. 96) to propose his protection motivation theory as an alternative. Subjective Expected Utility Models Subjective expected utility (SEU)models view humans as involved in choosing courses of action that they believe will be most likely to lead to rewards and avoid punishments (Lewin, 1935; Tolman, 1932). The basic structure of the model is a simple equation where the dependent variable is a multiplicative function of two,or sometimes more, independent variables.’ For example, Fishbein and Ajzen’s (1975) theory of attitude, as well as Vroom’s (1964) model of job satisfaction, may both be expressed as follows:

A = Cbiei where, A is attitude, b is belief strength (a subjective likelihood estimate), and e is the evaluation (valence) of a given belief. From a statistical standpoint then, the model predicts an significant interaction between belief and evaluation (see also Becker’s 1974 Health Belief M ~ d e l ) . ~ Rogers’ (1975) original protection motivation model relied on a similar formulation but added a third predictor and thus the expectation of a three-way interaction (see Figure 3). He argued that the three dimensions

Figure 3. Rogers’ (1 975) original Drotection motivation

299

Communication Theory

along which individuals gauge a potential threat are its severity /noxiousness, the likelihood of exposure to the threat, and the degree of effectiveness of whatever response might be available to combat it. This latter variable has been termed response efficacy. It is worth noting that fear does not appear in the model at all. Somewhat later, Beck and Frankel (1981) aptly noted that knowledge of an effective response may matter little if the threatened individual is unable to execute the response. Their point is that personal efficacy, as well as response efficacy, is necessary to the implementation of an intended action. This claim won widespread acceptance with the result that personal efficacy has been incorporated into SEU models of fear appeals. The addition of response efficacy brings the number of predictor variables to four and implies an interaction involving all of them. While there is some variety of opinion regarding the optimal means of combining these four types of information (cf., Rogers, 1981 with Sutton & Eiser, 1984; Sutton & Hallett, 1989), there is a fairly clear consensus on certain other issues. First, the expected three- and four-way interactions have not been observed. Consequently, Rogers (1983) revised his protection motivation model to predict only main effects and two-way interactions. Second, researchers seem to concur that all four of the dimensions delineated above are, at one time or another, important sources of information for individuals attempting to determine what response to make to a potential threat. Hence, an effective fear appeal should in some way deal with all four types of information. Two questions remain in our examination of the subjective expected utility models; What is the nature of fear? and What role does fear play in the persuasive process? Attention to the latter question will be helpful in answering the former. Figure 4 provides an illustration of Rogers’ (1983) revised model. While the figure contains several terms not discussed here, and suggests numerous relationships, each of which may be worthy of consideration, the role of fear is of most immediate concern. According to Rogers’ model, fear causes and is caused by perceptions of the severity of the threat and its likelihood of occurrence. Thus, the part played by fear in protection motivation theory was returned to the position granted it by the parallel response model.

-

Of I

Figure 4. Rogers’ ( 1983) revised protection motivation model.

Iniclrrtlm I I

-

-ccgnitive

I

cop-

lebdr I

300

Rethinking the Study of Fear Appeals

In Sutton’s approach, fear is accorded a smaller part. Figure 5 provides a part diagram, taken from Sutton and Eiser (1984), depicting the expected relationships between the cognitive variables and behavioral intention, as well as the expected relationship between fear and behavioral intention. As the figure makes clear, fear is not hypothesized to have any causal impact on attitude or behavior whatsoever. Its status is that of a nuisance variable. In light of these diligent efforts to turn the study of the fear appeals toward cognitive variables, the degree to which the concept of fear is explicated will come as no surprise; there is no extensive conceptual definition of fear offered by any of these approaches. Rogers’ (1975) comes closest when he first objects to the visceral treatments of fear (p. loo), then goes on to create an cognitivized passion-protection motivation. In his revised model, fear is a reciprocal determinant of appraised severity. In the work of Sutton and his co!leagues (Sutton & Eiser, 1984; Sutton & Hallet, 1989), fear is not treated conceptually. However, it is operationally defined as a self-report on items such as nervous, scared, and afraid; and efforts are made to control the effects of fear so that the variables of real interest, i.e., the cognitive variables, can be studied. Understanding the Changlng Status of Fear As I hope the preceding review makes apparent, over the years the “fear”

in fear appeals has undergone a radical transformation. During the heyday of drive theories, fear was at the center of the theoretical stage. With the coming of the parallel response model, it was forced to share the limelight with the cognitive machinations of the danger control process. Yet, affect and fear were not coequal; danger control was given responsibility for adaptive responses to threat, while fear was the scapegoat responsible for irrational actions and difficult-to-interpret findings. Finally, as the subjective expected utility models gained prominence, fear was virtually excluded from the study of fear appeals. In the most recent investigations based strongly on the cognitive perspective, fear has been treated as a control variable (e.g., Sutton & Eiser, 1984; Wolf, Gregory, & Stephan, 1986), if it is measured at all (e.g., Rogers, 1985; Self & Rogers, 1990).

Figure 5 Sutton and Eiser’s ( 1984) proposed causal model of the effects of fear-arousing messages. Solid lines represent hypotheses, dotted lines represent possibilities.

301

Communication

Theory

The shrinking role of fear was evident from the language choices of researchers as well as the path diagrams that accompany those articles. As the importance of fear dwindled, some writers began to suggest that the term “threat communication” would be preferable to “fear communication” (Beck & Frankel, 1981; Rosen, Terry, & Leventhal, 1982). Clearly, that prescription was in keeping with the belief that fright/ anxiety was no longer at the heart of the study of fear appeals. Why did this happen? How did the fear go out of fear appeals? Certainly, one cause was the state of the literature itself. Rogers’ (1975) claim that the literature showed an “intractable pattern of conflicting results” (p. 94), was accurate at the time it appeared, but it could have been made with equal validity at almost any other point in the history of the area. Although it was important, the ambiguity engendered by the apparent paucity of consistent findings was not sufficient in and of itself to explain the virtual demise of the fear construct. Rather, it was significant because it allowed forces at work in the culture of social science to have greater play. In part, the drift toward cognition and away from affect reflected a larger movement in social science. One of the defining features of cognitive science is its de-emphasis on affect (Gardner, 1985, especially pp. 38-45). While many have seen the cognitive revolution as an exhilarating new approach to the study of human behavior, others have bemoaned its effects on noncognitive questions, decrying the effects of “cognitive imperialism” (Tomkins, 1981). In whatever way one judges the value of this trend, there is no denying either its occurrence or its extensive impact on social science. So, just as in many other areas of inquiry, theoretical developments in the study of fear appeals were responsive to the ebb and flow of social scientific fashion. But neither is cognitivism the whole story. There is one other theme that runs through the literature of fear appeals. That is its strong bias toward application. This characteristic of the literature is most evident in the topics of the fear appeals themselves which, almost without exception, are concerned with public health issue^.^ Just a few of the consistently researched topics include dental care (Evans, Rozelle, Lasater, Dembroski, & Allen, 1970; ter Horst, Hoogstraten, & de Haan, 1985), diet (Skilbeck, Tulips, & Ley, 1977), smoking (Insko, Arkoff, & Insko, 1965; Rogers & Mewborn, 1976; Schwarz, Servay, & Kumpf, 1985; Sutton 8c Hallet, 1989), breast self-examination (Meyerowitz & Chaiken, 1987; Siero, Kok, & Pruyn, 1984), sexually transmitted diseases (Self & Rogers, 1990), nuclear radiation (Roser & Thompson, 1991), drugs (Horowitz, 1972; Horowitz & Gumenik, 1970), and alcohol use (Kohn, Goodstadt, Cook, Sheppard, & Chan, 1982; Williams, Ward, & Gray, 1985). And, despite the range of topics, there is tremendous uniformity in the suasory structure of the message. In virtually every case, the essence of the argument is “If you value your health, then you should change your behavior.” 302

Rethinking the Study of Fear Appeals

A quandary resulted from this emphasis on behavior change in that, at many points in the research process, it appeared that attitudes were responsive to a different set of predictors than were behaviors (e.g., Janis & Feshbach, 1953; Leventhal, 1971, p. 1212; Leventhal, Singer, & Jones, 1965; Rogers & Mewborn, 1976; Rogers & Thistlewaite, 1970). Attitude change occurred as a function of fear-greater fear yielded greater attitude change-but behavior was more difficult to predict. It showed a weaker and less consistent relationship to fear, one that appeared to vary as a function of numerous third variables (see Leventhal, 1970). Since the overarching aim of much of the fear appeals research was to persuade the public to act in a more healthful fashion, fear itself became superfluous. Whereas the preceding paragraphs paint the changes in the fear appeals literature in broad strokes, the reasoned decisions of individual researchers should not be overlooked. For example, Leventhal’s shift to a cognitive emphasis was based not only on the fall of drive theory, but also on his findings for information effects (e.g., Leventhal & Singer, 1966). Rogers (1975)rejection of the emotional was based on the apparently unpredictable relationships between the physiological indices of affect and attitude change (e.g., Rogers & Deckner, 1975; but see also Mewborn & Rogers, 1979; Watson, Pettingale, & Goldstein, 1983). In both cases, the researchers made thoughtful decisions on the basis of the available data. Yet, because of the variability in findings, and the uncertainty that naturally arose from such variability, it seems quite likely that more global, cultural forces also shaped the study of fear appeals in powerful ways. Fear Deserves Reinstatement A case can be made for the claim that fear should be reinstated in the study of fear appeals, although it requires a brief digression to consider recent advancements in the conceptualization of attitude (Tesser & Shaffer, 1990). Traditional approaches to the study of attitude have defined the concept in terms of affect, belief, and behavior (or conation). According to different theorists, an attitude consists of all three components (e.g., Breckler, 1984), of only affect and belief (e.g., Bagozzi & Burnkrant, 1979), or of only affect (e.g., Fishbein & Ajzen, 1975). Contemporary theorizing has moved away from these treatments of attitude and towards a reconceptualization of the construct as “the categorization of a stimulus object along an evaluative dimension based upon, or generated from, three classes of information: ( 1 ) cognitive information, (2) affective/emotional information, and/ or (3) information concerning past behaviors or behavioral intentions” (Zanna & Rempel, 1988, p. 319). This move has a number of theoretical advantages. For one, it makes the matter of congruence among the components of attitude an issue to be resolved by empirical inquiry rather than definitional fiat. Moreover, attitudes can be studied either as relatively stable, dispositional entities or as short-lived evaluations that are created to deal with 303

Communication

Theory

immediate and temporary demands in the environment. Both perspectives have been employed in the past, but without much attention to how they might be reconciled (cf., Ajzen, 1984 with Schwartz, 1973). But finally, and of principal concern to the study of fear appeals, is the conceptual distinction between the evaluation itself and one of its bases-affect. Researchers working in the areas of political opinion (Abelson, Kinder, Peters, & Fiske, 1982), contraceptive behavior (Fisher, 1984), responses to victimization (Tyler & Rasinski, 1984), and advertising (Batra & Ray, 1986; Edell & Burke, 1987; Petty, Gleicher, & Baker, 1991 ) have all demonstrated reliable relationships between affect and attitude. What is particularly compelling about each of these studies is that affect shows an impact on attitude and/or behavior that is independent of either cognition or behavior. In short, there exists a growing body of work, both theoretical and empirical, that indicates the general value of examining the impact of various feelings on attitude. Consider, as well, the findings on fear in particular. Quantitative summaries of the fear appeals literature concur that fear is reliably and substantially related to acceptance of the message (Boster & Mongeau, 1984; Mongeau, 1990; Sutton, 1982). Application of meta-analytic methods has given some coherence to that apparently “intractable pattern of conflicting results’’ to which Rogers referred (1975, p. 94). The most recent of these three meta-analyses (Mongeau, 1990), reports an average correlation coefficient of .20 between perceived fear and attitude, and a coefficient of .17 between perceived fear and behavior. The implications of these meta-analytic results are numerous. First and foremost, it is apparent that fear does seem to play a meaningful role in the persuasive process. However, it is not a simple effect, but one that is apparently qualified by interactions with age and anxiety (Boster & Mongeau, 1984; Hale & Mongeau, 1991; Mongeau, 1990). And, while all of the theories reviewed earlier in this paper predict some kind of interaction, none of the theories are compatible with the meta-analytic findings. We have a set of empirical generalizations absent an explanation. Second, the association between manipulated fear and perceived fear ( r = .34 in Mongeau, 1990), while nontrivial, is in many respects not all that informative. Since the meta-analysts were largely unable to examine the effects of message variables on fear induction (because the messages were not available), we still lack reproducible knowledge concerning how to construct fear arousing messages. In the absence of a viable theory of fear appeals and a plethora of previous operationalizations, we know only that experimenters’ intuitions about the construct of fear appeals have been pretty good. Overall then, while the meta-analytic findings argue strongly for the inclusion of fear in the study of fear appeals, those same findings underscore how little is known about the operation of fear appeals after almost four decades of empirical research .5 304

Rethinking the Study of Fear Appeals

Together there are three kinds of evidence that support the claim that fear should be returned to the study of fear appeals. First, there are several studies, done in a variety of contexts, that show an effect for affect after controlling for cognitions and/or behavior. Those studies provide strong evidence of the potential for affect to uniquely influence attitude, but they typically treat affect in a global fashion. For example, Abelson et al. (1982) examined affect simply as positive or negative. T o the extent that negative affects such as anger, fear, and sadness have discriminable effects on attitude the link between evidence and claim is weakened. There are also the meta-analytic results. Resting on the foundation of many studies and many subjects, the strength of these results lies in their reliability. What they demonstrate so convincingly is that fear covaries with attitude change. However, because of the limitations of many of the primary studies, the meta-analyses cannot speak to the causal role of fear. Fortunately, a third form of evidence exists. There is a handful of primary-data studies of fear appeals that measure both fear and cognition, and conduct analyses that allow for an estimate of the unique effects of fear. Sutton and his colleagues accomplish this in four studies (reported in two papers; Sutton & Eiser, 1984; Sutton & Hallett, 1989). Path analyses of each of the four data sets estimated the effects of fear and the SEU cognitive variables on intention to attempt to stop smoking. Three of the four analyses revealed a significant and direct effect for fear. These findings are buttressed by the work of Roser and Thompson (1991) who used a signalled stopping technique to assess cognitive activity during presentation of the message. Their analysis also showed a significant and direct effect for fear after statistically controlling for cognition. Rogers and Mewborn (1976) fit a path model to their data in which the effects of fear on intention are mediated by perceived severity.6 Although counting significance tests in a set of small sample studies (as these are) is a procedure that is biased in favor of Type I1 error (Hunter & Schmidt, 1990), four of the five studies mentioned above point toward a reliable and unique effect for fear. In sum, there are several reasons for putting the fear back into the study of fear appeals: because it is in keeping with recent advances in attitude theory; because empirical studies show a relationship between affect and attitude; because the meta-analyses reveal a durable association between fear and attitude; and because a majority of primary-data studies that controlled for cognition still showed an effect for fear on intention. Individually, these pieces of evidence are not especially compelling. When considered collectively, however, they make a strong case against continuing to ignore the role of fear in fear appeals. How can fear be reinstated? What is obviously missing is a viable theory of fear appeals.’ At a minimum, such a theory must be responsible to the three issues that drove this analysis of previous theories; (1) what 305

Communication

Theory

is fear? (2)what is it about messages that creates fear? and (3)how does fear figure into the persuasive process? The term that appears in each question is fear. Thus, it is apparent that before we can begin to formulate responses to either of the latter two questions, we must concern ourselves with the former, i.e., nature of fear. For guidance we may turn to Willie Sutton, the bank robber (no relation to S.R. Sutton, the fear appeals researcher). When asked why he robbed banks, Sutton replied “Because that’s where the money is.” That compelling logic suggests that we might work toward a conceptual delineation of fear by examining the emotion literature. An Emotional Perspective on Fear Appeals

My aim in this section was to exploit existing theory and research on emotion for the purpose of sketching a new perspective on fear appeals. While the perspective that I hoped to achieve is theoretical, it does not constitute a theory. Rather than strive for a high degree of explanatory power, my emphasis was on heuristic value. Hence, the strategy that I pursued in reviewing the emotions literature was one of isolating those aspects of existing work that showed substantial consensus, and then using those features of the literature to ask questions about fear appeals. It is worth mentioning that the emotion literature is dynamic terrain. When consensus on a given issue was less than complete, I leaned toward the position that appeared to have the greatest potential value for the study of fear appeals. In my estimation, the following propositions have attained the status of axioms: Axion 1 : Emotions arise from an assessment of the situation vis a vis the individual. Frijda (1986) offers a broad synthesis of current theories of emotion. He argues that: Emotion may be viewed as the result of stimuli interacting with dispositional entities like response tendencies, motives, goals, expectations; or equivalently, as the result of stimuli being relevant to obtaining incentives or reinforcementsor to maintaining well-being. Positive emotions can be seen as produced by stimuli representing match with expected or desired situations: with achievement of goals, satisfaction of motives, realization of response tendencies, acquisition of incentives. Negative emotions can be seen as the result of stimuli representing mismatch with any of the above. (p. 265)

The essential aspect of this view is the process of assessing match. Consequently, it is not surprising that theoretical elaboration of that appraisal process is at the center of many theories of emotion (e.g., Lazarus, 1991; Scherer, 1984; Smith & Ellsworth, 1985; Weiner, 1986). Axiom 2: Emotions have a biological basis. Emotions exist as part of a hard-wired communication system, the function of which is to enhance 306

Rethinking the Study of Fear Appeals

the survival of the species. Oatley and Johnson-Laird (1987) suggest that emotions act to facilitate transitions between plans in a system of multiple goals. That is, emotions function so as to move the human system into a particular mode and to maintain it in that mode. The function of fear in particular is protection (Plutchik, 1980). In this view then, emotions are information that alerts one aspect of the human organism to the status of some other aspect. But the communication system to which emotions are central is not a strictly intraorganismic affair. For species such as humans, in which the offspring are helpless for extended periods after birth and are the target of predatory pressures, a hard-wired communication system between the provider and recipient of care is a necessity (Izard & Malatesta, 1987). The survivability of the species is enhanced if the infant can signal its internal state and the caregiver can recognize and interpret those signals (see Cappella, 1991 for a more extended treatment of these issues). One medium by which much of this information exchange takes place is facial expression. Another is vocal behavior. Axiom 3: The cognitive mechanism that conducts appraisal is shaped by learning. Although Axion 2 specifies a biological basis for emotions, this is not to say that all emotional experience stems from innate factors nor that humans are born with a fully developed and operational emotional system. Rather, the rudimentary aspects of the system are innate, but the basic form and function of the emotional system is elaborated upon via learning as the organism develops.* This feature of the emotional system allows for enormous flexibility in relating to the environment which, in turn, has advantages for survival. Individual organisms can abstract from their own experience changing person-environment relationships and adapt accordingly. Moreover, the capacity for knowledge acquisition allows for the group of conspecifics to transmit the products of group learning intergenerationally. Axiom 4: Emotions unfold over time. Conceptual analyses of the appraisal process sometimes draw on a distinction between assessment of the situation and the assessment of how to cope with the situation. For example, Lazarus (1990), distinguishes between primary appraisal, which refers to the judgments that one makes about the relevance and valence of environmental events for oneself, and secondary appraisal, which is concerned with the resources and options that one has for coping with those same environmental events.’ Both types of appraisal may influence emotional response. If one recognizes a danger in the environment, that may produce fear. Continued consideration of the stimulus may cause fear to segue into anger if the individual perceives her or himself as more powerful than the stimulus, into relief if the individual judges her or himself to have initially misjudged the stimulus, or into greater fear if the individual assesses her or himself as incapable of dealing with the danger. Although this example highlights the influence of cognition on emo307

Communication Theory

tion, the reverse causal sequence may also obtain, i.e., affect may causally impact cognition. Recently, Isen (1987) has provided a review of the influence of positive affect on a variety of cognitive operations including memory at encoding, memory at retrieval, the use of intuition, and reliance on heuristics. Other work, more immediately relevant to persuasion, shows an interaction between valence of affective state and argument strength on cognitive responding (Bless, Bohner, Schwarz, & Strack, 1990). However, in day-to-day life, cognitive and affective processes are so closely interwoven as to resist easy segmentation. So, rather than conceive of the causal process as one in which cognition causes emotion which in turn causes cognition ad infinitum, it has been suggested that the interplay might be more aptly characterized as a cognitive-emotional fugue (Lewis, Sullivan, & Michalson, 1984, p. 264). The music metaphor highlights both the extent to which cognition and emotion are interwoven, while still suggesting more than a trivial degree of conceptual independence between the two. Moreover, just as different instruments or players are showcased in different sections of a composition, distinct internal processes may contribute disproportionately to the subjective gestalt as emotions unfold over time. Axiom 5: People make efforts to regulate their emotions. Emotional regulation refers to attempts to augment or inhibit particular affective states. Although such efforts need not be actively conscious, there is little doubt that people do attempt to manage their feelings. Two examples illustrate this: Zillman (1988) has argued that individuals selectively use television to alter their affective state, while Dillard (1990) has claimed that persons choose to enter or exit interpersonal influence episodes in order to manage their arousal (see also Isen, 1987). Various means of emotional regulation are possible. The list below draws heavily from Folkman and Lazarus’ (1990, pp. 318-325) work on coping. 1. Attention control. An individual might avoid unpleasant emotion by engaging in some other activity, e.g., sports or television viewing, or by makmg efforts to limit the input that is causing the emotion, e.g., turning away from the movie screen. In contrast, cognitive or physical activities might be postponed so as to savor a pleasant affect or individuals may exert effort to focus their attention on the source of positive stimulation so as to maximize its intensity. 2. Subjectively changing the person-environment relationship. Denial of the existence of the threatening stimulus, or denial of its relevance, can have the effect of lessening the intensity of the resulting negative emotion. The opposite sort of strategy is often employed by persons seeking to maintain hope, that is, they overestimate the likelihood of attractive events (McGuire, 1960). A more active version of this strategy is called distancing. This often takes the form of self-talk, e.g., “I counted to 10 to let my anger pass” or “I told myself that I wasn’t going to let it bother me.” 308

Rethinking the Study of Fear Appeals

3 . Objectively changing the person-environment relationship. Another option for controlling affect is to put oneself in those situations that are likely to initiate positive emotions and to escape those situations that are likely to bring about negative emotions. But, sometimes avoidance is not possible and therefore actually altering the nature of the person-environment relationship is required. This is typically the aim of fear appeals. Fear Appeals: Explicatingthe Adjective

With the overarching structure provided by the axioms in place, it is possible to be more precise about what fear is. T o begin, fear is an emotion. Emotions are complex, but coherent, phasic responses to ongoing assessments of the person-environment relationship (cf., Kleinginna & Kleinginna, 1981). Emotions may be characterized by the pattern of responses to person-environment assessmentsin four domains; subjective experience, action tendency, expressive behavior, and physiology. Subjectively, the emotional response of fear is described by terms such as apprehensive, frightened, scared, and terrified. The corresponding action tendency is one of avoidance (Frijda, Kuipers, & ter Schure, 1989). However, this tendency does not invariably result in flight. Other behavioral instantiations of the protection function include (1) withdrawal into a protective enclosure, a strategy employed by oysters and other bivalves, (2) removal of the threatening stimulus, e.g., removing the leech from one’s skin, (3) active avoidance of situations that might place the organism in close proximity to the threatening stimulus either physically or psychologically (Archer, 1979). Next, consider the expressive aspects of fear. The substantial literature on emotional expression has shown fear to be characterized by raised eyebrows that are pulled together, by raised upper eyelid and tightened lower eyelid, and by horizontal stretching of the lips (Izard, 1971). Additionally, although the research base is small, vocal behavior may properly be considered an aspect of emotional expression (Buck, 1984; Scherer, 1981). In the physiological domain, a vast array of changes take place. For example, participants in a cross-cultural study report that fear was associated with feeling cold, increased heart rate, trembling, muscle tension, and an unpleasant feeling of arousal (Scherer, Summerfield, & Walbott, 1983). The validity of these perceptions is borne out by studies that have measured bodily processes more directly. Those investigations showed that, relative to baseline, a state of fear is accompanied by higher levels of skin conductance, respiration, heart rate, and cortical arousal, but lower blood flow to the extremities (Frijda, 1986, p. 163, provides a review). The physiological changes that help to define fear can be imprecisely, and only imprecisely, characterized as arousal. It is misleading to speak of a unitary arousal construct since the various physiological subsystems are only loosely dependent upon one another (see e.g., Davidson, 1978; Notarius & Herrick, 1989, pp. 400-402; Wagner, 1989) and may vary in their relationship to one another, converging in some instances and 309

Communication Theory

diverging in others (Lacey, 1967). One illustration of this phenomenon has already been mentioned. Fear, which might be indexed by an increase in heart rate (Levenson, Ekman, & Friesen, 1990), could be accompanied by vigorous movement away from the threat or by freezing (and this would depend on the organism and its appraisal of the stimulus). T o push the example further, fear is also characterized by a decrease in skin temperature (Levenson et al., 1990). Thus, the two physiological indices, heart rate and skin response, tend to be inversely correlated during fear, while the association between action and either heart rate or skin temperature will depend on other factors. Relationships among the four domains of emotional response are made more complex by the recognition that the systems need not exhibit parallel behavior over time. One illustration of this claim comes from work on excitation transfer theory. For example, Cantor, Zillman, and Bryant (1975) show that the perception of arousal decays more rapidly than does actual physiological arousal (indexed in that study by heart rate and systolic blood pressure). This same lack of parallelism may be observed within the multi-faceted physiological system. For example, electrocortical activity occurs far more rapidly than does electrodermal activity. The general point to be made here is that fear is more than simply the phenomenological experience of fear. Rather, in this application, fear is more usefully conceived of as a pattern of responses over time across all four systems." Insofar as we may wish to infer the presence of fear from the observation of these systems, our confidence in that inference should increase as the coherence of the pattern of responses increases (Ekman, 1977). Fear Appeals and the Cognitive Apparatus

In keeping with Axiom 1, several investigations have recently sought to establish that particular emotions are associated with specific patterns of cognitive appraisal. They have been quite successful (e.g., Ellsworth & Smith, 1988; Fridja et al., 1989; Reisenzein & Hofmann, 1990; Roseman, 1984; Roseman, Spindel, & Jose, 1990; Smith & Ellsworth, 1985, 1987). And, while the findings are not univocal for fear, there is substantial consensus across studies. Fear tends to be characterized by the perception of a stimulus that ( 1 ) is important, (2)is negatively valenced, (3) is impending, (4) will require considerable effort to deal with, i.e., presents an obstacle, and (5) is beyond the control of the actor. There are at least two implications of this list worth highlighting. One is that the appraisal variables bear considerable resemblance to the dimensions of cognitive assessment that fear appeals researchers have developed over the years. In fact, points ( 2 ) ,(3), (4), and (5) precisely parallel the following constructs; perceived severity of the noxious event, its likelihood of occurrence, response efficacy, and self-efficacy, respectively (cf., Figure 3). Such convergence of results is reassuring. It indicates that two relatively independent research traditions are moving towards similar findings. 310

Rethinking the Study of Fear Appeals

But, in a second, different respect, both the findings described above and current theories of fear appeals seem less than satisfactory because they ignore important features of emotion that are specified in Axioms 2 and 3. Specifically, both approaches assume a well-developed cognitive apparatus and considerable motivation to use it. While these assumptions are no doubt valid some of the time, Axiom 2 insists that the capacity to experience and express emotion is in place prior to the full maturation of the cognitive apparatus. In fact, humans may experience fear as early as nine months of age (Sroufe, 1984), though it seems doubtful that infants less than a year of age are capable of making careful, ratio-level judgments of severity, likelihood, response efficacy and self-efficacy.Thus, assuming that developmentalresearchers are correct about the ontogeny of fear and we are willing to retain the assumption that emotions are reactions to assessments of the environment (Axiom l ) , then it must be possible for fear to result from causes other than the appraisal patterns summarized in the preceding paragraph. Given the axiomatic framework developed earlier, the most parsimonious explanation is that certain configurations of environmental stimuli produce fear independent of learning. The facial and vocal behavior of persons in the environment are again suggested as likely sources of such stimuli. This reasoning does not suggest that the existing theories of fear appeals are wholly incorrect nor that the empirical tests of Axiom 1 are spurious-only that they are incomplete. One plausible move in response to this theoretical problem is to posit a hierarchical cognitive system that functions on both the innate and learned levels simultaneously (e.g., Leventhal & Scherer, 1987).” A study by McHugo, Lanzetta, Sullivan, Masters, and Englis (1985)helps to make this point. Subjects in this investigation viewed brief clips of televised appearances by then-President Reagan. Each clip contained a different facial expression: happiness, fear, anger, or neutrality. Exposure to the clips was sufficient to induce changes in selfreported affect, facial EMG, skin resistance, and heart rate, all of which were largely consistent with Reagan’s facial expressions. Importantly, the intensity of the subjects’ responses was moderated by their attitudes toward the President such that Reagan backers showed a stronger tendency to match his facial displays than did Reagan opponents. These findings are notable in two respects. First, they demonstrate the power of facial displays and automaticity of response even in mature organisms. This is consistent with the claim that emotions have a biological basis (Axiom 2). Second, because attitude shaped emotional response in concert with hard-wired processes, the results also demonstrate that the cognitive system is capable of modifying the operations of the innate emotional system (Axiom 3). Hence, this study underscores the presence of innate processes in affective responding, while at the same time pointing clearly to the need to consider the effects of learning.” In combination with the arguments advanced earlier in this section, these findings point toward the need to examine the message-affect linkage in terms of a hierarchical cognitive system. 311

Communication

Theory

Temporal Dynamics of Fear Appeals Attention to the temporal dimension of fear suggests a whole host of

questions. For example, how rapidly does fear follow from a fear appeal? Where is the apex of the emotional experience? How quickly or completely does fear decay? Questions such as these focus on the overtime behavior of fear once the emotion has been instigated. But ultimately they are too simple. Axiom 4 emphasizes the intimate, and reciprocally causal, relationship between affect and cognition. Add to that the recognition that these two conceptually distinct processes interact with one another over time and several interesting possibilities emerge. First, if message recipients are viewed as passive consumers who absorb the appeal at the same rate that it glides across the television screen or is discharged by the radio, then message and response are tightly linked and unidirectionally ordered, i.e., message causes emotion which causes some outcome. In this view, the researcher is encouraged to focus on content features of the message that may create, intensify, or diminish affective response. Message-instigated affective responses result from the placement of information and emotionally evocative stimuli at different points in the message. Some research has already examined the influence of variations in the structure of the fear appeal (e.g., Leventhal & Niles, 1965). Using a post-test-only design, Leventhal and Singer (1966) found that subjects were least fearful when confronted with a message in which the recommendations preceded the fear-arousing material versus fear-induction followed by recommendations or an intermingling of the two types of material. However, the findings for the effects of recommendation position on acceptance of the advocacy are mixed (cf., Leventhal & Singer, 1966 with Skilbeck et al., 1977). Investigations of the impact of the delay between fear arousal and presentation of the action plan concur that delay limits the effectiveness of the appeal (Cecil, Weiss, & Feinberg, 1978; Skilbeck et al., 1977). Such studies are valuable, but they are remarkably few in number and, because they typically use post-test-only designs, they do not permit inferences about the dynamic interplay between cognition and affect. Noncontent features of the message have been incorporated into investigations almost completely on a intuitive basis. Efforts to increase the strength of the fear manipulation have included variation in the size of the viewing screen (Leventhal & Trembly, 1968), presentation of the message in color versus black and white (Insko et al., 1965), the use of second person versus third person language (Ramirez & Lasiter, 1977), and variation in the graphic detail of the pictures that accompanied the text (Dziokonski & Weber, 1977). Typically, investigators manipulate several of these variables simultaneously (Witte, 1992a, is a good example). While that is a sound strategy for strengthening a manipulation such an approach makes it impossible to parcel out the effects of discrete variables on affect or attitude, much less to trace their impact on message processing over time. 312

Rethinking the Study of Fear Appeals

Second, if one moves away from the global, unidirectional perspective and places greater emphasis on the reciprocal aspect of the affectcognition relationship, then questions arise regarding the manner in which early message features may influence judgments of later message features. Johnson and Tversky (1983) provide an example. In their study,,mild fearlanxiety was induced in subjects who then judged a variety of risks, ranging from natural disasters to disease, as more likely than did the control group. This suggests that emotionally evocative stimuli placed early in the message, or even prior to the message, could produce affective responses that feed into higher-level cognitive judgments about components of the argument which could, in turn, contribute to overall fear arousal, acceptance of the advocacy, and so on. Third, there is also the possibility that emotions may causally influence one another, That is, the instigation of one emotion may yield up another emotion independent of any message effects. Opponent process theory predicts that following the removal of an affect-inducing stimulus, an affective response will occur that is opposite in valence to the initial response (Solomon, 1980). For the emotion of happiness (but not sadness), Mauro ( 198 8) presents both self-report and physiological findings that comport with that prediction. More central to the present focus, he also reports tentative evidence of an opponent process effect for fear (Mauro, 1986, cited in Mauro, 1988). To summarize, three perspectives on the temporal dynamics of fear have been sketched out. The first assumes that messages impact cognitions which in turn determine emotional response. The second perspective builds on the first in that it recognizes that emotion may shape cognition as well as the reverse. The third and final perspective puts forth the possibility that the occurrence of one affect may spark other affective responses that are unconstrained by either the message or the cognitive system. The point here is not that one approach is superior to the other, for they are not competing perspectives. Each one is likely to be useful at least some of the time depending, as always, on the investigator’s questions. These three perspectives should serve to illustrate the breadth of questions that can plausibly be asked about the operation of fear appeals when the temporal dimension is given explicit recognition. Individual Differences

Perusal of the studies that have investigated the role of trait anxiety and the effectiveness of fear appeals reveals a hodge-podge of findings. Fortunately, Boster and Mongeau’s (1984; also Mongeau, 1990) metaanalysis suggests that most of that variability is due simply to sampling error. They do, however, report an interaction between trait anxiety and perceived fear such that the greatest attitude change was manifested by those subjects low in trait anxiety and high in perceived fear. Recently, this pattern of results was replicated in a primary data study (Hale & Mongeau, 1991). While such findings raise a number of theoretical questions, they also make one point quite clearly: if trait anxiety is as impor313

Communication

Theory

tant to the operation of fear appeals as these data suggest, then it will be important to include a measure of it in every future study. Jepson and Chaiken (1990) also examined the relationship between trait anxiety and the processing of a fear appeal that advocated regular check-ups for cancer. One conclusion that emerged from their efforts was that trait anxious persons tended to devote less effort to processing the appeal. A parallel effect was observed for those individuals with chronic fear of cancer. Both findings are compatible with the attentional control mechanism outlined in Axiom 5 , i.e., individuals may attempt to regulate their unpleasant affect, fear in this instance, by diverting their attention. But, Jepson and Chaiken’s findings add a twist. Since acute fear did not produce decrements in processing, it appears that measures of both trait anxiety and chronic fear may be indexing a learned avoidant pattern of coping with a negative emotion. Perhaps, when individuals already have an attitude, but they still have to deal with the negative affect that is aroused by presentation of the topic, that they activate previously learned routines to direct their attention away from both the message and the topic. In sum, the results of this study suggest that individuals may manage their responses to messages differently depending on their knowledge of the topic and the level of practice that they have in thinking about it. This possibility implies that, along with an index of anxiety, future investigations should give serious consideration to assessing knowledge of the topic of the persuasive message. Furthermore, it suggests that attitude formation via fear appeal may be a substantively different process than attitude change.

Recommendations

Quite often fear appeals are employed in efforts to convince people to change “for their own good.”This is manifestly the case with public health messages, those suasory efforts fashioned by government to improve the well-being of the populace. The longstanding interest in public health has molded the study of fear appeals. And, while it is important not to lose touch with the applied questions, the perspective advanced here suggests that we need to expend more energy grappling with some basic theoretical issues. A series of broad implications for persuing knowledge in this area emerge from this emotional analysis of fear appeals. First, it is important to recognize that studies of the relationship between message features and affect are valuable in their own right. Pertinent to this point, O’Keefe (1990, p. 127) advances a useful distinction regarding the definition of message variables. He notes that message variables may be delineated either in terms of the intrinsic features of the message or in terms of the responses of the recipients of the message. The

314

Rethinking the Study of Fear Appeals

earliest studies of fear appeals tended toward the latter course-a fear appeal was whatever scared the audience. Research driven by the parallel response model and the subjective expected utility (SEU) approach gave more systematic attention to the intrinsic content features of the message and comparatively less attention to their affective impact. The emotional perspective suggests that attention to content should be maintained, but that in addition, much could be gained by examining some of the many noncontent features of every message. Two such features, considered briefly in this paper, are the facial and vocal behavior of the message sources. Obviously, many other possibilities exist. Regardless of their number, one conclusion is apparent: the status of the message-affect relationship should be elevated beyond that of a manipulation check. Communication science should be actively pursuing an understanding of the relationship between message form and emotion. For the sake of clarity, we might do well to adopt the term “threat communications” to refer to those messages that describe impending negative consequences for the message recipient. This would be in keeping with the definition-by-intrinsic-feature described by O’Keefe ( 1990). This would help to make clear that fear appeals ought to have one additional characteristic, i.e., they should induce fear. How that is accomplished is an essential question for a theory of fear appeals. Efforts to understand the fear-inducing properties of mass media messages are already underway (Cantor, 1991, offers a review). Second, it would be fruitful to grant greater emphasis to the role of affect in cognitive processing of persuasive messages. At present, it is possible to see that fear might influence cognition in several ways. Under different conditions, fear might either augment or diminish the amount of cognitive capacity available for message processing. Furthermore, fear might enhance or inhibit motivation to engage in message analysis. Changes in either motivation or capacity could stem from unconscious processes or from active efforts to manage one’s emotional state. Alternately, fear might be used as information upon which to base an attitudinal or behavior response. Another possibility is that fear will indirectly influence certain specific kinds of cognitive judgments such as the likelihood of occurrence of the noxious event or its appraised severity. In combinations these alternatives present an enormous array of possibilities. Yet, my third recommendation suggests that we expand upon that array even further. In keeping with the definition of emotion as a gestalt based on several different systems-phenomenogical, motivational, expressive, and physiological-it will be important to design studies that tap into more than one of these systems. Self-reports of affective states have been an important source of information in previous studies of fear appeals and certainly should not be abandoned. However, assessment of the behavior of other systems, particularly the multi-faceted physiological system, are needed to complete the picture.

315

Communication

Theory

Fourth, the temporal dynamics of persuasive messages deserve far greater attention than they have received. The drive theories, which suggested that fear must go up in order to come down, implied a certain message structure and the expectation of temporally-ordered affects. Later perspectives tended to treat messages as entities that required certain components; severity, susceptibility, response-efficacy, and selfefficacy. Nevertheless, those same perspectives seemed to conclude that the arrangement of the components was unimportant (or too complex to study). The emotional perspective emphasizes the possibility of various overtime effects as message, cognition, and affect engage one another. Examination of the temporal aspects of fear appeals holds enormous promise for advancing our understanding of this aspect of persuasion. The three models of temporal dynamics outlined earlier may help to clarify thinking regarding the type of question that one seeks to address. Fifth and finally, an emotional perspective on persuasion suggests that the study of fear appeals alone is too narrow a focus. Rather, fear appeals research should be situated within a matrix of questions that probe the relationship between affect and persuasion. In support of this conclusion is the observation that, in the real world, emotions rarely ever occur in pure form, i.e., unaccompanied by other emotions (e.g., Lazarus, 1991; Schwartz & Weinberger, 1980). In fact, there is evidence from fear appeals studies which indicates that anger often co-occurs with fear (Dabbs & Leventhal, 1966; Leventhal, Singer, &Jones, 1965). At minimum, fear appeals researchers must begin to control for the effects of other affects. However, a far more productive tack would be to broaden our sights beyond a single emotion and to aim for a general theory of affect and persuasion.

Summary

At the beginning of this article, I reviewed the social scientific theories of fear appeals, and tried to make clear that those theories did what theories are supposed to do: they guided research, suggested new questions, and offered frameworks for the interpretation of research findings. Yet, as they developed, they veered away from their original reason for being, until finally, the theories of fear appeals lost contact with the notion of fear. Several arguments were developed in favor of returning fear to a more prominent position in the study of fear appeals. If one accepts the conclusion of those arguments, that acceptance gives rise to the question of how to best accomplish the reinstatement of fear. A perspective developed from emotion theory was offered as a response to that question. Key features of the emotion literature were first identified and then used as a means of posing questions about fear appeals and re-examining some existing research. This analysis generated a series of specific implications for future research as well as five general

316

Rethinking the Study of Fear Appeals

recommendations. Taken as a group, these propositions and recommendations offer a point from which to begin the study of fear appeals anew.

'

Reviews of broader focus are available elsewhere (e.g., Boster & Mongeau, 1984; Higbee, 1970; Miller & Hewgill, 1966; Mongeau, 1990; Sternthal & Craig, 1974; Sutton, 1982). My discussion of these models emphasizes their common history and their similarities. However, there are sufficient differences among them that it is not unreasonable to cast them as competing theories (seee.g., Weinstein, 1993). Any model that contains a multiplicative term implies an interaction between the components of the product. If only main effects were intended the two (or more) terms in the equation would relate additively. The product term indicates that the effect of one variable is thought to depend upon the value of the other variable(s). An important, but not widely appreciated, feature of this model is its specification of a within-persons interaction term. That is, the model calls for product terms to be formed then summed within persons. Most tests of the model have summed the main effect terms then formed a product from those sums, a procedure dearly at odds with the theory. ' Whiie there are studies that make use of nonphysical threats (e.g., Hale & Mongeau, Powell & Miller, 1967; Robberson & Rogers, 1988), they form a distinct minority. '1991; Although I am dating the beginning of the study of fear appeals with the appearance of Janis & Feshbach (1953), we could just as easily attribute it to Aristotle. But, to do so is depressing since it is to conclude that we have learned nothing in over 2000 years rather than just 40. Shelton and Rogers (1981) also report a path analysis in which fear does not affect intentions either d i r d y or indirectly. Although Shelton and Rogers would probably disagree, it is not evident that fear should be expected to influence intentions since their message depicted dangers to whales rather than to the subjects themselves. During the period that this paper was in the review process, Wine's (1992a, 1992b) reformulation of the parallel response model appeared in the literature. It appears to be a very promising attempt to revitalize the study of fear appeals and, importantly, the initial tests of the theory are quite supportive. However, given the relatively few studies completed to date it is too soon to tell how the model will fare in the long run. * One wrinkle in this otherwise neat picture is Zajonc's (1980) daim that."preferences need no inferences" (p. 151).His assertion, and the data that support it, would appear to directly contradict the assumption implicit in Axiom 1, i.e., that thinking precedes feeling. However, this disagreement is more apparent than real since it hinges entirely upon how much cognitive activity is required to constitute thinking (Leventhal & Scherer, 1987). In the Current view, it is most useful to assume that cognitive activity may take many forms ranging from very rapid and often out-of-conscious pattern recognition to systematic, deliberative contemplation. In this broad conception, some degree of cognitive activity always precedes emotion. This same distinction is evident in the fear appeals literature in a slightly more elaborate form, but lacking the implication of temporal sequence. Rogers (1983) discusses threat appraisal as a combination of the likelihood and severity of an event, and coping appraisal as a combination of self-efficacy and response efficacy. But for present purposes, the essential aspect of the distinction is the temporal sequencing. lo Of course, the relationships among the systems is a topic worthy of study in its own right and some programs of research have made such relationships the target of their endeavors. One example is work on the facial feedback hypothesis, in which researchers have attempted to establish that facial expressions have a causal impact on physiological and experiential response (Laird, 1984; Matsumoto, 1987). I' To propose a cognitive system that is resmcted to two levels is to oversimplify the picture of the human mind. However, an appreciation of the complexities is not required for the argument. In passing, it is worth noting that the results of this study also support the model of amtude described earlier in this paper. It was claimed that amtudes are based on affect,

'

'

317

Notea

Communication Theory

cognition, and behavior and that attitudes can influence each of those three entities. In the McHugo et al. (1985) study we see that attitude towards Reagan shaped both emotional and physiological responses to the stimulus.

References

Abelson, R. P., Kinder, D. R., Peters, M. D., & Fiske, S. T. (1982), Affective and semantic components in political persons perceptions, lournal of Personality and Social Psychology, 42,619-630. Ajzen, 1. (1984) Attitudes. In R. J. Corsini (Ed.), Wiley encyclopedia of psychology, vol. 1. New York: Wiley. Archer, J. (1979).Behavioral aspects of fear. In W. Sluckin (Ed.), Fear in animals and man (pp. 56-124). New York: Van Nostrand Reinhold. Bagozzi, R. P., & Burnkrant, R. E. (1979). Attitude organization and the attitudebehavior relationship. Journal of Personality and Social Psychology, 37,913-919. Batra, R., & Ray, M. L. (1986).Affective responses mediating acceptance of advertising. Journal of Consumer Research, 13,234-249. Beck, K. H., & Frankel, A. (1981). A conceptualization of threat communications and protective health behavior. Social Psychology Quarterly, 44, 204-217. Becker, M. H. (Ed.) (1974). The health belief model and personal health behavior. Thorofare, NJ: C. B. Slack, Inc. Bless, H., Bohner, G., Schwarz, N., & Strack, F. (1990). Mood and persuasion: A cognitive response analysis. Personality and Social Psychology Bulletin, 16,331-345. Boster, F. J., & Mongeau, P. A. (1984).Fear-arousing persuasive messages. In R. Bostrom (Ed.), Communication yearbook, uol. 8 (pp. 330-375). Newbury Park, CA: Sage. Breckler, S. J. (1984). Empirical validation of affect, behavior, and cognition as distinct components of attitude. ]ournal of Personality and Social Psychology, 47,1191-1205. Buck, R. (1984). The communication of emotion. New York: Guilford. Cantor, J. R. (1991). Fright responses to mass media productions. In J. Bryant & D. Zillman (Eds.), Responding to the screen: Reception and reaction processes (pp. 169197). Hillsdale, NJ: Lawrence Erlbaum. Cantor, J. R., Zillman, D., & Bryant, J. (1975). Enhancement of experienced sexual arousal in response to erotic stimuli through misattribution of unrelated residual excitation. Iournal of Personality and Social Psychology, 32,69-75. Cappella, J. N. (1991). The biological origins of automated patterns of human interaction. Communication Theory, I , 4-35. Cecil, J. S., Weiss, R. F., & Feinberg, R. A. (1978).The reinforcing effects of the recommendation in threatening communication. Journal of General Psychology, 98,65-77. Cooper, L. (1932). The rhetoric of Aristotle. New York: Appleton-Century-Crofts, Inc. Dabbs, J. M., & Leventhal, H. (1966). Effects of varying recommendations in a feararousing communication. Journal of Personality and Social Psychology, 4,525-531. Davidson, R. J. (1978). Specificity and patterning in biobehavioral systems. Implications for behavioral change. American Psychologist, 33,430-436. Dillard, J. P. (1990).The nature and substance of goals in tactical communication. In M. J. Cody & M. L. McLaughlin (Eds.), Psychology of tactical communication (pp. 7090). London: Multilingual Matters. Dollard, J., & Miller, N. E. (1950).Personality and psychotherapy. New York: McGrawHill. Dziokonski, W., & Weber, S. J. (1977). Repression-sensitization, perceived vulnerability, and the fear appeal communication. Journal of Social Psychology. 102,105-112. Edell, J. A., & Burke, M. C . (1987). The power of feelings in understanding advertising effects. Journal of Consumer Research, 14,421-433. Ekman, P. (1977). Biological and cultural contributions to body and facial movement. In J. Blacking (Ed.), Anthropology of the body (pp. 34-84). London: Academic Press. Ellsworth, P. C., & Smith, C . A. (1988).From appraisal to emotion: Differences among unpleasant feelings. Motivation and Emotion, 12,271-302. Evans, R. I., Rozelle, R. M., Lasater, T. M., Dembroski, T. M., & Allen, B. P. (1970). Fear arousal, persuasion, and actual versus implied behavioral change: A new perspective utilizing a real-life dental hygiene program. Journal of Personality and Social Psychology, 16,220-227. 318

Rethinking the Study of Fear Appeals

Faust, C. H., &Johnson, T. H. (1935).Jonathan Edwards. New York: American Book. Fishbein, M., & Ajzen, 1. (1975).Be/ief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA:Addison-Wesley. Fisher, W. A. ( 1 984). Predicting contraceptive behavior among university men: The role of emotions and behavior among university men: The role of emotions and behavioral intentions. Journal of Applied Social Psychology, 14,104-123. Folkman, S . , & Lazarus, R. S. (1990). Coping and emotion. In N. L. Stein, B. Leventhal, & T. Trabasso (Eds.), Psychological and biological approaches to emotion (pp. 313332). Hillsdale, NJ: Lawrence Erlbaum. Freimuth, V. S., Edgar, T., & Hammond, S. L. (1987). Colleges students’ awareness and interpretation of the AIDS risk. Science, Technology, and Human Values, 1 2 , 3 7 4 0 . Frijda, N. (1986). The emotions. Cambridge, UK: Cambridge University Press. Frijda, N., Kuipers, P., & ter Schure, E. (1989). Relations among emotion, appraisal, and emotional action readiness. Journal of Personality and Social Psychology, 57,212-228. Gardner, H. (1985). The mind’s new science: A history of the cognitive revolution. New York: Basic Books, Inc. Hale, J. L., & Mongeau, P. A. (1991, May). Testing a causal model of the persuasive impact of fear appeals. Paper presented at the annual meeting of the International Communication Association, Chicago. Higbee, K. L. (1970).Fifteen years of fear arousal: 1953-1968. Psychological Bulletin, 72, 426-444. Horowitz, 1. A. (1972). Attitude change as a function of perceived arousal. Journal of Social Psychology, 87, 117-126. Horowitz, I. A., & Gumenik, W. E. (1970). Effects of volunteer subject, choice, and fear arousal on attitude change. Journal of Experimental Social Psychology, 6,293-303. Hunter, J. E., & Schmidt, F. L. (1990). Methods of meta-analysis: Correcting error and bias in researchfindings. Newbury Park, CA: SAGE. Insko, C. A., Arkoff, A. & Insko, V. M. (1965). Effects of high and low fear-arousing communications upon opinions toward smoking. Journal of Experimental Social Psychology, 1,256-266. ken, A. M. (1987).Positive affect, cognitive processes, and social behavior. In L. Berkowitz (Ed.), Advances in experimental social psychology, vol. 20 (pp. 203-253). New York: Academic Press. Izard, C. E. (1971). The face of emotion. New York: Appleton-Century-Crofts. Izard, C. E., & Malatesta, C. Z. (1987). Perspectives on emotional development, I: Differential emotions theory of early emotional development. In J. D. Osofsky (Ed.), Handbook of infant development (2nd ed., pp. 494-554). New York: Wiley. Janis, I. L., & Feshbach, S. (1953). Effects of fear-arousing communications. Journal of Abnormal and Social Psychology, 48,78-92. Jepson, C., & Chaiken, S. (1990). Chronic issue-specific fear inhibits systematic processing of persuasive communications. Journal of Social Behavior and Personality, 5 , 61-84. Johnson, E. J., & Tversky, A. (1983). Affect, generalization, and the perception of risk. Journal of Personality and Social Psychology, 45,20-31. Kleinginna, P. R., & Kleinginna, A. M. (1981). A categorized list of emotion definitions, with suggestions for a consensual definition. Motivation and Emotion, 5,345-379. Kohn, P. M., Goodstadt, M. S., Cook, G. M., Sheppard, M., & Chan, G. (1982). Ineffectiveness of threat appeals about drinking and driving. Accident Analysis and Prevention, 14,457-464, Lacey, J. I. (1967). Somatic response patterning and stress: Some revisions of activation theory. In M. H. Appley & R. Trumbull (Eds.), Psychologicalstress: Issues in research (pp. 14-37). New York: Appleton-Century-Crofts. Laird, J. D. (1984).The real role of facial response in the experience of emotion: A reply to Tourangeau and Ellsworth, and others. Journal of Personality and Social Psychology, 47,909-917. Lazarus, R. S. (1990). Constructs of the mind in adaptation. In N. L. Stein, B. Leventhal, & T. Trabasso (Eds.), Psychological and biological approaches to emotion (pp. 3-19). Hillsdale, NJ: Lawrence Erlbaum. Lazarus, R. S. (1991).Emotion and adaptation. New York: Oxford Press. Levenson, R. W., Ekman, P., & Friesen, W. V. (1990).Voluntary facial action generates emotion-specific autonomic nervous system activity. Psychophysiology, 27,363-384. 319

Communication Theory

Leventhal, H. (1970). Findings and theory in the study of fear communications. In L. Berkowitz (Ed.), Advances in experimental social psychology, vol. 5 (pp. 119-186). New York: Academic Press. Leventhal, H. (1971). Fear appeals and persuasion: The differentiation of a motivational construct. American Journal of Public Health, 61,1205-1224. Leventhal, H., & Niles, P. (1964). A field experiment on fear-arousal with data on the validity of questionnaire measures. Journal of Personality, 32,459-479. Leventhal, H., & Scherer, K. (1987). The relationship of emotion to cognition: A functional approach to a semantic controversy. Cognition and Emotion, 1,3-28. Leventhal,H., & Singer, R. P. (1966). Affect arousal and positioning of recommendations in persuasive communications. Journal of PersoMlityandSocialPsychology, 4,137-146. Leventhal, H., Singer, R., & Jones, S. (1965). Effects of fear and specificity of recommendation upon attitudes and behavior. Journal of Personality and Social Psychology, 2 , 20-29. Leventhal, H., & Trembly, G. (1968). Negative emotions and persuasion. Journal of Personality, 36,154-168. Lewin, K. (1935).A dynamic theory ofpersonality. New York: McGraw-Hill. Lewis, M., Sullivan, M. W., & Michalson, L. (1984). The cognitive-emotional fugue. In C. E. hard, J. Kagan, & R. B. Zajonc (Eds.), Emotions, cognitions, 6 behavior (pp. 264-288). Cambridge, UK: Cambridge University Press. Matsumoto, D. (1987). The role of facial response in the experience of emotion: More methodological problems and a meta-analysis. Journal of Personality and Social Psychology, 52,769-774. Mauro, R. (1988).Opponent processes in human emotions? An experimental investigation of hedonic contrast and affective interactions. Motivation and Emotion, 12,333-351. McCuire, W. J. (1960). A syllogistic analysis of cognitive relationships. In C. I. Hovland & M. J. Rosenberg (Eds.), Attitude organization and change (pp. 65-111). New Haven, CT: Yale University Press. McGuire, W. (1968).Personality and attitude change: An information-processing theory. In A. G. Greenwaid, T. C. Brock, & T. M. Ostrom (Eds.), Psychological foundations ofattitudes (pp. 171-196). New York: Academic Press. McHugo, G. J., Lanzetta, J. T., Sullivan, D. G., Masters, R. D., & Englis, B. G. (1985). Emotional reactions to a political leader's expressive displays. Journal of Personality and Social Psychology. 49,1513-1529. Mewborn, C. R., & Rogers, R. W. (1979). Effects of threatening and reassuring components of fear appeals on physiological and verbal measures of emotion and attitudes. Journal of Experimental Social Psychology, 15,242-253. Meyerowitz, B. E., & Chaiken, S. (1987). The effect of message framing on breast selfexamination attitudes, intentions, and behaviors. Journal of Personality and Social Psychology, 52,500-510. Miller, G . R. (1963). Studies on the use of fear appeals: A summary and analysis. Central States Speech]oumal, 24,117-125. Miller, G . R., & Hewgill, M. A. (1966). Some recent research on fear-arousing message appeals. Speech Monographs, 33,379-391. Mongeau, P. A. (1990). Fear-arousing persuasive messages: A meta-analysis revisited. Paper presented at the annual meeting of the Western Speech Communication Association, Spokane. Newcomb, T. M., Turner, R. H., & Converse, P. E. (1965). Social psychology. New York: Holt, Rinehart, & Winston. Notarius, C. I., & Herrick, L. R. (1989).The psychophysiology of dyadic interaction. In H. Wagner & A. Manstead (Eds.), Handbook of social psychophysiology (pp. 393419). London: John Wiley & Sons. Oatley, K., & Johnson-Laird, P. N. (1987). Towards a cognitive theory of emotions. Cognition and Emotion, 1,29-50. O'Keefe, D. J. (1990). Persuasion: Theory and research. Newbury Park, CA: Sage. Petty, R. E., Gleicher, F., & Baker, S. M. (1991). Multiple roles for affect in persuasion. In J. Forgas (Eds.), Emotion and social judgments (pp. 181-200). Oxford, UK: Pergamon Press. Plutchik, R. (1980). Emotion: A psychoevolutionary synthesis. New York: Harper & Row. 320

Rethinking the Study of Fear Appeals

Powell, F. A., & Miller, G. R. (1967). Social approval and disapproval cues in anxietyarousing communications. Speech Monographs, 34,152-159. Ramirez, A., & Lasater, T. M. (1977). Ethnicity of communicator, self-esteem, and reactions to fear-arousingcommunications.Journal of Social Psychology, 102,79-91. Reisenzein, R., & Hofmann, T. (1990).An investigation of dimensions of cognitive appraisalin emotion using the repertory grid technique. Motivation and Emotion, 14,l-26. Robberson, M. R., & Rogers, R. W. (1988). Beyond fear appeals: Negative and positive persuasive appeals to health and self-esteem.Journal of Applied Social Psychology, 18, 277-287. Rogers, R. W. (1975). A protection motivation theory of fear appeals and attitude change. Journal of Psychology. 91,93-114. Rogers, R. W. (1983). Cognitive and physiological processes in fear appeals and attitude change: A revised theory of protection motivation. In J. Caaoppo & R. Petty (Eds.), Sociul psychophysiology (pp. 153-176). New York: Guilford. Rogers, R. W. (1985). Attitude change and information integration in fear appeals. Psychological Reports, 56,179-152. Rogers, R. W., & Deckner, C. W.(1975). Effects of fear appeals and physiologicalarousal upon emotion, attitudes, and cigarette smoking. Journal of Personality and Social Psychology, 32,222-230. Rogers, R. W., & Mewborn, C. R. (1976). Fear appeals and attitude change: Effects of a threat’s noxiousness, probability of occurrence, and the efficacy of coping responses. Journal of Personality and Social Psychology, 34,54-61. Roseman, I. J. (1984). Cognitive determinants of emotion: A structural theory. In P. Shaver (Ed.), Review of personality and social psychology, vol. 5 (p. 11-36). Berkeley, CA: Sage. Roseman, I. J., Spindel, M. S., & Jose, P. E. (1990). Appraisals of emotion-eliciting events: Testing a theory of discrete emotions. Journal of Personality and Social Psychology, 59,899-915. Rosen, T. J., Terry, N. S., & Leventhal, H. (1982). The role of esteem and coping in response to a threat communication. Journal of Research in Personality, 16, 90-107. Roser, C., & Thompson, M. E. (1991, May). Fear appeals and the formation of active publics. Paper presented at the annual meeting of the International Communication Association, Chicago. Scherer, K. R. (1981). Speech and emotional states. In J. Darby (Ed.), The evaluation of speech in psychiatry. New York: Grune & Stratton. Scherer, K. R. (1984). On the nature and function of emotion: A component process approach. In K. R. Scherer & P. Ekman (Eds.), Approaches to emotion (pp. 293-317). Hillsdale, NJ: Erlbaum. Scherer, K. R., Summerfield, A. B., & Walbott, H.G. (1983). Cross-national research on antecedents and components of emotion: A progress report. Social Science Information, 22,355-385. Schwartz, G. E., & Weinberger, D. A. (1980). Patterns of emotional responses to affective situations: Relations among happiness, sadness, anger, fear, depression, and anxiety. Motivation and Emotion. 4,175-191. Schwartz, S. H.(1973). Normative explanations of helping behavior: A critique, proposal, and empirical test. Journal of Experimental Social Psychology, 9,349-364. Schwarz, N., Servay, W., & Kumpf, M. (1985). Attribution of arousal as a mediator of the effectiveness of fear-arousing communications. Journal of Applied Sociul Psychology, 15,178-188. Self, C. A., & Rogers, R. W. (1990). Coping with threats to health: Effects of persuasive appeals on depressed, normal, and antisocial personalities.Journal of Behavioral Medicine, 13,343-357. Shelton, M. L., & Rogers, R. W. (1981). Fear-arousing and empathy-arousing appeals to help: The pathos of persuasion. Journal of Applied Social Psychology, 11,366-378. Siero, S., Kok, G., & Pruyn, J. (1984). Effects of public education about breast cancer and breast self-examination. Social Scientific Medicine. 18, 881-888. Skilbeck, C., Tulips, J., & Ley, P. (1977). The effects of fear arousal, fear position, fear exposure, and sidedness on compliance with dietary instructions. European ]ournu1 of Social Psychology, 7,221-239. 321

Communication Theory

Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in emotion. Journal of Personality and Social Psychology, 48,813-838. Smith, C. A., & Ellsworth, P. C. (1987). Patterns of appraisal and emotion related to taking an exam. Journal of Personality and Social Psychology, 52,475-488. Soames Job, R. F. (1988). Effective and ineffective use of fear in health promotion campaigns. American Journal of Public Health, 78,163-167. Solomon, R. (1980). The opponent-process theory of acquired emotion: The costs of pleasure and the benefits of pain. American Psychologist, 8,691-712. Sroufe, L. A. (1984). The organization of emotional development. In K. Scherer & P. Ekman (Eds.), Approaches to emotion (pp. 109-128). Hillsdale, NJ: Lawrence Erlbaum. Sternthal, B., & Craig, C. S. (1974). Fear appeals: Revisited and revised. lournal of Consumer Research, 1,22-34. Sutton, S . R. (1982). Fear-arousing communications: A critical examination of theory and research. In J. R. Eiser (Ed.), Social psychology and behavioral medicine (pp. 303337). London: John Wiley & Sons. Sutton, S. R. (in press). “Shock tactics” and the myth of the inverted U. British ]ournu1 of Addiction. Sutton, S . R., & Eiser, J. R. (1984). The effect of fear-arousing communications on cigarette smoking: An expectancy-value approach. Journal of Behavioral Medicine, 7 , 13-33. Sutton, S. R., & Hallett, R. (1989). The contribution of fear and cognitive factors in mediating the effects of fear-arousing communications. Social Behauiour, 4,83-98. Tesser, A., & Shaffer, D.R. (1990).Attitudes and attitude change. In L. W. Porter & M. R. Rosenzweig (Eds.), Annual review of psychology, uol. 41 (pp. 479-523). Palo Alto, CA: Annual Reviews, Inc. ter Horst, G., Hoogstraten, J., & de Haan, W. (1985). Stimulating dental attendance in the Netherlands: A comparison of three conceptual frameworks. Community Dentistry and Oral Epidemiology, 13, 136-139. Tolman, E.C. (1932). Purposive behauior in animals and men. New York: Century Co. Tomkins, S. S. (1981).The quest for primary motives: Biography and autobiography of an idea. Journal of Personality and Social Psychology, 41,306-329. Tyler, T. R., & Rasinski, K. (1984). Comparing psychological images of the social perceiver: Role of perceived informativeness, memorability, and affect in mediating the impact of crime victimization. Journal of Personality and Social Psychology, 46, 308329. Vroom, V. H. (1964). Work and motivation. New York: John Wiley. Wagner, H. (1989). The peripheral physiological differentiation of emotions. In H. Wagner & A. Manstead (Eds.), Handbook of social psychophysiology (pp. 77-98). London: John Wiley. Watson, M., Pettingale, K. W., & Goldstein, D.(1983).Effects of fear appeal on arousal, self-reported anxiety, and attitude toward smoking. Psychological Reports, 52, 139146. Weiner, B. (1986). An amibutional theory of motivation and emotion. New York: Springer-Verlag. Weinstein, N. D.(1993). Testing four competing theories of health-protective behavior. Health Psychology, 12,324-333. Williams, R. E., Ward, D.A., & Gray, L. N. (1985). The persistence of experimentally induced cognitive change: A neglected dimension in the assessment of drug prevention programs. Journal of Drug Education, 15, 33-42. Witte, K . (1992a).The role of threat and efficacy in AIDS prevention. International Quarterly of Community Health Education, 12,225-249. Wine, K. (1992b). Pumng the fear back into fear appeals: The extended parallel process model. Communication Monographs, 59,329-349. Wolf, S., Gregory, W. L., & Stephan, W. G. (1986).Protection motivation theory: Prediction of intentions to engage in anti-nudear war behaviors. Journal of Applied Social Psychology, 16,310-321. Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American Psychologist, 35,15 1-175. Zanna, M.P., & Rempel, J. K. (1988). Attitudes: A new look at an old concept. In D. 322

Rethinkingthe Study of Fear Appeals

Bar-Tal & A. Kruglanski (Eds.), The social psychology of knowledge (pp. 315-334). Cambridge, UK: Cambridge University Press. Zillman, D. (1988). Mood management: Using entertainment to full advantage. In L. Donohew, H. E. Sypher, & E. T. Higgins (Eds.), Communication, social cognition, and affect(pp. 141-171). Hillsdale, NJ: Lawrence Erlbaum.

323