Some Issues Surrounding Research Effects of &Quot;Feeling Advertisements&Quot;

Andrew A. Mitchell, University of Toronto
ABSTRACT - It is argued that our current models of advertising effects are inadequate for explaining the effect of feeling advertisements. A number of theoretical issues concerning affective states-are discussed as well as a number of issues concerning alternative hypotheses of how 'feeling' advertisements may work and possible boundary conditions on their effects.
[ to cite ]:
Andrew A. Mitchell (1986) ,"Some Issues Surrounding Research Effects of &Quot;Feeling Advertisements&Quot;", in NA - Advances in Consumer Research Volume 13, eds. Richard J. Lutz, Provo, UT : Association for Consumer Research, Pages: 623-628.

Advances in Consumer Research Volume 13, 1986      Pages 623-628

SOME ISSUES SURROUNDING RESEARCH EFFECTS OF "FEELING ADVERTISEMENTS"

Andrew A. Mitchell, University of Toronto

ABSTRACT -

It is argued that our current models of advertising effects are inadequate for explaining the effect of feeling advertisements. A number of theoretical issues concerning affective states-are discussed as well as a number of issues concerning alternative hypotheses of how 'feeling' advertisements may work and possible boundary conditions on their effects.

INTRODUCTION

There has been increased interest recently in understanding the role of emotion, feelings and affect on human behavior in both psychology (e.g. Bower 1981; Zajonc 1980) and consumer behavior (e.g. Holbrook 1985; Kassarjian 1985). To some extent, this increased interest may be viewed as a reaction against strict information processing models which implicitly (or explicitly) view human behavior to be largely based on the processing and interpretation of verbal information. Humans, however, also experience various affective states such as happiness and sorrow, so these states would also seem to comprise a critical element of human behavior. Unfortunately, there is no general consensus as to the functional role of these affective states. Simon (1967), for instance, has argued that emotions should be regarded as an interrupt in the human information processing system, while others have argued that affective states directly affect the processing of information (e.g. Bower 1981; Isen 1984). Others view affective states as a behavioral homeostatic process as well as a fundamental signaling system (e.g. Plutchik 1984).

In studying affective states, there are two distinctly different approaches that may be taken. First, affective states may be treated as a phenomenon in and of itself. Research using this approach is directed at developing a taxonomy or categorization of different affective states (e.g. Tomkins 1982) or at understanding what variables cause the various affective states to be evoked (e.g., Weiner 1982; Roseman 1983). In this latter case, affective states are treated as a dependent variable. Alternatively, research say be directed at understanding the effect of affective states on some other human activity. For instance, does an individual's affective state effect decision making? Under this approach affective states are treated as an independent variable.

In this paper, we will primarily be concerned with the second approach. More specifically, we will be concerned with the effect of affective states within an advertising context. However, as we will argue in this paper, we will need to have some theoretical understanding of affective states in order to understand their possible effects. Finally, in this paper the term affective state will be used in a generic sense to include all emotional and feeling states. Flood specifically will be viewed as a less intense form of emotion (e.g. Isen 1984).

The issue of how affective states within an advertising context may affect consumer behavior seems to be an important one since we are faced with a basic contradiction. Most of our current theories of how advertising works are strict information processing motels. In these motels, consumers must make either verbal responses to advertisements (support arguments) or acquire knowledge about the advertised brand in order to change attitudes which, in turn, are viewed as the primary (only) mediator of behavior. On the other hand, a very large proportion of the advertisements that appear on television and in magazines contain little product attribute information. Many seem to be designed simply to create an emotional reaction in the consumer. Following Zielski (1982), we will call these 'feeling' advertisements. In doing so, we wish to differentiate them from advertisements which contain emotional appeals. The effects of these latter types of advertisements presumably can be explained with our current models.

In order to explain the effect of 'feeling' advertisements with our current motels, we have to either (1) assume that individuals make considerable verbal responses to these types of advertisements, or (2) assume that these advertisements don't have any affect on consumer behavior. Alternatively, we can recognize that our current models do not adequately explain the effects of these types of advertisements. At this particular point in time, this latter alternative seems to be the most appropriate since there is both anecdotal evidence that 'feeling' advertising campaigns may increase sales and evidence from laboratory research that the affective states of individuals have an affect on attitude formation in an advertising context. Consequently, the first and second assumptions do not seem to be teniable.

THEORETICAL ISSUES

Research on moods and emotions and theories of emotions remains a rather murky area in psychology, both conceptually and in terms of empirical results. The purpose of this section is to briefly review some theoretical issues concerning affective states and some critical empirical results that are the most relevant with respect to developing a model that will explain the effects of 'feeling' advertisements. This review is not meant to examine all the critical conceptual issues surrounding emotion nor to necessarily cover all the various theories of emotion. In this discussion, we will focus primarily on theories and research that attempt to integrate emotion and moot into our current understanding of the human information processing system. Since we have good understanding of this system, it would seem the most fruitful to understand how moods and emotions may affect this system instead of attempting to develop a new model of how 'feeling' advertising may work. In addition, we will only be concerned with how 'feeling' advertisements may work given exposure to the advertisement. It should be noted, however, that other researchers have suggested that consumers may be more likely to attend to these types of advertisements (e.g. Ray and Batra 1983).

The first issue that will be discussed and certainly one of the most critical concerns the conceptualization and measurement of emotion. At least some of the confusion surrounding emotions occurs because it has been measured in many different ways. For instance, emotional states have been measured by using facial expressions (e.g. Ekman 1984), specific behaviors (e.g. Schachter and Singer 1962) and physiological measures (e.g. Lazarus et.al. 1965).

Most emotion theorists agree that an emotion has four different elements: (a) subjective feelings, (b) physiological arousal, (c) expressive behavior (e.g. facial expression) and (d) a behavioral response (e.g. shouting at someone). Following Plutchik (1982), we would like to conceptualize emotion as a hypothetical construct and the different elements as possible measures of the hypothetical construct. Since it is well known that individuals can mask their true emotional state with their expressive behavior and their behavioral response, and may not correctly report their subjective feelings, measures of affective states based on these three elements will probably contain measurement error and may, in some cases,

There are also problems with physiological measures. First, cognitive-arousal theories of emotion (e.g. Mandler 1975, 1982; Schachter and Singer 1967) indicate that physiological arousal is a necessary, but not a sufficient condition for someone to be in an affective state. In other words, someone may be in a state of physiological arousal, but not in an emotional state. Second, the empirical evidence indicates that the relationship between physiological arousal and emotion does not appear to be as strong as once thought (e.g. Reisenzein 1983). Finally, there is considerable disagreement as to the type of arousal that accompanies an emotion. Mandler (1985), for instance, argues that it is automatic arousal while others argue that cortical arousal must also be considered in emotional reactions (e.g. Izard 1982). Of all these possible measures, then, measures of subjective feeling states, would seem to be the most appropriate as long as there is no reason for subjects to misrepresent their feelings. In addition, this review suggests that physiological measures should only be used in conjunction with measures of subjective feeling states.

A second issue concerns emotion as motivation or drive. The behavioral school viewed emotion as a drive or a motivator to action. According to this view, affective states, with the accompanying increase in arousal, motivated the individual into action. For instance, a state of anger, would motivate the individual to strike at the target of the anger. In the persuasion literature, this view of emotion resulted in the fear drive model of Janis and Feshbach (1953). Under this motel, increases in the level of fear should provide motivation to reduce the cause of the fear. For instance, increasing the fear associated with lung cancer should increase the motivation to stop smoking. As is well known, however, experimental research indicated that in some situations high levels of fear may be less effective than low levels of fear. Consequently, an inverted U shape relationship was hypothesized to exist between fear levels and behavioral compliance .

A number of experimental findings within the fear appeal literature has cast doubts on this model of emotion as drive or motivation. For instance, most studies have been unable to find an interaction between level of fear and efficacy of suggested behavior (e.g. Mewborn and Rogers 1979; Leventhal 1970). If fear acts as a motivator, then higher levels of fear should cause more behavioral compliance when the suggested behavior is efficacious. In addition, no effects on behavioral compliance have been found when the fear appeal precedes or follows the message suggesting the appropriate behavior for reducing the cause of the fear (e.g. Leventhal and Singer 1966). If fear acts as a drive, one would also expect more behavioral compliance when the fear appeal precedes the suggested behavior. These findings have led Leventhal (1970) to propose an alternative to the fear-drive model, his Parallel Response Model. According to this model, individuals make two parallel responses to a fear appeal message. The first is a response to the cause of the fear (e.g. lung cancer) while the second is a response to the emotion or fear itself. When the level of fear is low, individuals devote most of their energy to reducing the cause of the fear and behavioral compliance is high. However, when the fear is high, individuals focus on strategies for coping with or reducing the level of fear and little energy to the cause of the fear, so behavioral compliance is low. Consequently, this model also results in an inverted U shaped relationship between level of fear and behavioral compliance.

The view of emotion as drive or motivation has also declined in most theories of emotion, but this is probably due more to the decline of the behaviorist school in psychology as opposed to contrary empirical evidence. Exceptions are Tomkins (1984), who views emotions as enhancing the behavioral response of the individual and Buck (1985) who has recently presented a model that integrates both motivation and emotion.

The third issue concerns the number of different emotions and the relationship between emotions. As might be expected there is no general agreement as to the number of different emotions that exist or, for that matter, a precise definition of emotion and now it differs from other psychological constructs. There seems to be general agreement concerning the fundamental emotions of happiness, sadness, fear, anger, interest, contempt, disgust, surprise, shame and guilt (e.g. Izard 1977; Tomkins 1982), however, other theorists have suggested that others such as pride and hope may also exist (e.g. Lazarus et.al. 1980).

A related issue concerns the structural relationship between affective states. A number of studies using a number of different methodologies have examined this issue. In most of these studies, subjects evaluated either facial expressions of subjective feeling states along a number of dimensions. This data was then factor analyzed or submitted to a multidimensional scaling program. Although the number of dimensions found varies across studies, the two dimensions of pleasantness and level of activation seem to consistently appear across studies (e.g. Frijda 1969; Russell 1980). Other researchers, focusing on how the appraisal of the environment affects emotional response, have suggested additional dimensions of attentional activity, control, depth of experience, legitimacy and certainty (e.g. Roseman 1984; Weiner 1982). A recent empirical study by Smith and Ellsworth (1985) which attempted to control for methodological problems in previous empirical studies, found five dimensions underlie different emotional states. These are pleasantness, responsibility/ control, certainty, amount of attention, amount of effort and situational/control. The results of these studies are important because we need to understand what dimensions underlie different affective states in order to produce these states with advertising or in the laboratory. As will be mentioned later, most of the current research examining the effect of affective states in an advertising context have examined affective states along the pleasantness dimension only.

The fourth issue concerns the extent to which affective states affect cognitive processing. As mentioned earlier, one approach to understanding the effect of affective states is to understand how these states may affect the processing of information. Currently, there is a growing literature on these effects. For instance, affective states have been found to affect decision making (e.g. Isen and Means 1983), the categorization of ambiguous stimuli (e.g. Isen and Daubman 1984), perceptions of risk (Johnson and Tversky 1983), the helping of others (e.g. Isen and Levin 1972; Moore Underwood and Rosenhan 1973) and the evaluation of objects (e.g. Isen et.al. 1978). In addition, a number of studies have found asymmetric effects due to pleasant and unpleasant affective states (e.g. Isen 1984).

What is less clear is why these effects occur. Here a number of different hypotheses have been proposed (e.g. Isen 1984). One is a retrieval hypothesis. Here it is hypothesized that an individual's affective state will effect what information is retrieved from memory. For instance, a number of studies indicate that when individuals are in a pleasant affective state, they are more likely to retrieve positively toned information (e.g. Teasdale and Fogerty 1979). The second is an encoding hypothesis. Here it is hypothesized that the affective state of the individual effects what information is encoded in memory. For instance, individuals are more likely to encode positively toned material when they are in a pleasant affective state (e.g. Bower et.al. 1981). However, direct tests of these hypotheses have yielded conflicting results. A further clouting of the issue is caused by the use of different manipulations in the studies to create affective states, so there is some concern that these different manipulations may confound the results. For instance, Bower et.al. (1981) used hypnosis to induce an affective state and found an encoding effect, while Hasher et.al. (1985), use naturally occurring affective states and found no effects.

In summary, the issue here is, if different affective states cause differences in evaluations and judgments, to what extent can these differences be explained by differences in cognitive processes. For instance, if subjects in a pleasant affective state form more positive evaluations of products, to what extent are these differences caused by differences in the encoding or retrieval of more positive product related information.

The final issue concerns the representation of emotions in memory. There are a number of empirical studies that indicate that memory for affective states persist over time. For instance, Bower (1981) has found evidence for state dependent learning. Individuals that learn information in a particular affective state will retrieve more of the learned information if they are in the same affective state at retrieval.

The current, though somewhat controversial approach to the representation problem is that affective states may be represented as a node in a semantic network (Clark and Isen 1981; Clark 1982; Bower 1981). An extension of this model is that the arousal state of the individual is also represented in memory. Clark, Milberg and Ross (1983), present empirical evidence supporting this proposition.

The main argument against this representation is that the number of linkages between a particular affective node and all the related nodes would be large and if a spreading activation model is used to explain information retrieval, there would be only a very small probability that a particular linked node would be activated. This would make the finding of state dependent learning very unlikely. Given the robustness of semantic network models of memory, however, one could always create refinements in the model that could handle these objections. For instance, one might hypothesize some other type of retrieval process for information linked to an affective node or one might argue that there are many different nodes for a particular emotion that may be differentiated by other contexts that occurred at learning.

Although there are a number of critical unresolved issues in psychology surrounding both the conceptualization of affective states and their effects on information processing, we should not let this deter us in investigating the effects of 'feeling' advertisements. There are a number of critical issues that can be examined given the present state of theoretical development on affective states. These will be discussed in the next section.

ADVERTISING ISSUES

In this section, a number of theoretical issues concerning the effect of affective states within an advertising context will be discussed. In general, there is increasing evidence that the affective state of individuals affects their responses to advertising messages (e.g. Milberg and Mitchell 1985; Srull 1984). The issues that will be discussed in this section focus on understanding the boundaries and causes of these effects.

The first issue is whether or not these effects can be explained by differences in cognitive processes. For instance, a number of studies indicate that individuals form more positive evaluations of advertised brands when they are in a pleasant affective state (e.g. Srull 1984). These effects might be explained by either the encoding and retrieval hypotheses. If individuals are in a pleasant affective state during exposure to an advertisement they may encode more favorable information about the advertised brand which. in turn. may cause them to evaluate the advertised brand more favorably. Alternatively, if the retrieval hypothesis is valid, then individuals in a pleasant affective state when evaluating a brand would retrieve more positive information which would also cause them to evaluate the advertised more favorably. Obviously, the encoding hypothesis has the most relevance for the advertiser since advertisers have some control over consumers' affective states through the design of advertisements and the placement of advertisements in specific television shows. In addition, if the encoding hypothesis is valid, it suggests that 'feeling' advertisements will have an effect only if the advertisement contains product information.

In a study designed to examine the validity of the encoding and retrieval hypotheses, Milberg and Mitchell (1984) manipulated the affective state of subjects at both encoding and retrieval. In addition, each product message contained both positive and negative information. The results of the study indicated very strong effects due to the affective state of the subject, however, no differences were found in the valence of the information recalled across conditions. These results suggest that the encoding and retrieval hypotheses cannot explain the attitudinal differences that were found, so that we must now look elsewhere for an explanation of these effects.

The second issue is whether attitude toward the advertisement (e.g. Mitchell and Olson 1981) can explain these effects. First, it might be expected that advertisements that create a pleasant affective state will be liked more than advertisements that create a neutral or unpleasant affective state. If this is true, then attitude toward the advertisement may explain advertisement induced affective state effects. This of course, needs to be demonstrated empirically. An alternative hypothesis is that affective states may produce a response bias. Whenever individuals are in a pleasant affective state they may simply evaluate everything more positively. Consequently, both attitude toward the advertisement and brand attitudes would both be evaluated more positively.

There are a number of ways of testing these rival hypotheses. First we might be able to find some experimental stimuli that could be used in an advertisement (e.g., music) that would have differential effects on the valence of the affective state and the evaluation of the advertisement. For instance, we might be able to find music that would create pleasant affective states, but would cause a neutral evaluation of the advertisement. Alternatively, we might create two experimental conditions where an experimental stimuli is used to induce an affective state prior to and during an advertising message. If attitude toward the advertisement were the same in both conditions, a response bias explanation would seem appropriate.

The third issue concerns the stability of these effects. All the studies that have found that affective states affect the evaluations of the advertised brand have taken measures of brand attitudes shortly after the affective state manipulations. Consequently, although these effects occur in the laboratory, they may disappear shortly after subjects leave the laboratory.

One boundary condition that may affect the stability of these effects is whether or not the subject forms an evaluation of the brand during exposure to the advertisement. Srull (1984), for instance, manipulated whether or not subjects formed an evaluation of an advertised brand during exposure to an advertisement. In this study, pleasant, neutral and unpleasant affective states were induced prior to seeing the advertising message The results of this study indicated that pleasant and unpleasant affective states had an effect on brand attitudes only when subjects formed an evaluation of the advertised brand during exposure to the advertisement.

The fourth issue is whether affective states have an effect on brand attitudes in an attitude change situation. All of the recent research that has obtained affective state effects on brand attitudes is an attitude formation situation. Previous research (e.g. Janis et.al. 1965) has found attitude change effects with different affective states, however, these studies contain methodological flaws. Consequently, it still is necessary to demonstrate unambiguously that affective states can have an effect in an attitude change situation.

A fifth issue concerns possible differences in effects if the affective state is induced prior to seeing the advertising message or during the message. These differences would be analogous to whether the affective state was induced by the program content prior to the advertisement or whether the affective state was induced by the advertisement. As mentioned previously, Srull (1984) found that when the affective state is induced prior to receiving the advertising message, the affective state has an effect only when subJects form an evaluation of the brand during exposure to the message. It is hypothesized here, that if the affective state is induced during exposure to the message, it will have an effect regardless of whether or not a brand attitude is formed during exposure to the advertisement. Under these conditions, the affective state will become linked to the brand in memory. Consequently, when this attitude is formed at a later point in time, the affective state will still have an effect. As the time period between exposure to the advertisement and the formation of the attitude increases, however, these effects might be expected to weaken since it is generally believed that affective states induced by recalling a particular event where that affective state was present, will be weaker than the affective state that occurred at the time of the event

The sixth issue concerns the effect of different types of affective states. All of the previous research examining the effects of affective states on brand attitudes in an advertising context has induced either a pleasant or unpleasant affective state. However, as was discussed earlier, pleasantness is only one of the dimensions that has been found to underlie different affective states. It may well be that other dimensions of affective states will also have an effect on brand attitudes, purchase behavior, or the evaluation of a product usage experience. This will be more likely to occur if it is found that affective states formed during exposure to an advertisement continue to have an effect even if the subject did not form an evaluation of the brand during exposure to the advertisement. It is also hypothesized that if these other dimensions have an effect, they are most likely to occur at the point of purchase or in the evaluation of the usage situation. In addition, it is hypothesized that if the different dimensions of affective states have an effect. these effects will be very product class specific.

The final issue concerns the differential effects of attitudes with a strong cognitive base and attitudes created through pure affect. Theoretically, we can now construct the same attitudes, in terms of valence, based on either product information or affective reactions. These latter reactions may be created either through the creation of an affective state during exposure to the advertisement or by using affect inducing non-informative photographs in the advertisements (e.g. Mitchell 1985). The issues here involve potential differences in the stability of the attitudes that are formed and differences in the attitude behavior relationship. Here, it might be hypothesized that attitudes with an informational base will be more stable, but have a weaker attitude behavior relationship.

DISCUSSION

In this paper, I have provided reasons why it is important to understand the effect of 'feeling' advertisements and briefly outlined some important theoretical issues concerning affective states that are critical in developing theoretical models of how 'feeling' advertisements may work. In addition, a number of important issues were presented surrounding the effects of 'feeling' advertisements.

A number of problems should also be pointed out that may occur in conducting research in this area. First, it is difficult to develop pure experimental manipulations of affective states. Most manipulations of affective states will also have some other effect on the information processing system. For instance, the use of music in an advertisement to induce a particular affective state may also act as a distractor. This means that there will be alternative explanations for the results that are found. Consequently, it is important to design tightly controlled experiments or use converging operations to rule out alternative explanations (e.g. Gartner, Harke and Erickson 1956). Second, it may be difficult to develop pure manipulations of different affective states since most theories of emotion are cognitive. This means that we need to create different cognitions in order to create different affective states. Consequently, if different effects are found, it will be difficult to determine what is causing the effect, the different cognition or the different affective states. The final problem is that individuals may experience many different emotions at the same tine (e.g. Ekman 1984). This will probably be especially true if we use complex stimuli such as advertisements. If we want to investigate the effect of specific affective states with advertising, we will probably need to create advertisements to make sure that we are inducing only the affective state that is of interest. It also means that if we want to measure affective states induced by actual advertisements, we will need to use measurement procedures that are able to measure combinations of emotions.

REFERENCES

Averill, J.R. (1980), "A Constructivist's View of Emotion." In R. Plutchik and H. Kellerman (Eds.) Emotion: Theory, Research and Experience, New York: Academic Press.

Bower, G.H. (1981), "Mood and Memory," American Psychologist, 36, 129-43.

Bower, G.R. and P.R. Cohen (1982), "Emotional Influences in Memory and Thinking." In M.S. Clark and S.T. Fiske (Eds.) Affect and Cognition, Hillsdale, N.J.: Lawrence Erlbaum Associates.

Bower, G.H., S.G. Gilligan and R.P. Monteiro (1981), "Selectivity of Learning Caused by Affective States," Journal of Experimental Psychology: General, 110, 451-73.

Buck, R. (1985), "Prime Theory: An Integrated View of Motivation and Emotion," Psychological Review, 92, 389-413.

Clark, M.S. (1982), "A Role for Arousal in Explaining the Link Between Feelings States, Judgments and Behavior." in M.S. Clark and S.T. Fiske (Eds.) Affect and Cognition, Hillsdale, N.J.: Lawrence Erlbaum Associates.

Clark, M.S. and A.M. Isen (1982), "Toward Understanding the Relationship Between Feeling States and Social Behavior." In A. Hastorf and A.M. Isen (Eds.) Cognitive Social Psychology, New York: Elsevier.

Clark, M.S., S. Milberg and J. Ross (1983), "Arousal Cues, Arousal Related Material in Memory: Implications for Understanding Effects of Mood on Memory," Journal of Verbal Learning and Verbal Behavior, 22, 633-49.

Ekman, P. (1984), "Expression and the Nature of Emotion. In K.R. Scherer and P. Ekman (Eds.) Approaches to Emotion. Hillsdale, N.J.: Lawrence Erlbaum Associates.

Ekman, P., R.W. Levenson, and W.V. Friesen (1983), "Autonomic Nervous System Activity Distinguishes Among Emotions," Science, 221, 1208-10.

Frijda, N.H. (1969), "Recognition of Emotion." In L. Berkowitz (Ed.) Advances in Experimental Social Psychology, Vol. 4, New York: Academic Press, 167-223.

Gardner, W.R., H.H. Harke and C.W. Erickson (1956), "Operationalism and the Concept of Perception," Psychological Review, 63, 149-59.

Gorn, G.J. (1982), "The Effects of Music in Advertising on Choice Behavior: A Classical Conditioning Approach," Journal of Marketing, 46, 94-101.

Hasher, L., R.C. Rose, R.T. Zacks, H. Sanft and B. Dorin (1985), "Mood, Recall, and Selectivity Effects in Normal College Students," Journal of Experimental Psychology: General, 114, 104-18.

Holbrook, M.R. (1985), "Emotion in the Consumption Experience: Toward A New Model of the Human Consumer." In R.A. Peterson, W.D. Hoyer and W.R. Wilson (Eds.) The Role of Affect in Consumer Behavior: Emerging Theories and Applications, Lexington, M.A.: D.C. Heath and Company.

Holbrook, M.R. and J. O'Shaughnessy (1984), "The Role of Emotion in Advertising," Psychology and Marketing, 1, 45-64.

Isen, A. (1984), "Toward Understanding the Role of Affect in Cognition." In R.S. Wyer, Jr. and T.K. Srull (Eds.) Handbook of Social Cognition, Vol. 3, Hillsdale, N.J.: Lawrence Erlbaum Associates.

Isen, A.M. and R.A. Daubman (1984), "The Influence of Affect on Categorization," Journal of Personality and Social Psychology, 47, 1206-17.

Isen, A.M. and P.F. Levin (1972), "The Effect of Feeling Good on Helping: Cookies and Kindness," Journal of Personality and Social Psychology, 21, 384-88.

Isen, A.M. and B. Means (1983), "Positive Affect as a Variable in Decision Making," Social Cognition, 2, 18-31.

Isen, A.M., T.E. Shalker, M.S. Clark and L. Karp (1978), "Positive Affect, Accessibility of Material in Memory, and Behavior: A Cognitive Loop?" Journal of Personality and Social Psychology, 36, 1-12.

Izard, C.E. (1977), Human Emotions. New York: Plenum Press.

Izard, C.E. (1982), "Comments on Emotion and Cognition: Can There be a Working Relationship?" In M.S. Clark and S.T. Fiske (Eds.) Affect and Cognition, Hillsdale, N.J.: Lawrence Erlbaum Associates.

Janis, I.L. and S. Feshbach (1953), "Effects of Fear Arousing Communications" Journal of Abnormal and Social Psychology, 48, 78-92.

Janis, I.L., D. Kaye and P. Kirschner (1965), "Facilitating Effects of 'Eating While Reading' on Responsiveness to Persuasive Communications," Journal of Personality and Social Psychology, 11, 181-6.

Johnson and Tversky (1983), "Affect, Generalization, and the Perception of Risk," Journal of Personality and Social Psychology, 45, 20-31.

Kassarjian, H.H. (1985), "Consumer Research: Some Recollections and a Commentary," Fellows' Address, Association for Consumer Research Conference. Las Vegas. NE.

Lazarus, R.S. (1982), "Thoughts on the Relations Between Emotion and Cognition," American Psychologist, 37, 10-19.

Lazarus, R. S., J.C. Coyne and S. Folkman (1984), "Cognition, Emotion and Motivation: The Doctoring of Humpty-Dumpty." In K.R. Scherer and P. Ekman (Eds.) Approaches to Emotion, Hillsdale, N.J.: Lawrence Erlbaum Associates.

Lazarus, R.S., A.D. Kanner and S. Folkman (1980), "Emotions: A Cognitive-Phenomenological Analysis." In R. Plutchik and H. Kellerman (Eds.) Emotion: Theory, Research and Experience, Vol. 1, New York: Academic Press.

Lazarus, R.S., E.M. Opton, M.S. Nomikes and N.O. Rankin (1965), "The Principle of Short-Circuiting of Threat: Further Evidence," Journal of Personality, 33, 622-35.

Leventhal, H. (1980), "Toward a Comprehensive Theory of Emotions." In L. Berkowitz (Et.) Advances in Experimental Social Psychology, Vol. 13, New York: Academic Press.

Leventhal, H. (1970), "Findings and Theory in the Study of Fear Communications." In L. Berkowitz (Et.) Advances in Experimental Social Psychology, Vol. 5, New York: Academic Press.

Leventhal, H. and R.P. Singer (1966), "Affect Arousal and Positioning of Recommendations in Persuasive Communications," Journal of Personality and Social Psychology, 4, 137-46.

Mandler, G. (1982), " a e Generation of Emotion: A Psychological Theory." In R. Plutchik and H. Kellerman (Eds.) Emotion: Theory, Research and Experience, New York: Academic Press.

Mandler, G. (1975), Mint and Emotion, New York: John Wiley and Sons.

Mewborn, C.R. and R.w. Rogers (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, 241-53.

Milberg, S. and A.A. Mitchell (1984), "An Examination of Alternative Hypotheses of How Moot May Affect Brand Attitudes and Intentions." Paper presented at the American Psychological Association Conference in Toronto, Ontario.

Mitchell, A.A., "The Effect of Visual and Verbal Components of Advertisements on Brand Attitudes and Attitude Toward the Advertisement." Working Paper, Faculty of Management Studies, University of Toronto, Toronto, Ontario M5S 1V4.

Mitchell, A.A. and J.C. Olson (1981), "Are Product Attribute Beliefs the Only Mediator of Advertising Effects on Brand Attitudes?" Journal of Marketing Research, 18, 318-22.

Moore, B.S., B. Underwood and D.L. Rosenhan (1973), "Affect and Altruism," Developmental Psychology, 8, 99-104.

Plutchik, R. (1984), "Emotions: A General Psychoevolutionary Theory." In t.R. Scherer and P. Ekman (Eds.) Approaches to Emotions, Hillsdale, N.J.: Lawrence Erlbaum Associates.

Plutchik, R. (1982), "A General Psychoevolutionary Theory of Emotion." In R. Plutchik and H. Kellerman (Eds.) Emotion: Theory, Research, and Experience, New York: Academic Press.

Plutchik, R. (1980), Emotion: A Psychoevolutionary Synthesis, New York: Harper and Row.

Ray, M.L. and R. Batra (1983), "Emotion and Persuasion in Advertising: What We Do and Don't Know About Affect." In R.P. Bagozzi and A.M. Tybout (Eds.) Advances in Consumer Research, Vol. 10, Ann Arbor: Association for Consumer Research.

Reisenzein, R. (1983), "The Schachter Theory of Emotion: Two Decades Later," Psychological Bulletin, 94, 239-64.

Roseman, I. (1984), "Cognitive Determinants of Emotions: A Structural Theory." In P. Shaver (Et.) Review of Personality and Social Psychology: Vol. 5, Emotions, Relationships and Health, Beverly Hills, C.A.: Sage.

Russell, J.A. (1980), "A Circumplex Model of Affect," Journal of Personality and Society Psychology, 39, 1161

Schachter, S. and J.E. Singer (1962), "Cognitive, Social and Physiological Determinants of Emotional State," Psychological Reviews, 69, 379-99.

Simon, H.A. (1967), "Motivational and Emotional Controls of Cognition," Psychological Review, 74, 29-39.

Smith, C.A. and P.C. Ellsworth, "Patterns of Cognitive Appraisal in Emotion," Journal of Personality and Social Psychology, 84, 813-38.

Srull, R. (1984), "Effects of Subjective Affective States on Memory and Judgment." In T. Kinnear (Ed.) Advances in Consumer Research, Vol. 11, Ann Arbor: Association for Consumer Research

Teasdale, J.D. and S.J. Fogerty (1979), "Differential Effects of Induced Moot on Retrieval of Pleasant and Unpleasant Events from Episodic Memory," Journal of Abnormal Psychology, 88, 248-57.

Tomkins, S.S. (1984), "Affect Theory." In R.R. Scherer and P. Ekman (Eds.) Approaches to Emotion. Hillsdale, N.J.: Lawrence Erlbaum Associates.

Tomkins, S.S. (1982), "Affect Theory." In P. Ekman (Ed.) Emotion in the Human Face. New York: Cambridge University Press, 353-95.

Weiner, B. (1982), "The Emotional Consequences of Causal Attributions." In M.S. Clark and S.T. Fiske (Eds.) Affect and Cognition, Hillsdale, N.J.: Lawrence Erlbaum Associates.

Zajonc, R.B. (1984), "The Interaction of Affect and Cognition." In R.R. Scherer and P. Ekman (Eds.) Approaches to Emotion, Hillsdale, N.J.: Lawrence Erlbaum Associates.

Zajonc, R.B. (1980), "Feeling and m inking: Preferences Need No Inferences," American Psychologist, 35, 151-75.

Zajonc, R.B. and H. Markus (1982), "Affective and Cognitive Factors in Preferences," Journal of Consumer Research, 9, 123-31.

Zielski, H.A. (1982), "Does Day After Recall Penalize 'Feeling' Ads?" Journal of Advertising Research, 22, 19-24.

----------------------------------------