Informational and Transformational Advertising: the Differential Effects of Time

Christopher P. Puto, Duke University
William D. Wells, Needham, Harper and Steers Advertising, Inc.
ABSTRACT - This paper presents formal definitions of informational and transformational advertising, together with the theoretical foundations for each. A series of theoretical propositions outlining the differential effects of repeated exposures over time on each type of ad are given, scales for measuring the informational and transformational content of ads are proposed and empirically validated, and suggested areas for further research are discussed.
[ to cite ]:
Christopher P. Puto and William D. Wells (1984) ,"Informational and Transformational Advertising: the Differential Effects of Time", in NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT : Association for Consumer Research, Pages: 638-643.

Advances in Consumer Research Volume 11, 1984      Pages 638-643

INFORMATIONAL AND TRANSFORMATIONAL ADVERTISING: THE DIFFERENTIAL EFFECTS OF TIME

Christopher P. Puto, Duke University

William D. Wells, Needham, Harper and Steers Advertising, Inc.

ABSTRACT -

This paper presents formal definitions of informational and transformational advertising, together with the theoretical foundations for each. A series of theoretical propositions outlining the differential effects of repeated exposures over time on each type of ad are given, scales for measuring the informational and transformational content of ads are proposed and empirically validated, and suggested areas for further research are discussed.

INTRODUCTION

Few topics in marketing and consumer behavior generate as much interest, research, and discussion as does advertising. Yet we still do not know very much about the factors underlying the complex relationships which exist between consumers and advertisements. Gardner (1982) suggests that the time has come for advertising researchers to be less concerned with adding to the existing vast array of empirical observations of advertising effects and more concerned with the development of useful theories about how advertising works. Recent work by Edell and Staelin (1983), Mitchell (1983), and Olson and Reynolds (1983) has contributed to our understanding of how advertising works by integrating current theoretical work in cognition with empirical findings obtained in an advertising context.

The purpose of this paper is to propose a theoretical structure of advertising effects based on the cognitive and affective elements of informational (e.g., highly cognitive) and transformational (e.g., highly experiential) advertising (Wells 1980). The approach taken here is first to define informational and transformational advertising in more formal terms. Then these two descriptive classifications of advertisements are traced to their theoretical antecedents in information processing theories of cognitive psychology and theories of emotion and persuasion in social psychology. This theoretical development leads to a series of propositions regarding the effect that each form of advertising should have on traditional measures of advertising effectiveness (e.g., day-after-recall and change in brand attitudes). A method for measuring the informational and transformational levels of advertisements is proposed and evaluated, and the key issues for future research in this area are discussed.

DEFINITIONS OF INFORMATIONAL AND TRANSFORMATIONAL ADVERTISING

Informational Advertising

Presumably every advertisement includes some element of information, be it nothing more than the identity of the advertised brand. We define an informational advertisement, however, as one which provides consumers with factual (i.e., presumably verifiable), relevant brand data in a clear and logical manner such that they have greater confidence in their ability to assess the merits of buying the brand after having seen the advertisement. An important aspect of this definition is that an advertisement can be designed with the intention of providing information, but it does not become an informational ad unless it is perceived as such by consumers. Rather than residing wholly within the advertisement, this definition resides in the consumers' perceptions of the ad's contents.

For an advertisement to be judged informational in accordance with the preceding definition, it must reflect the following characteristics:

1. Present factual, relevant information about the brand .

2. Present information which is immediately and obviously important to the potential consumer.

3. Present data which the consumer accepts as being verifiable.

Transformational Advertising

A transformational advertisement is one which associates the experience of using (consuming) the advertised brand with a unique set of psychological characteristics which would not typically be associated with the brand experience to the same degree without exposure to the advertisement. Thus, advertisements in this category "transform" the experience of using the brand by endowing this use with a particular experience that is different from that of using any similar brand. It is the advertisement itself which links the brand with the capacity to provide the consumer with an experience that is different from the consumption experience that would normally be expected to occur without exposure to the advertisement.

For an advertisement to be judged transformational, it must contain the following characteristics:

1. It must make the experience of using the product richer, warmer, more exciting, and/or more enjoyable, than that obtained solely from an objective description of the advertised brand.

2. It must connect the experience of the advertisement so tightly with the experience of using the brand that consumers cannot remember the brand without recalling the experience generated by the advertisement.

Transformational advertising is comparable to other psychological descriptors often applied to advertisements (e.g., mood, emotional, feeling, and image advertisements) in that it is essentially affect-based rather cognitive-based. However, these descriptors represent necessary but not sufficient conditions for a transformational advertisement. The transformation occurs when these descriptors are explicitly related by consumers to the experience of owning or consuming the advertised brand.

It is worth noting at this point that information and transformation are not mutually exclusive categories of advertisements. They are, however, exhaustive. Thus, any given advertisement can be classified as belonging to one of four basic categories: (I) High Transformation/ Low Information, (2) Low Transformation/High Information, (3) High Transformation/High Information, and (4) Low Transformation/Low Information. Each of these categories is hypothesized to produce varying effects on the traditional measures of advertising effectiveness, and these effects are the basis for a series of theoretical propositions which are given and discussed below.

THEORETICAL STRUCTURE

Information processing concepts developed within the various subdisciplines of psychology have received considerable attention in consumer behavior research in recent years, and the theoretical foundations of informational advertisements are well documented in this literature (cf. Bettman 1979; Fishbein and Ajzen 1975). Consequently, they need not be repeated here. This is not the case, however, with transformational advertising, and the remainder of this section will address this issue.

The theoretical antecedents of transformational advertising derive from three diverse research areas: (1) emotion (Clynes 1980; Rogers 1983); (2) vicarious classical conditioning (Berger 1962; Petty and Cacioppo 1981); and (3) motivation research (Dichter 1964). The essence of transformational advertising lies in the concept of generalized emotion advanced by Clynes (1980). Described as an emotional state in which no situational cause is involved, generalized emotion affects an individual's knowledge structure in several ways, two of which are germane to transformational advertising.

First, it tends to facilitate the selective recall of past experiences associated with the same state. In effect, it acts as a memory search function for a class of experiences which have in common a similar quality of emotion. Importantly with respect to transformational advertising, this process occurs rather effortlessly, and furthermore, the association is based on drawing similar feelings from past experiences into active memory rather than the recall of the actual experiences. Thus, an advertisement which encourages consumers to "Reach out and touch someone by showing a scene depicting warm human relationships does not produce an effect by prompting consumers to recall prior telephone conversations with specific individuals. Rather, it associates the similar feelings generated by past (but not necessarily similar) experiences with present experience of making a long distance telephone call.

A second characteristic of generalized emotion is that it acts as a selective focus to create new fantasies which will then be associated with this state and its subsequent expression (Clynes 1980, p. 297). This is evidenced by a television commercial no longer being aired which invited consumers to Come to Marlboro Country. Viewers were free to overlay their own feelings and fantasies onto the scene, and these feelings and fantasies then became permanently associated with the experience of smoking the advertised brand of cigarette.

In both or the preceding instances, the effect of the advertisement was to transform the experience of using (consuming) the advertised brand to reflect the feelings evoked during the exposure to the ad. Moreover, these feelings are not necessarily based on prior experiences with the brand or product class but rather are based on prior similar feelings, regardless of the original source.

How this process of generalized emotional association takes place, especially with regard to advertisements, is not yet defined. There are, however, at least three viable explanations which do not necessarily represent competing hypotheses. The first is that empathy and closely related concepts such as involvement (Krugman 1965) and personal relevance (Schlinger 1979) are mediating factors which enable the consumer to relate the experience and feelings generated by the ad to the experience of using the advertised brand.

Empathy has received scant research attention as a potential mediator of advertising effects. This may in part be due to the rather restrictive treatment of empathy in social psychology as primarily a mediator of helping behavior. Stotland (1969) defines empathy as ". . .an observer's reacting emotionally because he perceives that another is experiencing or is about to experience an emotion" (p. 272). For advertisers, a more useful definition of empathy is . . .the imaginative projection of a subjective state into an object so that the object appears to be infused with it (Merriam 1979, p. 369). We define empathy as a combination of these approaches such that, with respect to advertisements, empathy is an observer's vicarious emotional identification with the execution of an advertisement. Thus, an advertisement produces empathy among consumers if it causes them to identify emotionally with the actors, the situation, or some other element of the content of the ad. In the context of this discussion, involvement and personal relevance represent necessary but not sufficient conditions for empathy.

A second explanation for how transformational advertising works is through vicarious classical conditioning, which is a combination of classical conditioning and observational learning principles (Petty and Cacioppo 1981). In vicarious classical conditioning, the unconditioned stimulus is a strong, observable, psychological reaction on the part of a social model. The vicarious conditioning occurs when observers (consumers) witness a strong psychological reaction by the social model (actors or persons depicted in an advertisement) toward an otherwise neutral stimulus (a brand). Over time, the neutral stimulus will come to elicit the same psychological response from the observers as it did from the social model. When no People are depicted in the advertisement, the effect becomes one more closely associated with traditional classical conditioning in which the viewer's response to the advertisement is conditioned to his/her response to the brand presented in the advertisement (Petty and Cacioppo 1980).

A third possible explanation stems from recent work in cognition and social cognition (c.f ., Rogers 1983; Zajonc 1980) which suggests the possibility that Paivio's (1971) dual coding theory (i.e., items can be encoded in memory through either a verbal or an imaginal code) may be extended to include a third code, emotion, which interacts with the visual and verbal codes to form a more complete representation of the cognitive process. Rogers reviews work by Kirker (1981), Posner and Snyder (1975), and Zajonc (1980) which ... begins to offer a convergent data set that is compatible with [an emotional code] (1983, p. 295, emphasis from original quotation).

An additional theoretical question concerns the relationship between transformational advertising and consumer buying decisions. That is, Why should transformational advertising work? One reason stems from the idea generally credited in the marketing literature to Dichter (1949; 1964) that not all consumer buying decisions are necessarily cognitive in nature. While Dichter's research methodologies and reported results have been the subjects of considerable debate and criticism within the discipline, the reasoning underlying his proposals remains quite viable.

Many of these same ideas are embodied in Hirschman and Holbrook's (1982) recently proposed concept of hedonic consumption. In summarizing the literature in this area, Hirschman and Holbrook offer two propositions which are of particular importance to the discussion of transformational advertising:

1. In some instances, emotional desires dominate utilitarian motives in the choice of products;

2. Consumers imbue a product with a subjective meaning that supplements the concrete attributes it possesses (1982, p. 94).

Given that a marketer accepts these propositions, transformational advertising represents an appropriate marketing communications methodology for applying them.

Theoretical Propositions

The conceptual framework given above leads to several theoretical propositions concerning the differential effects of informational and transformational advertisements on variables of interest to advertising and communications researchers. Each of these propositions is given and discussed briefly below.

Proposition 1: To be effective as a persuasive communication, an advertisement must be informational, transformational, or both.

By definition, an advertisement which is neither informational nor transformational makes no positive contribution to the consumer's brand perceptions and will not perform well on any of the traditional measures of advertising effectiveness (e.g., day-after-recall tests or measures of changes in brand attitudes).

Proposition 2: Informational and Transformational advertisements differ with respect to their performance on standard day-after-recall tests.

By providing specific, objective data about a brand, informational advertisements facilitate the transfer of information to long term memory. Thus, informational advertisements should produce superior scores on day-after-recall tests compared to transformational advertisements, which tend to contain more emotional and more nonverbal elements. ID a direct comparison of "thinking" versus "feeling" commercials, Berger (1981) reported superior performance for thinking ads on standard day-after-recall tests. However, when a masked recognition test was used, the memory differential for the two types of commercials disappeared. While the memory trace for the transformational commercial may in fact exist, if the encoding was primarily visual or emotional, the relevant information will not be available for immediate verbal recall. After repeated exposures, the images or feelings conveyed in the transformational advertisement may be sufficiently well-formed in consumers' minds to permit them to be translated into an adequate verbal response on a recall test. Therefore, transformational advertisements should reflect an improvement in recall scores over repeated numbers of exposures to the point where there is essentially no difference between them and informational advertisements.

Proposition 3: Transformational advertisements will require several exposures over time in order to produce an effect on brand attitude change measures.

Because they operate mainly on emotions and are essentially indirect in nature, the effects of transformational advertisements on consumers' brand attitudes should be cumulative over repeated exposures. Thus, there should be little change in consumers' attitudes towards the advertised brand after one typical exposure to a transformational advertisement. Repeated exposures to a transformational advertisement over time, however, should produce substantial, favorable changes in brand attitudes.

MEASUREMENT ISSUES

In order for these propositions to be developed into testable hypotheses, methods for identifying the informational and transformational content of specific advertisements are needed. One method for doing this is to have judges examine the advertisements in conjunction with the definition given above and assign them to one of the four quadrants in the information/transformation space. This presumes, however, that these constructs can be readily assigned to discrete categories when, in fact, the degree or extent of the informational or transformational nature of an ad may be important in evaluating its effects.

It may also be possible to identify transformational ads through the use of projection techniques such as sentence completion, picture interpretation, or depth interviews with individual subjects. These techniques, however, remain open to the same criticisms directed toward the motivation research proponents of the 1950's and 1960's, namely that the findings are extremely dependent on the subjective interpretations of the researchers.

Because our definitions are dependent on the consumers' subjective perceptions of the content of the advertisement, we feel that relying on the investigator's interpretations of this content further compounds the measurement difficulty. Therefore, we propose to use a series of scale items which attempt to tap the transformational effects of individual ads across groups of target consumers. The successful development of such a scale would permit researchers to compare the relative transformational nature of various ads. Using a standard measure also facilitates the comparison of findings across multiple research projects dealing with this subject. Similar reasoning applies to the measurement of an ad's perceived informational content.

In the sections which follow, we briefly outline the procedures used in item selection and item analysis. Throughout these sections, we have endeavored to adhere as closely as possible to the suggested paradigm for construct measurement given by Churchill (1979) and to the scale development procedures given by Nunnally (1978)*

Scale Item Development

Candidate items for the informational scale were derived from items used in prior research on the informational content of advertisements (c.f., Aaker and Norris 1982: Resnik and Stern 1977), to which were added a series of items derived from the definition of informational advertisements given above.

In order to tap the emotional and experiential aspects of the transformation construct, we examined prior research relating to empathy (cf. Stotland 1969) and personal relevance (cf. Schlinger 1979). The majority of the published research on empathy is concerned with measuring the empathic tendencies of individuals rather than of objects or circumstances. Since our primary research concern is the latter, it was necessary to use the definition of transformational advertising to develop original items for this aspect of the transformation scale. Items tapping the personal relevance aspect of the transformation scale were derived from the viewer response profile measures given by Schlinger (1979) and from the revised Needham, Harper, and Steers communications measures (unpublished). The items comprising both scales and the order in which they appeared on the test instrument are given in Table 1.

Methodology

One of the authors (CP) reviewed approximately 400 television commercials and selected 20 which (1) were either mainly informational or mainly transformational and (2) promoted products of interest to the planned test audience (college students). These 20 commercials were then independently judged on this same basis by two judges who were knowledgeable in the definitions but were blind to the prior classifications. The two judges and the author were in 100 percent agreement on the classification of 13 commercials (5 mainly informational and 8 mainly transformational), and these were used in the validation phase of the research. Recall that information and transformation are not mutually exclusive categories; however, commercials judged to be predominant on just one dimension were selected in this preliminary stage of the research in order to establish a basis for discriminating between the two constructs. The 10 product classes and the order of presentation for the 13 commercials are given in Table 4 below.

The subjects used in the scale validation were 130 undergraduate psychology students who were participating in partial fulfillment of a course requirement. The subject were tested in a theater setting in two approximately equal groups. For each group the commercials were shown on large screen projection television, and the subjects were told that the purpose of the research was to obtain their opinion of various television commercials. Each commercial was shown once. Immediately after seeing a commercial, subjects reported their prior exposures to it on three-point scale (never, a few times, many times) an their overall opinion of it on a five-point scale anchored by very favorable and very unfavorable. These measure were followed by the 23 scale items for which the subject indicated their degree of agreement or disagreement on six-point scale anchored by Strongly Agree and Strongly Disagree. This process was repeated for each of the 13 commercials. The order of presentation of the commercial was the same for both groups.

TABLE 1

INFORMATION AND TRANSFORMATION SCALE ITEMS

Scale Item Analysis

Recall that each of the 130 subjects responded to each scale item for 13 different commercials. To minimize con founding due to variability among the test commercials, al measures were first computed within each commercial an then, where appropriate, were averaged across commercial to obtain descriptive summary measures.

The first of these summary measures are the average item to-total scale correlations given for both scales in Table 2. The low item-to-total correlations for items 1, 12, and 21 in the information scale (see Table 2 A) suggest that they may not be good candidates for retention in future work using this scale. We have elected to retain them, however, for several reasons. The low correlation for item 1 may be an artifact of the prior exposure of these particular commercials. a is item is specifically related to new information, and it would appear to be of value in a typical pretest environment in which commercials are viewed for the first time. Item 12 is a component of the definition of the information construct, and we can offer no explanation for its low correlation with the rest of the scale. Since it does not reduce the reliability of the scale (see next section), we recommend that it be retained while the scale undergoes further refinement. Item 21 is, retrospectively, incorrectly phrased, and in future uses of this scale it should be reworded as follows: I would have more confidence in my ability to judge the merits of buying (this brand) now that I have seen this commercial.

The average item-to-total scale correlations for the transformation scale are given in Table 2 (B), and they are generally at a satisfactory level. The rather broad ranges for items 3, 14, 16, and 17 suggest that they are subject to considerable variability across different classes of commercials. Deletion of these items did not, however, improve the scale reliability coefficients for the individual commercials, and thus we recommend their retention in the scale at this point.

TABLE 2

AVERAGE ITEM-TOTAL CORRELATIONS FOR INFORMATION AND TRANSFORMATION SCALES

Scale Reliability Measures

Cronbach's coefficient alpha (Nunnally 1978) is reported for each scale in Table 3. These reliability coefficients were computed separately for each commercial and then averaged across the 13 test commercials to obtain the figures reported in the table. The ranges on these coefficients across commercials are quite narrow, and the mean values indicate that the scales exhibit a very satisfactory degree or reliability.

Construct Validation

Informational and transformational scale values for each commercial were obtained within subjects by summing the items comprising each scale and dividing the sum by the number of items in each respective scale. Mean scale values across subjects were then computed for each commercial and are given in Table 4. The information and transformation scale means for each commercial differ at the P < .001 level of significance (one-tailed t-test). Furthermore, for each commercial, the statistically higher scale mean accurately predicts the judges' prior assignment of that commercial as being either primarily informational or primarily transformational. Thus, there is 100 percent agreement between the judges' prior determinations and the informational and transformational scales for each of the 13 test commercials.

Another way to interpret these data would be to compare the information and transformation scale means for each commercial with the mid-point of each scale. The result would be to equate this mid-point with an arbitrary or relative zero' point on the scale. In the case of the six-point scales used in this research, those commercials below the 3.5 mid-point would be classified as being low on that dimension while those above it would be classified as being 'high.

TABLE 3

AVERAGE RELIABILITY COEFFICIENTS FOR INFORMATION AND TRANSFORMATION SCALES

Using this interpretation, it is possible for a commercial to be significantly higher (in a statistical sense) on one dimension versus the other while still being low on both dimensions in comparison to the scale mid-points (see, for example, commercials No. 2, 6, and 10 in Table 4). Pending further empirical work designed to establish norms for "low" and "high" values of information and transformation, we recommend that the actual scale values rather than arbitrary discrete categories be used in testing hypotheses regarding the relative effectiveness of the informational and transformational content of specific advertisements.

TABLE 4

SCALE MEANS FOR EACH COMMERCIAL

CONCLUSIONS

We have defined information and transformation as two descriptive constructs for classifying advertisements according to the way in which they affect consumers' perception of their content. The theoretical foundations for each o these constructs were presented, and they were related to advertising through a series of research propositions concerning the effects that varying levels of information and transformational content in advertisements will have o the traditional measures of advertising effectiveness. We have proposed and tested a series of scales for measuring these constructs.

In the brief space remaining, we will enumerate some of the many fertile opportunities for additional research. First Wells (1980) has proposed that some media (e.g., radio and television) are more suitable for transformational advertising than others (e.g., print). The research report ed here suggest that television commercials are capable o functioning as a transformational medium. The application of this to other media remains an important and open question.

Second, the viewing of these test commercials occurred under conditions of considerable artificiality, but this was necessary in order to validate the scales. Now that al adequate measure of an advertisement's transformational content is available, an important next step involves identifying the length of time required for a transformational advertisement to be effective under normal exposure conditions.

Third, we have proposed that advertisements can be places in a two-dimensional space according to their information and transformational content. In the present paper however, we have reported only marginal evidence in support of the four quadrants comprising this space. Further work in identifying and describing advertisements in all four quadrants is an essential part of this theory.

Fourth, and quite possibly the most important, is the issue of the advertisement's ability to transform the actual experience of using the brand. This is a critical component of the theory, and it remains to be empirically demonstrated in a controlled experimental environment. Some work i already in progress along these lines, but the opportunities for additional contributions are virtually boundless We feel that these concepts rest on firm theoretical grounds, and we hope that we have been successful in providing a framework within which they can be systematically examined.

REFERENCES

Aaker, David A. and Norris, Donald (1982), "Characteristic of TV Commercials Perceived as Informative," Journal of Advertising Research, 22, 61-70.

Berger, David (1981), -Measuring the Effectiveness of Emotional and Rational TV Commercials, presented at the Advertising Research Foundation Mid-Year Conference, Chicago Ill.

Berger, S. M. (1962), Conditioning Through Vicarious Instigation, Psychological Review, 69, 450-466.

Bettman, James R. (1979), An Information Processing of Consumer Choice, Reading, Mass.: Addison-Wesley.

Churchill, Gilbert A. Jr. (1979), "A Paradigm for Developing Better Measures of Marketing Constructs," Journal of Marketing Research, 16, 64-73.

Clynes, Manfred (1980), The Communication of Emotion Theory of Sentics, in R. Plutchik and H. Kellerman (eds.) Emotion: Theory, Research, and Experience, New York: Academic Press.

Dichter, Ernest (1949), A Psychological View of Advertising Effectiveness, Journal of Marketing, 14, 61-66.

Dichter, Ernest (1964), Handbook of Consumer Motivation, New York: McGraw-Hill.

Edell, Julie A. and Staelin, Richard (1983), The Information Processing of Pictures in Print Advertisements, Journal of Consumer Research, 10, 45-61.

Fishbein, Martin and Ajzen, Icek (1975), Belief, Attitude Intention and Behavior: An Introduction to Theory and Research, Reading, Mass.: Addison-Wesley.

Gardner, Burleigh, (1982), A Conceptual Framework for Advertising, Chicago: Crain Books.

Hirschman, Elizabeth and Holbrook, Morris (1982), Hedonic Consumption: Emerging Concepts, Methods and Propositions," Journal of Marketing, 46, 92-101.

Kirker, W. S. (1981), Self-Reference and Emotion, unpublished master's thesis, University of Calgary, Canada.

Krugman, Herbert E. (1965), The Impact of Television Advertising: Learning Without Involvement, Public Opinion Quarterly, 29, 349-356.

Merriam, G. (1979), Webster's New Collegiate Dictionary, Springfield, Mass.: G. and C. Merriam Company

Mitchell, Andrew A. (1983), 'Cognitive Process Initiated by Exposure to Advertising,- in Richard Harris (ed.), Information Processing Research in Advertising, (Hillsdale, N.J.: Lawrence Erlbaum).

Nunnally, Jum C. (1978), Psychometric Theory, 2nd Ed., (New York: McGraw-Hill).

Olson, Jerry C. and Reynolds, Thomas J. (1983), "Understanding Consumers' Cognitive Structures: Implications for Advertising Strategy, in Larry Percy and Arch G. Woodside (eds . ), Advertising and Consumer Psychology, (Lexington, Mass.: Lexington Books).

Petty, Richard E. and Cacioppo, John T. (1981), Attitudes and Persuasion: Classic and Contemporary Approaches, Dubuque, Iowa: William C. Brown Company.

Posner, M. and Snyder, C.R.R. (1975), Attention and Cognitive Control, in R. L. Solso (ed. ), Information Processing and Cognition, New York: Wiley.

Resnik, Alan and Stern, Bruce L. (1977), An Analysis of Information Content in Television Advertising," Journal of Marketing, 41, 50-53.

Rogers, T.B. (1983), Emotion, Imagery, and Verbal Codes: A Closer Look at an Increasingly Complex Interaction, in J.C. Yuille (ed.), Imagery, Memory and Cognition, Hillsdale. N.J.: Erlbaum.

Schlinger, Mary Jane (1979), A Profile of Responses to Commercials, Journal of Advertising Research, 19, 37-46.

Stotland, Ezra (1969), Exploratory Investigations of Empathy, in L. Berkowitz (ed. ), Advances in Exploratory Social Psychology, Vol 4, New York: Academic Press.

Wells, William D. (1980), How Advertising Works , Needham, Harper and Steers Advertising, Inc., Chicago.

Zajonc, R.B. (1980), 'Feeling and Thinking, American Psychologist, 35, 151-175.

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