Emotional Product Differentiation By Classical Conditioning (With Consequences For the "Low-Involvement Hierarchy")

Werner Kroeber-Riel, University of the Saarland (W. Germany)
In saturated markets, informative advertising plays a but secondary role. The attitudes towards brands are formed, above all, by emotional advertising and other emotional marketing activities, mainly according to the principles of classical conditioning. Based on empirical results the conditions for an effective conditioning are formulated. Classical conditioning should be regarded separately from other forms of attitude learning. It can be characterized as passive learning with predominant emotional responses. Low-involvement learning can be explained with classical conditioning. This leads to a modification of the so-called low involvement hierarchy of consumer behavior.
[ to cite ]:
Werner Kroeber-Riel (1984) ,"Emotional Product Differentiation By Classical Conditioning (With Consequences For the "Low-Involvement Hierarchy")", in NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT : Association for Consumer Research, Pages: 538-543.

Advances in Consumer Research Volume 11, 1984      Pages 538-543


Werner Kroeber-Riel, University of the Saarland (W. Germany)

[The author wishes to express his gratitude to Professor R. P. Bagozzi and Professor Jerry Olson for ideas on the subject.]


In saturated markets, informative advertising plays a but secondary role. The attitudes towards brands are formed, above all, by emotional advertising and other emotional marketing activities, mainly according to the principles of classical conditioning

Based on empirical results the conditions for an effective conditioning are formulated. Classical conditioning should be regarded separately from other forms of attitude learning. It can be characterized as passive learning with predominant emotional responses. Low-involvement learning can be explained with classical conditioning. This leads to a modification of the so-called low involvement hierarchy of consumer behavior.


"Product Differentiation" refers to marketing activities which cause a consumer to differentiate one brand from another competing brand. This is achieved when the consumer (1.) perceives that the (objective) quality of a brand is different from that of competing ones or (2.) attaches emotional feelings with a brand which differentiates it from others.

Emotional product differentiation attempts the second goal, it attempts to give a brand an emotional value which distinguishes it from other brands. E.g.: A cigarette brand can be associated with prestige feelings; it distinguishes itself from competing brands with project different feelings such as nature, freshness, manliness, enjoyment, etc..

Emotional product differentiation is mainly achieved by advertising, but also by design and packaging of the products, by POS, and by other marketing activities. We will limit ourselves here to advertising.

As a design of such advertising we can consider an ad which has the following components:

- the brand name or a picture of the brand

- a picture which induces emotional feelings

- a headline (and possibly a short text).

This type of advertising dominates in many markets, for example in those which sell beer, cigarettes, perfume, detergents. etc..

These are saturated markets. In these markets, the products are mature; they only show minor differences in quality. As demonstration: In the opinion of two thirds of Europeans, the leading brands of beer, cosmetics, and electrical household appliances are of the same quality (Kanter 1981).

The possibilities for objective differentiation of the products are slim in these markets. In addition, the consumer shows little interest in the differences of quality of mature goods.

Information about the product, therefore, does not play an important role. The consumer's preference for a brand is-mainly created by emotional brand experiences. A content analysis of advertisements may show, which emotions are addressed to consumers to establish emotional product differentiation. To give an example, we here report some emotional brand experiences, which are mediated in some common product fields by printed ads in German magazines

for alcoholic drinks:  (1) prestige, (2) tradition (3) companionship;

for tobacco and cigarettes: (1) enjoyment (2) self-confidence (3) activity

for cosmetics: (1) eroticism (2) nature and freshness (3) prestige.

The figures indicate the rank of frequency. Whether the emotional content of ads will really lead to corresponding feelings of the consumer and whether these feelings will be correctly associated with the brands in the mind of the consumer remains an open question. One can, however, assume that an analysis of the advertisements will at least partially show what the emotional brand experiences of the consumer are


We now pose the question as to the effects of emotional advertising. Which laws does the mediation of emotional brand experiences follow?

According to theories of verbal learning, we can formulate the following hypothesis:

If a "neutral" brand name is repeatedly presented together with an emotional stimulus, the brand name will assume an emotional meaning.

In this way, the brand name receives the ability to release emotional reactions in the consumer. Instead of a brand name, one can also speak more generally of the symbolic representation of the brand in the advertisement (as in a picture, symbol, etc.).

This hypothesis refers to the model of associative learning. This is an abstract model which is compatible with various principles of learning. One of the learning principles which can be used for a more exact explanation of associated learning is classical conditioning (Lefrancois 1972).

In the language of classical conditioning, the brand name is the "conditioned stimulus" (CS). It will be emotionally loaded by being presented repeatedly with an "unconditioned stimulus" (UCS) which is the emotional stimulus.

Some basic research on conditioning of words was done by Staats and Staats (1958). Nationality names such as "Swedish" and "Dutch" were used as neutral stimuli and strong emotionally loaded words were used as the unconditioned stimuli. The nationality names received an additional emotional meaning by repeated presentation with the emotionally loaded words (UCS). They became conditioned stimuli (CS).

What has been designated according to the investigations of Staats and Staats as the emotional meaning of a name, is considered to be the attitude towards the object in social psychology and in marketing research (Staats (1958) makes this point very clear). This equation will also be made evident by the methods used for measurement: The meaning of a brand name and the attitude toward the brand are conveyed by the same methods among others by over-all-ratings of the affective evaluation or by the semantic differential.

Taking the basic research of Staats and Staats as a starting point, the hypothesis can be formulated that advertisements could change attitudes towards products by classical conditioning (Staats 1968; Mednick, Pollio, Loftus 1973). This hypothesis was tested by us in a 2x2-experiment in which advertisements were tried out using the model "brand name + emotionally loaded picture". The first results of this experiment were published in a working paper- (Kroeber-Riel, von Keitz 1980); Kroeber-Riel (1980) presented a more complete description of the experiment.

Slide advertisements - exposed for five seconds - were presented to subjects in a simulated film theater. Among others. "Hoba soap" was selected as a neutral brand name. The pictures conveyed emotional events concerned with eroticism, social happiness, and exotic landscapes. They had no relationship to the product; that is, they did not transfer any information about the product.

The following variables were manipulated between subject:

the emotional intensity of the pictures (strong--weak) and the additional textual information (present--absent)

within subject:

number of repetitions (up to thirty).

The following dependent variables were measured: The emotional reactions to the brand name, the attitude towards the brand and the specific beliefs as well as the readiness to try the product.

At the beginning of the experiment, it was found out by psychobiological measurements and by verbal ratings that the brand name "Hoba soap" had practically no emotional meaning. After the conditioning, the name alone - exposed twenty-four hours after the final conditioning period - aroused significant emotional reactions.

Before the conditioning, the values on the evaluative (five point) ratings of the semantic differential were between 3.9 and 4.2; after conditioning, they were 0.4 units better (lower). These differences are significant. The intensity of the emotional meaning which had been acquired by the brand name was confirmed by additional psychobiological measurements (of the electrodermal reactions).

The presence of written text did not have any influence on the success of the conditioning. It is, therefore, possible to change the emotional meaning of a brand name and with it the attitude towards the brand by emotional persuasion without any information being given about the product. In order to explain these results, we should recall that the process of attitude formation by classical conditioning occurs without intervening cognitive reactions (see section 5).


The following is a summary of the requirements for effective conditioning which were established in the experiment referred to in Kroeber-Riel (1980). Unfortunately, there are no further experiments which document the emotional product differentiation by means of classical conditioning.To corroborate our findings, we have to rely on experiments from basic research.

1. The emotional stimulus (UCS) must be relatively strong.

The weak emotional stimuli did not have any effect in the experiments of Kroeber-Riel. Their use was an important cause for the lack of success in conditioning experiments (Ertl, Oldenburg, Vormfelde-Siry, Vormfelde 1971). Emotional stimuli which release biologically preprogrammed reactions such as erotic stimuli are especially strong and as such effective for use in conditioning.

2. Numerous repetitions are necessary for conditioning. It is emphasized in the literature that the first presentations of the stimuli have very little or no effect (Bower, Hilgard, 1981, p. 50). In the Staats and Staats experiment, 18 conditioning "trials" were necessary in order to give the "neutral" stimulus an emotional meaning. In our experiment, there was no conditioning success after 18 repetitions. Thirty repetitions distributed over ten days were necessary to achieve a significant effect. It seems that twenty to thirty repetitions are necessary in order to do the conditioning for brand names.

In this context, it is remarkable that the mere repetition of the brand name in the advertisement can result in an increase in liking for the brand name (the brand) without presentation of further emotional stimuli. This "mere exposure effect" was demonstrated by Zajonc (19685

3. Passive stimulus acquisition and processing increases the chances of conditioning.

Limiting the cognitive participation of the subject is an important characteristic of classical conditioning. In the conditioning process the recipients are not aware of the ad influence (Staats and Staats, 1958). They process the stimuli automatically without reflecting them cognitively.

If active cognitive processing and control of the exposed stimuli are present, one can expect voluntary barriers against the intended advertising influence (compare Rudy, Wagner 1975, p.280; for different approaches see Ross, Ross 1976, p. 107 ff. ). This is also to be expected in emotional product differentiation when the recipients reflect that the "information" presented by ads has obviously nothing to do with the product itself.

4. Pictures are better UCS than words.

This is caused by two effects of pictures: First, the processing of pictures requires less mental effort on the part of the recipient than the processing of words (Haber 1981). In general, pictures are processed automatically. This enables the recipient to acquire stimuli passively.

Secondly, pictures are better suited than words to effect emotional reactions. This is related to the fact that emotional behavior is mainly related to activities of the right hemisphere, and this activity is more strongly stimulated by pictures than by words (Tucker 1981).

To formulate an order of the effectiveness of emotional stimuli, one notices that real objects present the most effective UCS for conditioning. Real objects can be substituted by pictures (King 1978). The pictorial models are more effective than verbal representations of the objects.

In addition to these conditions, there are numerous other conditions which have to be kept in mind, among others, the length of stimulus exposure and their order should be considered. The literature on emotional conditioning is full of information on this.

However, several questions on conditioning by means of advertisements remain open. For example, it is well known .that conditioning is most effective when the neutral stimulus is presented before the emotional stimulus (UCS) and when the time between the presentation of the neutral and emotional stimulus is no longer than 0.5 seconds. This condition can hardly be met in printed advertising, because in general, one's attention is usually attracted by the emotional stimulus first (which means that printed advertisements deal with the less effective "backward conditioned response" in which the emotional stimulus is presented respectively acquired first).

A further question concerns the stability of the conditioned emotional response (induced by advertisement) and the repetition needed to maintain the proper reaction.

Finally, let us consider whether the advertisement described here (brand name + emotional picture) can be called a "real" conditioning. The search for the limits of such a definition does not seem very productive. According to everything we know today, product differentiation by means of emotional advertising is essentially done according to conditions which are typical for classical conditioning. It is an open problem how to interpret the active cognitive information processing which always is present in the subject to some extent and modify the effects of conditioning c Mart in 19 In


Models of attitude learning by means of advertising should include at least two basic determinants:

- predominance of cognitive or emotional responses

- activeness or passiveness of the subject.

Activity and passivity are the extreme ends of a continuum: The presented stimuli are either acquired and processed actively with directed attention (activation) or they are hardly noticed and acquired with minimal mental participation.

Independently of this, learning can be performed mainly by cognitive or emotional reactions.

Using these determinants, we intend to distinguish here between some manifestations of attitude learning which are important for the understanding of classical (emotional) conditioning.

Predominance of cognitive responses:

In consumer research, above all, such attitude changes are studied which can be explained by cognitive attitude models, especially the Fishbein model. According to this model, attitudes are a function of :he perception and evaluation of product attributes.In :he language of Fishbein: "The attitude toward an object is dependent from the individual's beliefs about :he object and the evaluative aspects of the beliefs" (Fishbein 1972).

In order to produce attitude formation or change experimentally, the subjects are usually exposed to verbal information about a product in such a way that the information is acquired and processed with attention. After one or two presentations of the informative stimuli, attitude formation or change occur based on cognitive responses to the advertisement and the induced changes in cognitive structure (see for example Lutz 1975; Olson, Toy, Dover 1982).

This process can be characterized as active attitude learning with the predominance of cognitive reactions. A more passive cognitive learning may be expected when the influencing stimuli can be processed by the subject without or with little mental effort. This is a reliable factor when product information is presented through pictures.

Mixed cognitive and emotional responses:

Of course, cognitive responses are nearly always accompanied by emotional responses. However, they only play a subordinate role in forming a predominantly cognitive attitude thus being neglected as "quantite neglegeable".

Now we should like to consider emotional responses explicitly, which is required when strong emotional stimuli are applied in advertising and the effected emotional responses rank equally with the cognitive responses and codetermine attitude formation.

This is assumed when product information is presented in a strong emotional context. For example: An ad presents a car and a girl, with or without additional verbal text.

The persuasion effects of such an emotional context have been examined but insufficiently (Smith, Engel 1968; Kanungo, Pang 1973; Mitchell, Olson 1981; and 'Kroeber-Riel 1983). These experiments were carried out under such conditions as to ensure an attentional and active processing of the advertising stimuli. In this case, the emotional context causes an effectful climate of perception leading to a selective perception and evaluation of product attributes. Besides, it effects specific emotional associations towards the product. As the quoted experiments verify, both processes interfere considerably with attitude formation. This may be called active attitude learning with mixed cognitive and emotional responses because both product specific (cognitive) beliefs and emotional impressions come into force.

Predominance of emotional responses:

Now the advertisement does not contain any product information, neither in the form of a picture of the product nor in the form of direct or indirect information about the brand.The advertisement fits the stimulus pattern "brand name + emotional stimulus".

In an experiment by Mitchell and Olson (1981), an ad was presented which contained nothing more than the brand name "cosmetic tissues brand L" and a picture of a sunset. In this case, active learning took place due to the experimental conditions which had been set up; this learning induced a change in attitude towards the brand after two presentations. In the experiment by Kroeber-Riel (1980) described above, passive subjects were confronted with advertisements of a similar type. Attitude change occurred after thirty repetitions.

In both experiments, the learning process was dominated by emotional reactions, even though it followed different rules in active and passive subjects.

One can assume that the product related knowledge would remain for the most part unchanged in such emotional attitude learning. Changes in product specific beliefs, which may be noted after the learning process are probably a consequence (not the cause) of the emotional evaluation of the product.

In summary: By means of emotional advertisements it is possible to change attitudes towards brands without intervening cognitive responses. The brand related emotional reactions which are stimulated by the advertisement lead directly to an attitude towards the brand. These possibilities of emotional learning are not recognized by the cognitive attitude models but confirmed by recent causal models. According to these causal models the affective responses to an object represent the attitude to the object without additional - cognitive - beliefs (Bagozzi 1982; Burnkrant, Page 1¦.89)


It is generally accepted that low-involvement learning is characterized by:

- passivity on the part of the subject

- preference for pictorial stimuli

- frequent repetition of stimuli

- non-intentional processing of the stimuli.

If these characteristics are compared with the conditions for effective conditioning mentioned above far-reaching compatibility can be recognized. The classical (emotional) conditioning is one form, possibly the most important form, of low-involvement learning (see also Trommsdorff, Schuster 1981, pp. 750).

This view is confirmed when the market conditions in which the low-involvement behavior of consumers occurs are taken into consideration. Consumer involvement is low when the products have only minor quality differences from one another (Ray 1973, p. 9; DeBruicker 1979 p. 119 ff.; Lastovicka, Gardner 1979). This is especially the case in saturated markets with mature products. It is exactly in these markets that product differentiation by means of emotional conditioning is the preferred strategy of influencing consumers.

The insight that classical conditioning is a typical form of low involvement learning provides an important link between empirical results in low involvement behavior of consumers and the general learning theories which has been missing until now.

Furthermore, this suggests a modification of the socalled hierarchy of consumer behavior: In this connection, let us consider the well-known model of low-involvement hierarchy of Ray (1973). In this model, the usually hypothesized order "first attitude, then behavior" is reversed for low involvement conditions to "first behavior, then attitude". To support this, Ray refers to Krugman (1965) on low-involvement learning by television. Ray wrote (1973, p. 8-9):

"Television ads may not directly change attitude, they might after overwhelming repetition make possible a shift in cognitive structure. Consumers may be better able to recall the name or idea of the product. Then the next time they are in a purchasing situation, that names come to mind, the buy and attitude is subsequently changed as a result of experience with the product. Thus, this low-involvement hierarchy is a cognitive-connotative-affective one.

Thereby Ray assumes a two-step sequence of behavior: In the first step, before purchase, brand awareness and familiarity results from learning by repetition. An attitude has not been formed at this point because the influence of television advertising is not sufficient to produce a structured attitude. But brand awareness is enough to bring about the purchase of the product. After the purchase, when the consumer has experience with the brand, attitude formation occurs and an affective relationship to the brand is established.

This hypothesis corresponds to a cognitive way of looking at things. According to this way, the consumer evaluates the perceived characteristics of the product when he uses it. Only now - as a result of the evaluation - the attitude towards the product is formed.

In addition to this sequence of behavior, which results from cognitive attitude learning - which corresponds correctly to patterns of consumer behavior - there are, nevertheless, other behavioral sequences which are based on emotional attitude learning, and which conform to low-involvement conditions:

It is before the purchase that an attitude towards the brand can be formed by passive repetition learning. This is achieved by means of emotional conditioning which the (low-involvement) consumers are exposed to (among others by TV).

Also after the purchase, the attitude to the brand can as well be developed by means of emotional influence as by cognitive learning: As was mentioned in the section on "Emotional Product Differentiation" concerning saturated markets, the perception and evaluation are of secondary importance for the consumer. This is also true during the consumption stage of the product. Also in this stage, the emotional brand experience seems to become the main determinant of the consumer's attitude to the brand.

Naturally, we do not assume that attitude learning takes place exclusively in one or the other of these two ways. Both cognitive and emotional learning, passive and active learning interact in many ways. What we are trying to explain here are the basic possibilities of attitude formation.

The basic difference between low involvement behavior and high involvement behavior cannot generally be searched in reversing the formula "attitude ---o behavior" when considering low involvement behavior. The difference can be better understood by realizing that the attitudes develop differently: With high involvement behavior, attitudes are formed by active learning, with low involvement behavior, attitudes are formed by passive learning. Emotional product differentiation by means of classical conditioning is a typical example of this passive attitude learning. (We are not discussing further forms of passive attitude learning as studied by Leavitt, Greenwald, et al 1981; Petty, Cacioppo 1981).

The empirical results which Ray (1973) collected in order to support his formulation of the low-involvement hierarchy (= behavior --> attitude) does not contradict our hypothesis of emotional attitude learning (both before and after the purchase):

Ray identified the effectivelessness of television advertising on the attitudes of low-involvement consumers when TV advertisements were repeated six times. As we know from research on conditioning six repetitions are hardly sufficient for attitude learning. (In addition, a strongly emotional advertisement is necessary in order to form an attitude.

In summary: Attitudes can be created directly by emotional advertising (product differentiation) in the case of low involvement behavior; this can be done independently of knowledge and use of the Product.

The effectiveness of emotional advertising is not as is often claimed mainly dependent on the medium (TV or print) but on the extent to which the advertising uses emotional pictures which can be effective when looked at cursorily. In that printed advertising also uses emotional pictures, it can be as well as TV advertising considered an effective medium for low-involvement learning.

This is especially true in Germany where TV advertisements play a secondary role due to government regulations. Emotional product differentiation is mainly performed in Germany by the popular magazines such as Stern. The growing importance of emotional product differentiation and with it the low involvement learning of consumers is shown, among other ways, by the fact that the space used for pictures in ads has grown considerably in saturated markets in the last few years. According to our content analysis, it rose, for example, in the advertisements for alcoholic drinks from 25 per cent to 80 per cent in the last twenty years!


Bagozzi, R. P. (1982), "A Field Investigation of Causal Relations Among Cognitions, Affect, Intentions, and Behavior", Journal of Marketing Research, XIX, Nov. 1982. 562-585.

Bower, G. H. and E. R. Hilgard (1981), Theories of Learning, 5th ed., Englewood Cliffs: Prentice Hall.

Burnkrant. R. E. and T. J. Page Jr. (1982), "An Examination of the Convergent, Discriminant, and Predictive Validity of Fishbein's Behavioral Intention Model." Journal of Marketing Research, XIX, Nov. 1982, 550-562.

DeBruiker, F. St. (1979), "An Appraisal of Low-Involvement Consumer Information Processing," Attitude Research Plays for High Stakes, (eds.) J. C. Maloney and B. Silverman, Chicago: American Marketing Association, 112-130.

Ertl, F., E. Oldenburg, U. Vormfelde-Siry and D. Vormfelde (1971), "Klassische Konditionierung konnotativer Bedeutungen" ("Classical Conditioning of Connotative Meanings"), Zeitschrift fur Psychologie, 179. 2. 226-246.

Fishbein, M. (1973), "The Search for Attitudinal-Behavioral Consistency", Behavioral Science Foundation of Consumer Behavior, (ed.) J. S. Cohen, New York: McMillan Publishing Company.

Haber, R. N. (1981), "The Power of Visual Perceiving," Journal of Mental Imagery, 5, 1-40.

Kanter . D. L. (1981), "It Could Be: Ad Trends Flowing from Europe to US", Advertising Age, 49.

Kanungo, R. N. and S. Pang (1973), "Effects of Human Models on Perceived Product Quality", Journal of Applied Psychology, 57, 2. 172-178.

King, D. L. (1978), "Image Theory of Conditioning. Memory, Forgetting, Functional Similarity, Fusion, and Dominance", Journal of Mental Imagery, 2, 47-62.

Kroeber-Riel, W. (1980), Konsumentenverhalten (Consumer Behavior), 2nd ed., Munchen: Vahlen-Verlag.

Kroeber-Riel, W. (1983), "Effects of Emotional Pictorial Elements in Ads - Analyzed by Means of Eye Movement Monitoring", paper presented to the Proceedings of the 14th ACR Annual Conference in Chicago.

Kroeber-Riel, W. and W. von Keitz (1980), Conditioning of Attitude - A Non-informational Approach in Attitude Formation. Working Paper - International Series, Institute for Consumer and Behavioral Research, University of the Saarland, Saarbrucken, W. Germany.

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

Lastovicka, J. L. and D. M. Gardner (1979), "Components of Involvement", Attitude Research Plays for High Stakes, (eds.) J. C. Maloney and B. Silverman, Chicago: American Marketing Association, 53-73.

Leavitt, C., A. G. Greenwald and C. Obermiller (1981): "What is Low Involvement Low in?", Advances in Consumer Research, Vol. VIII, (ed.) K. B. Monroe, Arlington: Association for Consumer Research, 15-20.

Lefrancois, G. R. (1972), Psychological Theories and Human Learning: Kongors Report, Belmont: Wadsworth Publ. Co.

Lutz, R. J. (1975): "Changing Brand Attitudes Through Modification of Cognitive Structure", Journal of Consumer Research, 1, 4, 49-59.

Martin, J. (1975): "Psychophysiology and Conditioning", Research in Psychophysiology, (eds.) P. H. Venables and M. J. Christie, London: Wiley, 364-388.

Mednick, S. A., H. R. Pollio and E. F. Loftus (1973), Learning, Englewood Cliffs: Prentice Hall.

Mitchell, A. A. and J. C. Olson (1981), "Are Product Attribute Beliefs the Only Mediator of Advertising Effects on Brand Attitude?", Journal of Marketing Research, XVIII, 318-332.

Mitchell, A. A. and J. C. Olson (1977), "Cognitive Effects of Advertising Repetition", Advances in Consumer Research, IV, (ed.) W. D. Perreault, Arlington: Association for Consumer Research, 213-220.

Olson, J. C., D. R. Toy and Ph. A. Dover (1982), "90 Cognitive Responses Mediate the Effects of Advertising Content on Cognitive Structure?", Journal of Consumer Research, 9, 3, 245-262.

Petty, E. P. and J. I. Cacioppo (1981), "Issue Involvement as a Moderator of the Effects on Attitude of Advertising Content and Context", Advances in Consumer Research, VIII, (ed.) K. P. Monroe. Arlington: Association for Consumer Research, 20-24.

Ray, M. L. (1973), Marketing Communication and the Hierarchy-of-Effects, Working Paper of the Marketing Science Institute. Boston.

Ross, L. E. and S. M. Ross (1976), "Cognitive Factors in Classical Conditioning", Handbook of Learning and Cognitive Processes Vol. 3: Approaches to Human Learning and Motivation, (ed.) W. K. Estes, Hillsdale: Lawrence Erlbaum Associates, 103-129.

Rudy, J. W. and A. R. Wagner (1975), "Stimulus Selection in Associative Learning", Handbook of Learning and Cognitive Processes, Vol. 2: Conditioning and Behavior Theory, (ed.) W. K. Estes, Hillsdale, Lawrence Erlbaum Associates 269-303.

Shaffer, D. R., R. Reardon, E. G. Clary, and C. Sadowski (1982), "The Effects of Information on Perspectives and Attitude Change", Journal of Social- Psychology, 117 (1), 125-133.

Smith, G. H. and R. Engel (1968), "Influence of a Female Model on Perceived Characteristics of an Automobile", Proceedings of the 76th Annual of the American Psychological Association, 681-682.

Staats, A. W. (1968), Learning, Language, and Cognition, New York: Holt, Rinehart and Winston, Inc..

Staats, A. W. and C. K. Staats (1958), "Attitudes Established by Class.cal Conditioning", Journal of Abnormal Social Psychology, 57, 37-40.

Trommsdorff, V. and H. Schuster (1981), "Die Einstellungsforschung fur die Werbung" (Attitude Research for Advertising), Die Werbung (Advertising), Vol. 1, (ed.) B. Tietz, Munchen: Verlag Moderne Industrie.

Tucker, D. M. (1981), "Lateral Brain Function, Emotion, and Conceptualization", Psychological Bulletin, 89, 1, 19-46.

Zajonc, R. B. (1968), "Attitudinal Effects of Mere Exposure", Journal of Personality and Social Psychology Monograph Supplement, 9, 1-27.

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