The Flip Side of the Persuasive Equation? Does a Product Influence a Spokesperson's Public Image

Tina Kiesler, University of Southern California
ABSTRACT - Previous research on person perception and impression formation provides the theoretical background for an experiment to assess the influence of three variables on consumers' impressions of a spokesperson: the type of product endorsed, the amount of meaningful information consumers have about the spokesperson, and the presence or absence of a known celebrity as the spokesperson. Contrary to expectations, little evidence was found for the influence of any of these variables on consumers' impressions of spokespeople for two different products.
[ to cite ]:
Tina Kiesler (1988) ,"The Flip Side of the Persuasive Equation? Does a Product Influence a Spokesperson's Public Image", in NA - Advances in Consumer Research Volume 15, eds. Micheal J. Houston, Provo, UT : Association for Consumer Research, Pages: 62-68.

Advances in Consumer Research Volume 15, 1988      Pages 62-68

THE FLIP SIDE OF THE PERSUASIVE EQUATION?

DOES A PRODUCT INFLUENCE A SPOKESPERSON'S PUBLIC IMAGE

Tina Kiesler, University of Southern California

ABSTRACT -

Previous research on person perception and impression formation provides the theoretical background for an experiment to assess the influence of three variables on consumers' impressions of a spokesperson: the type of product endorsed, the amount of meaningful information consumers have about the spokesperson, and the presence or absence of a known celebrity as the spokesperson. Contrary to expectations, little evidence was found for the influence of any of these variables on consumers' impressions of spokespeople for two different products.

INTRODUCTION

Advertising is a fundamental source of information to consumers. Each day we receive advertising messages from numerous sources. Whether it is Bill Cosby touting the joys of eating Jello Pudding or a "typical" housewife marveling at the wonders of a particular laundry detergent, spokespeople are an important ingredient in an advertising campaign. To date, much of the relevant research on spokespeople has examined the credibility of spokespeople as persuasive communicators. Factors such as expertise, trustworthiness and attractiveness have been found to enhance a spokesperson's influence on others' attitudes and behavior (see Cooper and Croyle 1984; Petty and Caccioppo 1981 for reviews).

The research on spokespeople has been one sided. Communicators have been concerned with finding a persuasive spokesperson to "sell" their product, service or message. This paper examines the influence that a particular product may have on consumers' perceptions of the spokesperson marketing that product. If the product does influence consumers' perception of spokespeople, it is important that prospective spokespeople realize such effects. Further, when a spokesperson also serves as an advocate for other products there may be an interaction effect such that those other products may differentially influence consumers' impressions of the spokesperson.

Research on impression formation and change has focussed on the effects of schemas, or stereotypes, on individuals' perceptions of others. People often form impressions of, and make inferences about, others based on stereotypical or categorical cues such as occupation, age, race, and gender (Fiske and Taylor 1984).

These cues are particularly influential when a person is first forming an impression of another. For example, suppose a male meets a female for the first time at a party. His impression of her may differ depending on information that was provided prior to the introduction. The female may be intelligent, attractive and outgoing. However, if the male were told that she is a professor, his impression of her is likely to differ from one formed after being told that she is a fashion model. In the first case a professor schema will be evoked and particular attention will be paid to the fact that she is intelligent. Upon knowing this, he may infer that she reads a lot, enjoys philosophical conversation, and possesses other attributes consistent with the impression of professors as a group of people. However, if he is told that she is a fashion model, he may base his impression on another schema--that of fashion models. Her attractiveness now becomes more relevant and he may make inferences about her vanity, superficiality, and other factors that may be relevant to a schema of fashion models.

This example highlights how schemas can "guide" information processing. That is, providing a schema label (such as "professor" or "fashion model") influences the inferences made about a person. Research indicates that once an individual places a person or objects into a particular category (schema), the person is likely to interpret information in a way that is consistent with the schema label (Zadny and Gerard 1974). They tend to remember items that- are consistent with the schema label and often fail to remember items that are irrelevant to the label (Bower, Black and Turner 1979; Hastie and Kumar 1979; Srull 1981; among others). They even claim to remember information which, although consistent with the category, was never given about the particular stimulus.

For example, Cantor and Mischel (1977) gave subjects lists of adjectives that described a person. In one case subjects were told that the person described was an extrovert. The lists contained words that were moderately related to extroversion (i.e., entertaining). On a later recognition task, people falsely recognized prototypical words which were never given in the list (i.e., outgoing).

A schema has both facilitative as well as biasing effects on the impression formation process. It facilitates the process by implying a set of expectations about the person which can simplify information processing. For example, a spokesperson for athletic equipment may be categorized as an athlete and thus may be considered to be athletically skilled, healthy, and competitive. On the other hand, the label may bias perceptions because information that is not directly relevant to the label may not be fully processed (i.e., the athletic spokesperson may also be compassionate, honest, etc.) and thus does not influence one's impression of the person. These biases may five rise to incomplete, inaccurate, or unrealistic beliefs and expectations about a person.

A product endorsement may serve to cue a particular schema and thus may influence impressions in a manner consistent with the schema. If a product serves as a schema cue for the impression formation process, it then becomes important to understand the process by which impressions might be influenced by the product(s) one chooses to endorse.

HYPOTHESES

Past research on the effects of stereotyping indicates that once an individual is perceived to be a member of a particular category, stereotypical attributes are imputed to the individual. For instance, Taylor and her colleagues (Taylor, Fiske, Etcoff and Ruderman 1978) provide evidence that stereotypical gender traits are attributed to an individual based solely on one's knowledge of their gender. Taylor et al. provided subjects with taped conversations between men and women. They had male and female versions of each speaker (in which speech speed and intonations were matched). Later the men were rated as more influential, confident, analytic, and negative. Men were also rated as less warm and sensitive than women.

Based on findings such as these, it is expected that subjects will form an impression of a previously unknown spokesperson based on whatever schema is cued. The cued schema may then influence later inferences about the person. When forming an impression of a spokesperson, the product endorsement may serve as the schema cue. In order to form a meaningful impression of the spokesperson, subjects will rely on the evoked schema. Later judgments will reflect the schema used to process the information. Therefore, a subject's evaluations of a spokesperson's personality attributes will be influenced by the type of product that he or she endorses.

Expectation 1: Subjects will later rate a spokesperson higher on personality traits that are characteristic of the schema that was cued by the product than will subjects who received a different (product) schema cue.

A great deal of literature in marketing indicates that a consumer's level of knowledge about a product class influences the way she processes product information as well as the way she makes judgments and choices (Bettman and Park 1980; Edell and Mitchell 1978; Johnson and Russo 1981, 1984; Park and Lessig 1981; Sujan 1985; among others).

Knowledge may influence initial impressions as well. For example, Linville (1982) found that individuals made less extreme evaluations about people in their own groups ("in-group members") than they did about people in groups other than their own ("out-group members"). She attributed this to the complexity of the respondents' knowledge structures. People have more complex knowledge structures about the groups to which they belong than they do about groups that they do not belong to.

In the context of the present experiment, it may be that people who have less meaningful information about a spokesperson will make more extreme judgments about the person. Further, since they have less relevant information, they may be more dependent upon the given schema cue as a guide for their inferences. They are expected to rely upon stereotypic expectations regarding the spokesperson and thus they may later judge the person to possess "more" of personality traits that are characteristic of the schema cue than will subjects who have more meaningful information.

Expectation 2: Subjects with less meaningful information about the spokesperson will rate the spokesperson higher on personality traits that are characteristic of the cued schema than will subjects with more meaningful information.

Frequently, celebrities rather than "ordinary" people are used as spokespeople for products. It is likely that a product endorsement may have a different influence on the impressions formed of celebrity spokespeople than non-celebrity spokespeople since the celebrity already has a strong public image. Consumers have cues other than just the product from which to form an impression of a celebrity. Thus the schema cue (product endorsement) may have less influence in guiding consumers' impressions of a spokesperson who is a known celebrity than an unknown spokesperson. Furthermore, since consumers may already have accumulated information about the celebrity spokesperson, the amount of meaningful information that is provided to them may DO longer be relevant. Thus, it is expected that

Expectation 3: Expectation 1 and Expectation 2 will be descriptive of consumers' judgments regarding an unknown spokesperson. The schema label and the amount of information will have less influence on subjects' judgments of a celebrity spokesperson.

Expectations 1-3 posit that consumers will infer certain personality traits to be descriptive of a spokesperson depending on the product that is endorsed, the amount of information given, and whether or not the spokesperson is a celebrity. The basis for the expectations is that the product serves as a schematic cue for processing information and making inferences about a person.

In a sense, the schema cue triggers a category, or stereotype, that one has formed over time and has "stored" in memory. When a person perceives another person (or an object) as belonging to a particular category, a person may draw on their past knowledge and experience regarding members of that category and may then make inferences about the person based on the stereotype (as in the professor/fashion model example). A great deal of research indicates that schematic cues do guide information processing in this manner.

Much of the evidence for such effects has been provided by examinations of recall and recognition measures gathered in experimental settings. Individuals seem to show greater recall of information that is consistent with an evoked schema than of information that is irrelevant to the schema. For example, Cohen (1981) showed subjects a videotape of a woman and her husband eating dinner and having an informal birthday party. In one condition, subjects were told that the woman worked as a librarian. In another condition, subjects were told that she was a waitress. When subjects were given the librarian cue, they more accurately remembered items that were characteristic of that occupational role (i.e., listens to classical music, wears glasses) than did subjects given the waitress cue. Subjects given the waitress cue more accurately remembered "waitresslike" items (i.e., she drinks beer and watches television).

It is also likely that subjects might recall item that are relevant to the schema cue but are not provided in a description of the spokesperson. Recall intrusions of this type are an indication of subjects' elaboration of the information. For example, if one believes that used car salesmen are generally sleazy, one may incorrectly recall being told that the used car salesman down the street is sleazy when in fact he was not told that. Both types of recall--better recall of schema-consistent information and schematic intrusions--are considered to be evidence of schematic processing (Bower, et al. 1979). Thus it is expected that:

Expectation 4a: Recall of traits characteristic of one schema cue will be greater for those who were given that cue than for subjects given a different cue.

Expectation 4b: Subjects will misremember more items that are characteristic of the schema cue given to them than will subjects who are given a different schema cue.

Since subjects who receive less meaningful information about the spokesperson are expected to rely upon the schema cue to a greater extent, the recall of schema-consistent information and schematic intrusions are expected to be particularly evident for subjects who receive less meaningful information.

Expectation 4c: Expectations 4a and 4b will be more prevalent in the recall of subjects who are given little meaningful information about the spokesperson than for subjects given more meaningful information.

When the spokesperson is a celebrity, however, subjects are expected to rely upon the public image of the celebrity to a greater extent than the schema cue or the amount of information (as stated in Expectation 3). In a sense, the celebrity's public image may act as a cue for processing information about the spokesperson. If subjects are relying upon the image of the celebrity we may expect to see recall intrusions that are relevant to the celebrity's public image.

Expectation 4d: Subjects who are given a known celebrity as a spokesperson will show (a significant number of) recall intrusions that are consistent with the image.

METHOD

In this study, spokespeople were selected for two different causes: the homeless and a defense contractor. These two types of "product" endorsements were used because they were found to imply very different schemas. They are also products where some type of emotional involvement is assumed on the part of the spokesperson. One may be able to make stronger inferences about a spokesperson for the homeless, an endeavor in which some emotional involvement is assumed, than if the person is a spokesperson for products such as coffee or deodorant. These product cues should serve as a strong test of the expectations.

A pretest was conducted with 20 subjects to ascertain the characteristic, or schema-relevant, personality traits of each type of spokesperson. Traits were chosen for use in the experiment if they were perceived by the pretest subjects to be characteristic of one type of spokesperson and not the other.

SubJects

Seventy-two subjects volunteered to participate in the study. They responded to flyers posted on the UCLA campus. All subjects were paid 10 dollars for their participation in the study.

Independent Variables

Three independent variables were manipulated in this study. There were two levels of each variable, resulting in an eight cell design (n = 9 per cell). The tree independent variables are (1) The schema cue (type of product endorsed): Homeless vs. Defense Contractor, (2) The amount of meaningful information in the description: Less Meaningful vs. More Meaningful, and '3) The image of the spokesperson: Celebrity vs. Unknown spokesperson.

Procedure

Subjects were told that they would be participating in a study of how consumers form impressions of spokespeople and products. They were told that they would receive a booklet in which a spokesperson would be described. They were asked to read the description and then to answer the questions that followed. Each booklet contained a written description of a spokesperson on the first page and relevant measures on the following pages. Subjects were randomly assigned to an experimental condition.

In one condition, subjects were told on the first page of their booklet that they would be reading a description, and forming an impression, of a spokesperson for the homeless. In another condition they were told that they would be reading a description, and forming an impression, of a spokesperson for a defense contractor.

Additionally, half of the subjects were told that the spokesperson was Bob Hope and the other half were not given any specific information about the spokesperson's identity. Bob Hope was used as a celebrity image because pretest subjects felt that he could realistically be a spokesperson for either type of product and he is also a well-known figure with an identifiable image. Thus, the first page of the booklet determined the schema cue that each subject received as well as the presence or absence of a celebrity image.

The spokesperson was described to the subject on the second page of the booklet. Subjects received one of two versions of the description: a description with less meaningful information or a description with more meaningful information. The description with less meaningful information contained four personality traits which were relevant to a spokesperson for the homeless (Spokesperson H traits), four personality traits which were relevant to a spokesperson for a defense contractor (Spokesperson DC traits), and 12 traits which were irrelevant to both types of spokespeople. This description was intended to have less meaningful information than the alternative description which contained 10 Spokesperson H traits, 10 Spokesperson DC traits, and no traits irrelevant to both types of spokespeople. Thus both descriptions had the same amount of information, what differed was the meaningfulness of this information with respect to the two possible schemas used as cues. The order in which the traits were presented in the description was varied to control for possible order effects in impression formation and in recall.

At the conclusion of the description, subjects were asked to write down their impression of the spokesperson. This measure was given to provide closure for the subjects and to solidify the impressions they had formed. As far as subjects knew, this was the conclusion of the study.

The dependent measures of interest were asked after an unrelated intervening task (a questionnaire for a separate study by another researcher) had been completed. The intervening task was necessary to make sure that subjects were not basing their judgments and recall on information in short-term memory. The schema cue provided at the beginning of the study, as well as the amount of information given and the presence or absence of a celebrity image, was expected to influence judgment and recall of particular attributes of the spokesperson over time

Dependent Variables

Two dependent variables were measured in the study: subjects' ratings of the spokesperson on eight different personality traits (relevant to Expectations 1-3) and subjects' recall of the information given to them in the original description of the spokesperson (relevant to Expectations 4a-4d).

Subjects were asked to rate the spokesperson on a 9-point scale on eight traits (where 9 indicated that the personality trait was very representative of the person and 1 indicated that the trait was not at all representative of the person). Four of the traits used in the judgment task were characteristic of spokespeople for the homeless (idealistic, sympathetic, individualistic, and open-minded) and four of the traits were characteristic of spokespeople for a defense contractor (clean, educated, self-critical, and definite). A mean rating was computed for each of the two sets of personality traits and these two ratings were then used as dependent measures in the analysis.

The personality traits used in the judgment task were also elicited via the pretest and differed from those given in the original description. All eight of these traits were perceived by pretest subjects to be very characteristic (mean rating of 6.3 or greater on a nine-point scale) of one type of spokesperson and not the other (mean rating of 5.0 or less).

Subjects were then asked to recall the information given to them in the description of the spokesperson at the beginning of the study. They were then asked a variety of questions to ascertain their degree of knowledge about defense contractors and homeless people. Upon completion of the questionnaire, subjects were thanked, debriefed, and were then paid for their participation.

RESULTS AND DISCUSSION

The data were analyzed using a 2 (Schema Cue: Homeless vs. Defense Contractor) X 2 (Information: More vs. Less) X 2 (Image: Celebrity vs. Unknown) completely randomized analysis of variance. There were nine observations per cell. Subjects' personality trait judgments and recall were used as dependent variables in separate analyses.

An average rating was computed for the four traits relevant to spokespeople for the homeless and for the four traits relevant to spokespeople for a defense contractor. These two averages were the dependent variables for judgments. Recall was analyzed according to the number of correct traits recalled which were relevant to each type of spokesperson, the number of memory intrusions relevant to each type of spokesperson, and the number of memory intrusions characteristic of Bob Hope's celebrity image (i.e., funny). The recall data were coded by a judge who was blind to the treatment condition of the subjects.

Results of Judgment Data

It was expected that subjects in the homeless condition would rate the spokesperson higher on personality traits that pretest subjects considered characteristic of such spokespeople (Spokesperson H traits) than would subjects who were given the defense contractor cue. These ratings were expected to be more extreme for subjects given less meaningful information about the spokesperson. However, the cue and the amount of information were not expected to be significant influences upon the impression formed by subjects who had a celebrity in mind at the time they read the description of the person.

The analysis was performed separately on the spokesperson H traits and the spokesperson DC traits.

Traits Relevant to Spokespeople for the Homeless: A significant Amount of Information x Image x Cue interaction was found (F [1,64] = 6.31, p<.05). There was also a significant main effect for Image (F [1,64] = 3.93, p<.05) such that subjects in the unknown spokesperson condition rated the spokesperson higher on Spokesperson H traits (mean = 5.84) than did subjects in the celebrity spokesperson condition (mean = 5.32). Of greater interest is the three-way interaction and these means are graphed in Figure 1.

The results were not as expected. The only significant difference was found for subjects who received a defense contractor cue and more meaningful information. In this situation, subjects rated the spokesperson higher on Spokesperson H traits when they had an unknown spokesperson to consider than when Bob Hope was presented as the spokesperson (t [16] = 4.57, p<.01). This pattern was not shown for subjects in the homeless cue condition, however.

Subjects given the homeless cue did not perceive the Spokesperson H traits to be any more characteristic of the spokesperson than did subjects given the defense contractor cue (contrary to Expectation 1). Furthermore, subjects with less meaningful information did not make more extreme judgments than did subjects with more meaningful information (contrary to Expectation 2). The presence of a celebrity spokesperson did not influence =the judgments of subjects given the homeless cue (in part, contrary to Expectation 3) but did influence the judgments of subjects given the defense contractor cue. In this case, the presence of absence of a celebrity as the spokesperson differentially influenced the judgments of only the subjects given more meaningful information.

Traits Relevant to Spokespeople for Defense Contractor: A significant main effect for cue was found (F [1,64] = 5.26, pc.OS) such that subjects who were given the defense contractor cue rated the spokesperson higher on Spokesperson DC traits (mean = 7.35) than subjects given the homeless cue (mean = 6.70). There was also an Image x Cue interaction (F [1,64] = 7.20, pc.01). The interaction is depicted in Figure 2.

Subjects given the defense contractor cue rated the spokesperson higher on relevant traits when an unknown spokesperson was provided than when the spokesperson was a celebrity (t [34] = 2.31, p<.05). The presence of an image did not influence subjects' ratings on these traits when they had been given the homeless cue (t [34] = 1.59, ns). Evidently, subjects given the defense contractor cue were sensitive to the presence or absence of a celebrity image, whereas those given the homeless cue were not. This was true for their judgments regarding Spokesperson DC traits (regardless of the amount of information) as well as Spokesperson H traits (only when given more information).

FIGURE 1

THE THREE-WAY INTERACTION BETWEEN CUE, IMAGE, AND AMOUNT OF INFORMATION FOR SPOKESPERSON H TRAITS

FIGURE 2

IMAGE x CUE INTERACTION FOR JUDGMENTS REGARDING SPOKESPERSON DC TRAITS

Results Of Recall Data

Manipulation Check: Amount of Information: It was assumed that subjects in the more information condition would have significantly more information in memory with which to form an impression than would subjects in the less information condition. The manipulation can be verified if subjects in the more information condition recall more (correct) attributes than subjects in the less information condition. The manipulation was verified by a significant main effect for amount of information for recall of Spokesperson H traits (F [1,64] = 9.04, p<.01) as well as the recall of Spokesperson DC traits (F [1,64] = 8.97, p<.01). In both cases, subjects in the more information condition recalled more information than subjects in the less information condition.

Correct Recall and Spokesperson Intrusions: It was hypothesized that subjects' recall of the spokesperson would provide some additional support for the hypotheses regarding judgments. If subjects are relying on the schema cue then they should recall more of the schema-consistent traits (Expectation 4a) as well as show more memory intrusions for traits relevant to the schema (Expectation 4b) than subjects given the other schema cue. It was also thought that the amount of information given to subjects would interact with the influence of the schema cue (Expectation 4c).

Expectations 4a-c were not supported by the data. There was no significant main effect for schema cue nor was there a Cue x Amount of Information interaction for recall of Spokesperson H or DC traits. Thus, there is little indication that consumers were relying upon the schema cue to form an impression of the spokesperson.

Image Intrusions: Subjects given a celebrity image for a spokesperson are expected to rely on that image when forming an impression of the person. If they are relying upon the image, they should show memory intrusions that are consistent with the celebrity's image. Subjects in the celebrity condition did show intrusions consistent with Bob Hope's image (t [35] = 2.24, p<.05). This indicates that subjects were utilizing the image when processing the information about the spokesperson.

CONCLUSION

A great deal of research has examined the impression formation process. That people form impressions and make inferences based on stereotypic or categorical cues is not new (Taylor and Crocker 1981). However, the range of cues that have been examined is somewhat limited. We do know, for example, that people often form impressions of others based on age, race, gender, and occupation (see Fiske and Taylor 1984). People do seem to be quite sensitive to salient cues when they form impressions of others. This study attempted to extend that research to an advertising domain and to examine the influence of a very salient cue, a product, OD the impression formation process for a particular type of person, a spokesperson.

Interestingly, the expected findings were not obtained. Since the expectations posited in the paper were based upon research on person perception in general, the findings indicate that there is something unusual about the way consumers form impressions of spokespeople.

The failure to support the hypotheses might indicate that people perceive the product cue as a less meaningful cue than other cues that are often used in person perception research (such as occupation or gender). It could also be that the impression formation process for spokespeople is very different than that for people in other occupational and societal roles. Spokespeople are often advertising a product because they want the paycheck that is associated with such work. In fact, David Ogilvy, of Ogilvy & Mather, has stopped using celebrities in his advertisements because he claims that the audience assumes that the celebrity has been bought off (Ogilvy 1983). Ogilvy's belief is substantiated by research done by Video Storyboard Tests, an advertising research company (see Sherman 1985).

In part, to account for just such an effect, the celebrity image and no image condition were included in the study. Also, the products chosen for this study were ones where some type of emotional involvement is assumed on the part of the spokesperson. Yet, the results did not show a strong effect for the type of product, the image, or the amount of information in the impression formation process for spokespeople.

Assuming no fatal flaws in the study, the results indicate that further examination of the impression formation process is needed. Spokespeople may be perceived to be special cases when it comes to forming an impression of them. For example, people learn about spokespeople by seeing them or hearing them communicate through some medium. Further research should examine the differences between media-mediated impression formation and interpersonal-mediated impression formation.

The cues used in the impression formation process for spokespeople may not be product-related cues. Person perception literature has examined only a limited number of cues as factors in the impression formation process. Further research might focus on different cues that might be used in such a situation. An in-depth, perhaps even qualitative, study would be a positive next step in this direction. What are the factors that influence a spokesperson's public image? The public should be able to provide some insight in that respect.

This also highlights the artificiality of the experimental study in this situation. It is not unusual for experimenters to use paper and pencil measures in laboratory studies of person perception and impression formation, and to achieve significance in their results. Much of the research on person perception has been done with paper and pencil measures (however, see Cohen 1981 and Zadny and Gerard 1974 for a different approach). The use of paper and pencil measures in an experimental setting, however, may be more appropriate for typical impression formation studies than for an examination of others' perceptions of spokespeople due to the media-related nature of their presentation. Special care should be given to matching the presentation of the stimuli with its typical presentation in the real world.

In general, our knowledge about impression formation and change has been built from one type of research setting. Perhaps future studies should examine the impression formation process in various settings. More realistic and involving situations may bring new insight to the phenomenon.

References

Bettman, James R. and C. Whan Park (1980), "Effects of Prior Knowledge and Experience and Phase of the Choice Process on Consumer Decision Processes: A Protocol Analysis," Journal of Consumer Research, 7 (December), 234-248.

Bower, Gordon H. John B. Black, and Terence J. Turner (1979), "Scripts in Memory for Text," Cognitive Psychology, 11, 177-220.

Cantor, Nancy and Walter Mischel (1977), "Traits as Prototypes: Effects on Recognition Memory," Journal of Personality and Social Psychology, 35 (January), 38-48.

Cohen, Claudia E. (1981), "Person Categories and Social Perception: Testing Some Boundaries of the Processing Effects of Prior Knowledge," Journal of Personality and Social Psychology, 40 (3), 441-452.

Cooper, Joel and Robert T. Croyle (1984), "Attitudes and Attitude Change," in Annual Review of Psychology, Vol. 35, eds. M. R. Rosenzweig and L. W. Porter, Palo Alto, CA: Annual Reviews, Inc., 395-426.

Edell, Julie and Andrew Mitchell (1978), "An Informational Processing Approach to Cognitive Responses," in Research Frontiers in Marketing; Dialogues and Directions: 1978 Educators' Conference Proceedings, ed. S. Jain, Chicago: American Marketing Association, 178-183.

Fiske, Susan T. and Shelley E. Taylor (1984), Social Cognition, Reading, MA: Addison-Wesley Publishing Co., Inc.

Hastie, Reid and Purohit Kumar (1979), "Person Memory: Personality Traits as Organizing Principles in Memory for Behaviors," Journal of Personality and Social Psychology, 37 (January), 27-38.

Johnson, Eric J. and J. Edward Russo (1981), "Product Familiarity and Learning New Information," in Advances in Consumer Research, Vol. 8, ed. K. B. Monroe, Ann Arbor, Ml: Association for Consumer Research, 151-155.

Johnson, Eric J. and J. Edward Russo (1984), "Product Familiarity and Learning New Information," Journal of Consumer Research, 11 (June), 542-550.

Linville, Patricia W. (1982), "The Complexity-Extremity Effect and Age-Based Stereotyping," Journal of Personality and Social Psychology, 42 (February), 192-211.

Ogilvy, David (1983), Ogilvy on Advertising, New York: Vintage Books.

Park, C. Whan and V. Parker Lessig (1981), "Familiarity and its Impact on Consumer Decision Biases and Heuristics," Journal of Consumer Research, 8 (September), 223-230.

Petty, Richard E. and John T. Caccioppo (1981), Altitudes and Persuasion: Classic and Contemporary Approaches, Dubuque, Iowa: Wm. C. Brown Company Publishers.

Sherman, Stratford P. (1985), "When You Wish Upon a Star," Fortune, 112 (August 19), 66-71.

Srull, Thomas K. (1981), "Person Memory: Some Test of Associative Storage and Retrieval Models," Journal of Experimental Psychology: Human Learning and Memory, 7 (November), 440-463.

Sujan, Mita (1985), "Consumer Knowledge: Effects on Evaluation Strategies Mediating Consumer Judgments," Journal of Consumer Research, 12 (June), 31-46.

Taylor, Shelley E. and Jennifer Crocker (1981), "Schematic Bases of Social Information Processing," in Social Cognition: The Ontario Symposium, Vol. 1, eds. E. T. Higgins, C. P. Herman and M. T. Zanna, Hillsdale, NJ: Lawrence Erlbaum Associates, 89- 134.

Taylor, Shelley E., Susan T. Fiske, Nancy L. Etcoff, and Audrey J. Ruderman (1978), "Categorical and Contextual Bases of Person Memory and Stereotyping," Journal of Personality and Social Psychology, 36 (July), 778-793.

Zadny, Jerry and Harold B. Gerard (1974), "Attributed Intentions and Informational Selectivity," Journal of Experimental Social Psychology, 10 (January), 34-52.

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