Effects of Product Knowledge on Compaison, Memory, Evaluation, and Choice: a Model of Expertise in Consumer Decision-Making

ABSTRACT - A consumer information processing model is proposed whereby product class experts and novices judge a specific brand's similarity to an ideal product in different ways. By virtue of their complex knowledge structures in memory, expert consumers can process and use information about both similarities and differences between a specific brand and a perceived ideal product. Novices, though, are restricted to processing only similarity information due to the rudimentary nature of their knowledge structures. The differential use of similarity and difference information in the brand/ideal comparison process has implications for brand evaluation and choice by consumer experts and novices.

Citation:

Ann E. Beattie (1982) ,"Effects of Product Knowledge on Compaison, Memory, Evaluation, and Choice: a Model of Expertise in Consumer Decision-Making", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 336-341.

Advances in Consumer Research Volume 9, 1982      Pages 336-341

EFFECTS OF PRODUCT KNOWLEDGE ON COMPAISON, MEMORY, EVALUATION, AND CHOICE: A MODEL OF EXPERTISE IN CONSUMER DECISION-MAKING

Ann E. Beattie, Carnegie-Mellon University

ABSTRACT -

A consumer information processing model is proposed whereby product class experts and novices judge a specific brand's similarity to an ideal product in different ways. By virtue of their complex knowledge structures in memory, expert consumers can process and use information about both similarities and differences between a specific brand and a perceived ideal product. Novices, though, are restricted to processing only similarity information due to the rudimentary nature of their knowledge structures. The differential use of similarity and difference information in the brand/ideal comparison process has implications for brand evaluation and choice by consumer experts and novices.

INTRODUCTION

Many decisions rely on domain-specific knowledge: decisions as basic as category membership ("Is this a Toyota?"), and as complex as making evaluations and choices ("Do I want to buy it?"). Domain-specific knowledge has been conceptualized as expertise in cognitive and social psychology, and more specifically as product familiarity in consumer psychology. Expertise influences attention to inputs when stimuli are encountered, the processing and structuring of that information in memory, and how information is used in making decisions.

Considerable evidence indicates that expert-novice differences in decision-making are based on the representation of knowledge in memory. These representations can be viewed as schemata: "cognitive structures of organized prior knowledge abstracted from experiences with specific instances" (Fiske & Linville, 1980). Such cognitive representations are built up through accumulated experience in a domain, and with increasing expertise or familiarity, schemata change in predictable ways. Expert-novice differences apparently lie both in the amount of information within a schema, and in the organization of that information.

For instance, Chase and Simon (1973) demonstrated that chess masters were better able to remember board positions than were novices, due to "a vast, organized long-term memory of specific information about chess-board patterns" (p. 279). Similarly, a well developed product schema should contain structured knowledge about general product class information, product attributes, brands, and use information (Marks & Olson, 1981). Experts' schemata, though, are not limited to specific items of knowledge. Experts also have more associations between items and more strategies for using information. Problem solving research (e.g., Larkin, McDermott, Simon & Simon, 1977) indicates that experts categorize problems to define optimal heuristics in obtaining solutions. Consequently, consumers with expertise in a product class may well incorporate purchase criteria and decision strategies in their product schemata (Olson 1978).

Schematic content and complexity have implications for recall and decisions made on the basis of new information. Discussion in this paper will center on how schema development, or expertise, can affect consumers' attention to various aspects of a product, and provide heuristics for evaluating and selecting brands. The basic model for processing information and making decisions will involve: attention to product attributes, a comparison to memorial information, and consequent decisions.

MODEL PREVIEW

In terms of product attributes, most consumers have a general idea of what a product can provide. In fact several theories suggest that consumers have conceptual images of ideal products and their attribute combinations (for example, see Kotler, 1980 on the ideal-product model). Yet, in a complex product category, such as stereo equipment, or automobiles, experts would seem to have an advantage over novices. The advantage lies in complex knowledge structures about product classes that experts hold in memory. Product familiarity provides this knowledge and can direct attention to a product's important features.

Expertise in a product domain can also facilitate the processing of new product information. The central hypothesis to be argued here is that consumer experts can make use of both similarities and differences when comparing a specific brand to an ideal product in memory. Product class experts can survey an array of brands, identify important attributes, compare these attributes with the ideal product attributes, and process and use information about the similarities and differences of specific brands with the ideal. Novices, on the other hand, disperse attention across all features of a product. They tend to judge brands in terms of overall similaritY to a simplistic, or externally defined ideal.

The expert's ability to use both similarity and difference information in comparing products to an ideal is a critical feature in distinguishing him from the novice--and has considerable effect on decisions involving brand evaluation and choice. Evidence for this theory of consumer expertise and decision making will be discussed under each component of the model: attention, prototypical representation, and comparisons.

EXPERTISE AND ATTENTION TO PRODUCT ATTRIBUTES

As a result of accumulated experience with a product class, experts' schemata contain knowledge of which attributes of a product are the most important (Einhorn, 1975; Johnson & Russo, 1980). In viewing various brands, then, expert consumers selectively attend to those attributes they consider relevant to decision making (Gardners 1981). Novice consumers do not have the necessary knowledge to distinguish important product attributes. Instead, their attention will be captured by salient perceptual features (see McArthur, 1980; Taylor & Fiske, 1978 for reviews).

For example, suppose a racquetball expert and novice both were in the market for a new racquet. The expert consumer will notice racquet weight and grip size, as two of the most important features in a racquet. The novice, though, is more likely to give equal but cursory attention to every feature (assuming no one feature is particularly salient). The well structured racquet schema of the expert directs attention to important product features. The novice has no such guide for selectively noticing critical product attributes. If a racquet has a flashy color, or some other nonessential but salient attribute, the novice is likely to be at the mercy of such perceptual tricks.

EXPERTISE AND PROTOTYPICAL REPRESENTATION

Keeping in mind the racquetball example, consider how the expert might assess the quality of a racquet she is considering. An expert's knowledge is not limited to identifying which features merit attention. The expert also knows the values each feature should take in a high quality product; that is, experts have decision criteria.

We will argue that individuals familiar with a category hold prototypes made up of exemplary object attributes. The prototype notion stems from work on categorization processes. Rosch (1978) has defined prototypicality as the clearest cases of category membership, that is, "goodness of membership" in a category. In fact, experts seem to have representations of prototypical elements in memory, around which their domain-specific knowledge centers (e.g., Goldin, 1978; Rosch, 1975). In the context of purchasing products, it is reasonable to suppose some consumers have notions of prototypical "ideals" for particular product categories .

Specifically, consumer experts should have "ideal" prototypes structured within a product schema, based on extensive experience with a product class. These ideals identify important product attributes, the levels these attributes should take in a high quality product, associations between product attributes, and each attribute's utility. This conception of an ideal prototype for experts is similar to well developed product schemata outlined by others (see Marks 6 Olson, 1981; Johnson 6 Russo, 1980). The difference lies in the hypothesis of a specific ideal prototype to which new products are compared. In contrast to experts, a novice consumer does not have sufficient experience with a product class to have built complex knowledge structures. It is unlikely, then, that novices would hold clear conceptions of an ideal prototype in memory.

In order to engage in any kind of comparison process, novices must select an externally defined ideal; for example, by word of mouth or consultation with an expert. It is more likely, though, that novices represent the ideal as the highest price brand in a product class. In terms of price-quality relationship theory, the most expensive brand available will be perceived as having the highest quality. Novice consumers can simply compare affordable products to the most expensive brand on the market shelves.

While both experts and novice consumers can hold perceived ideals for new product comparison, novices are at a disadvantage. They are limited to an externally defined ideal, while experts hold prototypical ideals, embedded in rich knowledge structures that are based on experience with a product class. This allows experts to assess product information differently than novices in judging brand similarity to an ideal.

EXPERTISE AND COMPARISONS: SIMILARITIES AND DIFFERENCES

Experts, with complex schemata, can isolate important product attributes and focus on both similarities and differences between a specific brand and a prototypical ideal. Novices should view the comparison process differently. Without a guide to important product features, novices only have the capacity to focus on product information in terms of overall similarity to an ideal.

Concrete support for differential ability to handle similarity and difference information by experts and novices comes from work on political information processing. Fiske and Kinder (1980) suggest that the political expert possesses sophisticated and elaborate political schemata, in comparison with the political novice. Processing political information, they conclude, varies as a function of political schema complexity. Specifically, their subjects read about a country that was labeled either Democratic or Communist, but the description actually contained attributes of both ideologies. When given a free recall task, novice subjects remembered only information that was ideologically similar to the countries' political label. Experts, though, recalled information that was both similar and different from the political label they were given. Experts, with well developed political schemata, apparently could use both attribute similarities and differences in processing political information. In that research, particular schemata were cued by a label. Novices evidently attended only to information that was similar to the label, having inadequate capacity with which to process both similarities and differences.

The rich, complexly structured schemata of experts allows for fine-grained interpretation of incoming information. Experts have a superior understanding of how new information relates to the body of knowledge they already possess. Thus experts may view "difference" information in comparing a product to an ideal as particularly attention-provoking and diagnostic. As Hastie (1980) notes, well-developed schemata may enable more complex perception and comprehension strategies in accounting for difference information. Lengthy and perhaps deeper processing produces stable and enduring memory traces for information concerning differences. Novices, with very little information about a product class stored in memory, would not be expected to find difference information particularly salient. With rudimentary background knowledge, it would be difficult for them to interpret how differences between a specific brand and an ideal would affect product performance.

In addition, the elaborated schemata of experts allow them to "chunk" information (Chase & Simon, 1973; Dyer & Fiske, 1981) so that several components of information are viewed as a single representation in memory. For the expert, a single more general concept can represent several other closely associated concepts (Marks & Olson, 1980). Novices to not have complex knowledge structures that would allow information chunking. Instead, they must view each item of information as a separate unit.

Both experts and novices are limited in the amount of information that can be processed simultaneously in memory. Unless some information is transferred to long term memory, the cognitive limitations of short term memory restrict the number of cognitive units that can be held at one time (five to seven). Since experts can chunk information, though, they have the capacity to process more information in STM than do novices, who are limited to single items of information. Experts then, have the ability to simultaneously process both similarity and difference information by virtue of chunking, while novices, with simple knowledge structures, process information by single items of similarity information. Thus experts are able to process both similarity and difference information because of: 1) the way they organize incoming information, and 2) the attention given to difference information as being particularly diagnostic.

INTEGRATION WITH EXISTING THEORIES

A body of literature regarding the biasing effect of prior beliefs suggests that people have a predilection towards noting and remembering information that is consistent (similar) with their previously held hypotheses. The data generated by these theories are not as incompatible with the present analysis as they may first seem. For example, research on hypothesis-confirmation (see Snyder & Gangestad, 1981) involves giving subjects a defined hypothesis and noting the kind of information they choose to use in testing that hypothesis. Results show that subjects tend to request or remember information that would serve to confirm the hypothesis rather than disconfirm it. This research does not attempt to draw distinctions between experts and novices--deviations in individual subjects' knowledge regarding the hypothesis are not measured. Subjects are given an externally (experimenter) defined hypothesis to work with, much like the external ideal product novices are proposed to work with here. The propensity to use confirming information, then, appears perfectly reasonable.

Another set of research on illusory correlation has shown that people are biased towards over-weighting the co-occurrence of associated (similar) events. For example, Chapman and Chapman (1969) make an expert-novice distinction in their work, with the finding that both clinicians (experts) and students (novices) base their evaluation of the correlation between symptoms and test results on the association of the two, rather than on the actual co-occurrence. Subjects in these studies were comparing information with preconceived theories (the association between symptom and test result). For novices, it is not surprising that similarities are over-weighted. For expert clinicians, who have performed the task countless times, the task may be rather simplistic. To the extent that such tasks are over-learned, attention becomes minimal, and performance 18 automatic (Norman, 1976). In a case where attention is perfunctory, it would not be surprising that difference information is overlooked. Furthermore, the illusory correlation phenomena depends on attention to the occurrence and non-occurrence of events. It is difficult to equate the non-occurrence of an event with information about differences when comparing information to an ideal.

The model of expertise outlined here, with special attention to the use of similarity and difference information, is also perfectly compatible with existing theories of decision making (e.g., Bayesian, regression), research approaches to consumer evaluation and choice (e.g., information processing, attitude formation models), and biases in decision making (e.g., availability and representativeness heuristics). In fact, it outlines a bias that novice consumers are often prey to judging specific brands in terms of overall similarity to a perceived ideal product without accounting for differences.

PREDICTIONS: COMPARISONS AND JUDGED SIMILARITY IN CONSUMER DECISIONS

The expert-novice distinctions outlined 80 far will emerge most clearly in the comparison of products to an ideal. As an example, consider the hypothetical product class in Table 1. The rather obvious task for consumers in making any decisions about an array of products is to select the brand that will provide optimal performance. For simplicity's sake, assume the same ideal in the product class represents perfection for both expert and novice consumers.

TABLE 1

AN EXAMPLE PRODUCT-ATTRIBUTE MATRIX

Thus, the ideal ranks a 10 on all four product attributes. Each of the three brands available to consumers exhibits different levels on the four attributes, with a value of 9 being closer, or more similar, to the ideal than a value of 5. -Yet each brand's attribute levels sum to 28 and average 7. The expert, with a well defined product schema, knows that two attributes (A and B) are important to product quality, while two attributes (C and D) are relatively unimportant. In comparing each brand's similarity to the prototype, then, the expert will view Brand X as closest to the ideal, followed by Brand Z, and lastly Brand Y. Since all brands evidence the same mean value across attributes, the judgment of similarity rests only on the pattern of attribute values--a pattern showing some attributes to be more similar to the ideal, and some as more different.

The novice's Judgment of brand similarity to the ideal does not take attribute importance into account. When asked for similarity judgments, the novice should (mistakenly) view Brand Z as the most similar to the ideal, followed by both Brands X and Y. The novice does not perceive and use difference information in the same way that an expert does. Brand Z exhibits the same mean attribute level as Brands X and Y, but has a variance of zero. Similarity is judged equally across all attributes for the novice- and it is Brand Z that shows the most overall similarity to the ideal.

PREDICTIONS: EXPERTISE AND MEMORY

Expertise affects both the amount of new information that is recalled, and the organization of that information. Existing evidence thus far indicates that the amount product information recalled by both expert and novice consumers depends on task instructions. Johnson and Russo (1980) found that when subjects were instructed to make a choice between products, both product experts and novices remembered less than moderately experienced consumers. [These recall results replicated Bettman and Park's (1980) findings for information search. Moderately experienced consumers searched for more information than did consumer experts and novices. Bettman and Park suggest that while novice consumers have difficulty understanding product information, and experts have no need of its moderately experienced consumers can both understand and use new product information.] In contrast to this curvilinear effect, when consumers were instructed to evaluate products a linear effect was observed for expertise, such that experts always recalled more product information than novices. Johnson and Russo suggest that choice and evaluation tasks require different patterns of information processing. Consumers who make choices use phased rules that eliminate alternatives, while in evaluation, expertise leads to a highly selective search of information, a cording to attribute importance.

The decision strategies outlined by Johnson and so are perfectly reasonable in terms of the information processing view suggested here. But how does consumer memory for product information occur when no specific task instructions (choice or evaluation) are suggested prior to viewing information? Now to product attribute similarities and differences to an ideal affect recall of new information in the more ecologically valid case of no instructions? Other evidence indicates that expertise does not necessarily affect the amount of information recalled when no task instructions are given (as in a real world situation). Rather, expertise can influence the organization of new information in memory - and it is here that attention to similarities and differences in schematic processing come into play.

Recall the political expertise study by Fiske and Kinder (1980). Political experts recalled information that was both similar and different from a schematic label, while novices remembered only label-similar information. There were no differences in the total amount of information recalled e But politically expert subjects organized information in a more sophisticated manner than did the political novices. Fiske and Kinder analyzed subjects' recall of label-similar and label-dissimilar attribute clusters, controlling for total recall of each type. Politically inexpert subjects organized information rather mechanically, by similarity. Political experts organized information by a more complex criterion, clustering information by differences from the schematic label. Fiske and Kinder suggest different processing rules for experts and novices--inexperienced people may proceed through information noticing schema-similar attributes and ignoring the rest. Experts seem to notice schema-similar and different attributes, collecting together difference information.

The evidence for expertise and memory lends support to our model of comparison to an ideal. Experts compare information to a prototype based on prior experience. They should recall both prototype-similar and -discrepant information, clustering it in complex ways. An experienced consumer comparing a brand to an internal product ideal can attend to and recall attribute similarities and differences. Novices compare information to more rudimentary knowledge structures, and recall only information that is similar to what little they know. Inexperienced consumers comparing brands to an external product ideal remember only those brand attributes that are similar to the ideal.

Expertise affects the integration of information into knowledge structures, in terms of similarities and differences, as argued 50 far. It also has effects on the way product evaluations and choices are made. Since evaluations and choices indicate different underlying processes, they will be discussed separately.

PREDICTIONS: EXPERTISE AND PRODUCT EVALUATIONS

Referring back to the simplified example in Table 1, evaluation of the various brands should follow the judged similarity of brands to the ideal prototype, for both experts and novices. That is, regardless of the degree of product familiarity or expertise, consumers will evaluate the brands seen as most similar to the ideal the most highly. But expertise can have an effect on how evaluations proceed, once the similarity judgment has been made.

The schematic processing of experts influences how evaluations are derived from similarity judgments. Work in person perception on "schema-triggered affect" (Fiske, Beattie & Milberg, 1981) has shown that when someone "fits" or matches the important features of a schema developed from experience with a person category, an affective response associated with the schema is triggered for the new person. The closer a match the new person provides to the schema, the more likely the person will receive the affect linked to the schema. People who only partially match the schema elicit "mixed" affective responses. Comparison of people to schemata based on past experience provides immediate evaluations when matches occur. Schema-triggered affect is hypothesized to short circuit the more cognitively laborious task depicted by traditional models of social perception (e.g., Anderson, 1967; Schneider, 1973).

Applied to product evaluation, the development of a complex product schema allows evaluation of products to proceed on the basis of schematic match. For an expert, comparison of products to an ideal (built from experience with a product class) elicits evaluations according to degree of fit. A product with attribute levels that are similar to the ideal should elicit positive evaluations. When a product is similar on some attributes and different on others, resulting evaluations will be mixed with both positive and negative affect contributing to overall evaluations Further, experts will judge similarity on the basis of important features only. Thus, schematic match, and consequently, schema-triggered affect will depend only on important attributes.

Novices have no notion of attribute importance, and will likely evaluate brands in terms of similarity to an ideal across all attributes. While novices can have an externally defined product ideal, they have no elaborated schemata to process information and assign affect. Their evaluations of brands should rely on overall product similarity to an ideal, and require more attention and cognitive processing, as in traditional models of product evaluation (i.e., Fishbein & Ajzen, 1975).

PREDICTIONS: EXPERTISE AND PRODUCT CHOICES

Once again in reference to Table 1, consumers' brand choice should follow the patterns of judged similarity to the ideal exhibited by experts and novices. Brand choice differences for experts and novices, though, can be considered in light of rank ordering complexity. Ranked choices represent a hierarchy that can be informally regarded as "most preferred" through "least preferred." People familiar with a domain tend to use a greater number of valuative dimensions in thinking about objects in that domain. [This is conceptually similar to Linville's (1981) theory that people familiar with specific person categories use greater number of dimensions in thinking about persons from that category than do people unfamiliar with the category.]

Product experts, then, are likely to view brands in a product set with greater distinction in preference ordering. Novices may tend to clump brands together in preference ordering, viewing fewer distinctions between them. When making choices among the brands represented in Table 1, those more familiar with the product class should use more levels in ranking choice alternatives than should inexpert consumers. This would reflect the judged brand similarity to the ideal pattern suggested for experts and novices. Experts should choose Brands X, Y, and Z respectively, while novices should choose Brand Z as the most preferred, followed by both Brands X and Y as equally preferred.

SUMMARY AND IMPLICATIONS

For experts, with well defined product schemata, prior knowledge can direct focus of attention to important attributes on new products. Novices, without structured schemata to guide attention, distribute their attention diffusely over all product attributes, unless a particular feature is perceptually salient. Attributes of the specific products being surveyed are then compared to the attribute quality levels of an ideal for that product class. Novice consumers engage in the product attribute comparison process by using a price-quality relationship to identify the "ideal" product from the market shelves. Experts, however, will have built an ideal prototype for a product class within an internal knowledge structure in memory. They will have the benefits of knowing optimal attribute levels for the ideal product, and the relationship between attributes O

The increased benefits of schematic information processing that experts possess allows them to more fully use similarity and difference information. Experts can attend to and remember a brand's similarities and differences to an ideal's attribute levels. Taken in conjunction with knowing which attributes to focus on, experts' evaluations and choices concerning brands can proceed on a more discriminating level than can novices'. Novices tend to focus on ideal similarity across all of a brand's attributes. In assessing specific brands, both experts and novice consumers want the "best buys" brands that most closely match ideal product prototypes. However, because of the difference in attention to brand attribute/ideal product attribute similarities and differences, expert and novice consumers will judge brand similarity to the prototype by different criteria.

Expert consumers judge similarity to an ideal on important attributes, while novice consumers judge overall similarity across attributes. Evaluations of brands reflect these patterns, but experts' brand evaluations are triggered in memory by similarity to an internal product prototype. Choices also follow upon judged similarity, with experts evidencing more complex choice hierarchies than novices.

Organization of new product information in memory differs for expert and novice consumers Experienced consumers recall both similarities and differences between brand and ideal and cluster them accordingly. Inexperienced consumers remember and organize information only by attribute similarities. Considerable evidence suggests that consumer experts perceive, process, and use information differently than novices in assessing brands via stored prior knowledge.

The implications for marketing strategies are fairly straightforward. Considering the information processing strategies outlined here, different patterns of product attribute quality levels appeal to product-familiar and product-unfamiliar consumer segments. A product expert can use and remember product attribute quality information on important features. A manufacturer wishing to appeal to expert consumers will concentrate efforts on identifying and improving quality on these product elements. Strategies designed for product novices can concentrate on overall quality of a product, without necessarily having to expend resources to ensure top-quality on any particular set of attributes. When directing marketing efforts to product-unfamiliar consumers, an effective strategy will concentrate on perceptual saliency of the product and its packaging. Expert consumers have their own conception of important or salient product attributes, based on prior knowledge. They will be swayed more by attribute similarities and differences to a perceived ideal product. Novice consumers, without any such knowledge base, will rely on a product's overall similarities to an ideal product, in the absence of manipulated attention to specific attributes.

REFERENCES

Anderson, N. E. (1967), "Averaging Model Analysis of Set Size Effect in Impression Formation," Journal of Experimental Psychology, 75, 158-165.

Bettman, J. R. and Park, C. W. (1980), "Implications of a Constructive View of Choice for Analysis of Protocol Data: A Coding Scheme for Elements of Choice Processes," in J. C. Olson, ed., Advances in Consumer Research, Vol. 7, Ann Arbor: Association for Consumer Research, 148-153.

Chapman, L. J. and Chapman, J. (1969), "Illusory Correlations as an Obstacle to the Use of Valid Psychodiagnostic Tests," Journal of Abnormal Psychology 74, 271-280.

Chase, W. G. and Simon, R. A. (1973), "Perception in Chess," Cognitive Psychology, 4, 55-81.

Dyer, L. and Fiske, S. T. (1981), "Learning and Schema Development in Organization," unpublished manuscript, Carnegie-Mellon University.

Einhorn, H. J. (1974), "Expert Judgment: Some Necessary Conditions and an Example," Journal of Applied Psychology, 59, 562-571.

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

Fiske, S. T., Beattie, A. E. and Milberg, S. (1981), "Schema-Triggered Affect: An Application to the Initiation of Close Relationships," unpublished manuscript, Carnegie-Mellon University.

Fiske, S. T. and Kinder, D. R. (1980), "Involvement, Expertise, and Schema Use: Evidence from Political Cognition," in N. Cantor and J. Kihlstrom, eds., Personality, Cognition, and Social Interaction, Hillsdale, N. J.: Erlbaum

Fiske, S. T. and Linville, P. W. (1980), "What Does the Schema Concept Buy Us?," Personality and Social Psychology Bulletin, 6, 543-557.

Gardner, M. P. (1981), "An Examination of the Effects of Prominence in an Advertisement on the Criteria Used for Brand Evaluation," doctoral dissertation, Carnegie-Mellon University.

Goldin, S. E. (1978), "Memory for the Ordinary: Typicality Effects in Chess Memory," Journal of Experimental Psychology: Human Learning and Memory, 4, 605-616.

Hastie, R. (1980), "Schematic Principles in Human Memory," in E. S. Higgins, C. P. Herman and M. P. Zanna, eds., Social Cognition: The Ontario Symposium, Hillsdale, N. J.: Erlbaum.

Johnson, E. J. and Russo, J. E. (1981), "Product Familiarity and Learning New Information," in R. Monroe, ed., Advances in Consumer Research, Vol. 8, Ann Arbor: Association for Consumer Research.

Kotler, P. (1980), Marketing Management: Analysis, Planning, and Control, Englewood Cliffs, N. J.: Prentice-Hall.

Larkin, J., McDermott , J., Simon, D. and Simon, H. A. (1980), "Expert and Novice Performance in Solving Physics Problems," Science, 208, 1335-1342.

Linville, P. W. (1981), "Schema Complexity and Valuative Extremity in Age-Based Stereotyping," unpublished manuscript, Yale University.

Marks, L. J. and Olson, J. C. (1981), "Toward a Cognitive Structure Conceptualization of Product Familiarity," in R. Monroe, ed., Advances in Consumer Research, Vol. 8, Ann Arbor: Association for Consumer Research.

McArthur, L. Z. (1980), "What Grabs You? The Role of Attention in Impression Formation and Causal Attribution," in E. T. Higgins, C. P. Herman and M. P. Zanna, eds., Social Cognition: The Ontario Symposium, Hillsdale, N. J.: Erlbaum.

Norman, D. A. (1976), Memory and Attention: An Introduction to Information Processing, N. Y.: Wiley and Sons, Inc.

Olson, J. D. (1978), "Inferential Belief Formation in the Cue Utilization Process," in H. D. Hunt, ed., Advances in Consumer Research, Vol. 5, Ann Arbor: Association for Consumer Research.

Rosch, E. (1975),"Cognitive Representations of Semantic Categories," Journal of Experimental Psychology: General, 104, 192-233.

Rosch, E. (1978), "Principles of Categorization," in E. Rosch and B. R. Lloyd, eds., Cognition and Categorization, Hillsdale, N. J.: Erlbaum.

Schneider, De J. (1973) 9 "Implicit Personality Theory: A Review," Psychological Bulletin, 9, 294-309.

Snyder, M. and Gangestad, S. (1981), "Hypothesis-Testing Processes," in W. J. Ickes and R. F. Ridd, eds., New Directions in Attribution Research, Vol. III, Hillsdale, N. J.: Erlbaum.

Taylor, S. E. and Fiske, S. T. (1978), "Salience, Attention, and Attribution: Top of the Head Phenomena," in L. Berkowitz, ed., Advances in Experimental Social Psychology, NY: Academic Press.

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