Assessment of Consumer Knowledge and Its Consequences: a Multi-Component Approach

Aurier Philippe, University of Montpelier
Paul-Valentin Ngobo, University of Montpelier
ABSTRACT - Researchers have found that experts and novices differ in the amount, content and organization of their knowledge and as a result exhibit large variances when they perform some product-related tasks. However, no agreement has been reached on how to assess consumer knowledge. This paper reexamines the knowledge assessment issue, using wine as a focal product category. Four components of consumer knowledge are identified and validated: familiarity, objective product class information, objective and subjective expertise. Moreover, we find that these components have different effects on cognitive effort, category graded structure, prepurchase information search, choice attributes and non-aided recall tasks.
[ to cite ]:
Aurier Philippe and Paul-Valentin Ngobo (1999) ,"Assessment of Consumer Knowledge and Its Consequences: a Multi-Component Approach", in NA - Advances in Consumer Research Volume 26, eds. Eric J. Arnould and Linda M. Scott, Provo, UT : Association for Consumer Research, Pages: 569-575.

Advances in Consumer Research Volume 26, 1999      Pages 569-575

ASSESSMENT OF CONSUMER KNOWLEDGE AND ITS CONSEQUENCES: A MULTI-COMPONENT APPROACH

Aurier Philippe, University of Montpelier

Paul-Valentin Ngobo, University of Montpelier

ABSTRACT -

Researchers have found that experts and novices differ in the amount, content and organization of their knowledge and as a result exhibit large variances when they perform some product-related tasks. However, no agreement has been reached on how to assess consumer knowledge. This paper reexamines the knowledge assessment issue, using wine as a focal product category. Four components of consumer knowledge are identified and validated: familiarity, objective product class information, objective and subjective expertise. Moreover, we find that these components have different effects on cognitive effort, category graded structure, prepurchase information search, choice attributes and non-aided recall tasks.

The role of consumer knowledge on behavior has been extensively investigated by consumer researchers, with focal issues ranging from knowledge measurement (Brucks, 1986), the effects of knowledge on search (Bettman and Park, 1980), product judgments (Maheswaran et al. 1996) and choice (Mitchell and Dacin, 1996). Researchers have found that experts and novices differ in the amount, content and organization of their knowledge and as a result exhibit large variances when they perform some product-related tasks. However, in spite of the importance of this issue and the fact that some interesting results have been obtained, no agreement has been reached on how to assess consumer knowledge. Different measures have been used to operationalize conceptually distinct dimensions of knowledge (Brucks, 1986; Park et al. 1994). Almost every study of consumer knowledge has come with its own scale. Some studies have measured familiarity, past experiences (Bettman and Park, 1980) while others have measured perceived or objective knowledge (Brucks, 1985). This however is less likely to favor the generalizability of the findings (see Fiske et al. 1994).

The purpose of this paper is to reexamine the knowledge assessment issue. We extend current work (see for example Mitchell and Dacin, 1996; Park et al. 1994) by adopting a multi-component approach to consumer knowledge. Knowledge is a multidimensional construct because there are different types of knowledge (e.g. objective vs. Subjective, brand vs. product type knowledge). Consequently, the relationship between knowledge and product-related tasks might depend upon considered components (Bettman, et al., 1991; Brucks, 1985; Cole et al. 1992).

In this paper, we address two issues: (1) the operationalization of consumer knowledge and (2) its relationship with some product-related tasks. Moreover, our research hypotheses are tested using wine as the focal product category. Wine was chosen because it is a complex product, which demonstrates a great deal of variety. Wine is also an involving product where information search, expertise, perceived risk and opinion leadership play important roles (e.g. wine taster expert).

The rest of the paper flows as follows. The next section presents the theory and the hypotheses. That is we propose a theoretical structure of consumer knowledge and relates it to some wine-related tasks. The second section presents the study design. Research hypotheses are tested in the third section. The paper concludes with a discussion of the results.

1. THEORY

The Content and Organization of Consumer Knowledge

On the basis of our literature review, we distinguish two major dimensions of knowledge: familiarity (as accumulated consumption experiences) and product knowledge (as the sum of product class information and rules stored in memory). These two dimensions are supposed to have specific components (see Figure 1).

Familiarity

It is the behavioral component of knowledge. It may have two componentsBdepth and breathBas suggested by Zaichowsky (1984) and it results from the (mere) consumer activity with the product .

Product knowledge

Table 1 specifies the content of the second dimension of consumer knowledge. Two components can be obtained by opposing declarative knowledge (product class information) vs. procedural knowledge (expertise) (see Anderson’s 1983). The declarative knowledge (i.e. knowing facts) is at the basis of procedures development (i.e. possessing skills or expertise).

Product class information (PCI) (Park and al. 1994) is knowledge in terms of quantity of information stored in memory regardless of its truthfulness. It can be analyzed in terms of objective product class information (e.g. the actual nmber of brands and attributes listed) and in terms of subjective product class information (e.g. consumer perception of his/her product class information).

Expertise consists of different rules regarding the use of declarative knowledge or "production rules" referring to real objects or concepts. Expertise is thus a function of declarative knowledge (product class information) but not similar to it. Two people may have the same quantity of information and exhibit different performance on product judgment and choice (Green and Gilhooly, 1992). On the other hand, we might distinguish objective from subjective expertise. Objective expertise corresponds to the actual problem solving capacity of a consumer while subjective expertise corresponds to the perception of one’s ability to perform product-related tasks.

The Predictive Validity of the Components of Knowledge

In order to assess the external validity of the components of consumer knowledge as well as their associations with product-related tasks, we will rely on the Alba and Hutchinson’s (1987) framework of the consequences of expertise (see Table 2). [One has to acknowledge that, in the Alba & Hutchinson paper, familiarity is considered as the antecedent of expertise taped via five types of consequences. Here, we consider and measure four components of consumer knowledge (familiarity, objective product category information, objective expertise, subjective expertise). The five dimensions proposed by A&H are considered as their consequences of consumer knowledge.]

Given that there is less theoretical background to develop specific hypotheses regarding the differential effects of each component on each dependent variable, only generic hypotheses regarding the nature of the associations between consumer knowledge and its consequences will be developed.

FIGURE 1

THE DIMENSIONS AND COMPONENTS OF CONSUMER KNOWLEDGE

TABLE 1

DIFFERENT COMPONENTS OF PRODUCT KNOWLEDGE

Cognitive Effort and Automaticity

Cognitive effort and Automaticity will be analyzed in terms of the ease with which individuals happen to perform wine-related tasks. Prior research has found that familiarity and product knowledge result in more rapid task performance and increases the cognitive resources available for the other tasks (see Alba and Hutchinson, 1987). Thus we make the following hypothesis:

Hypothesis 1. Familiarity and product knowledge (i.e. PCI and expertise) should be negatively related to the cognitive effort required to perform wine-related tasks.

The Cognitive Structure

In this paper, cognitive structure is viewed in terms of the product category structure. Product category structure includes both the level of categorization (basic, subordinate, superordinate levels) and the graded structure of the product category . The level of categorization is defined as the natural level at which individuals categorize wines. Prior studies have shown that individuals low on knowledge tend to categorize concepts at a basic level (Sujan and Dekleva, 1987).

Hypothesis 2a. Familiarity and product knowledge should be negatively related to the categorization at the basic (perceptual) level.

Graded structure corresponds to the idea that, in a particular category (for instance high quality wines), membership (typicality) is a matter of degree and that some members are better examples than others. Graded structure, analyzed in terms of family resemblance (Mervis and Rosch, 1981), increases with familiarity and product knowledge (A&H, 1987).

Hypothesis 2b. Familiarity and product knowledge should be positively related to the variability of family resemblance judgments in a category.

Degree of analysis

Degree of analysis refers to the "extent to which consumers access all and only the information that is relevant and/or important for a particular task" (Alba and Hutchinson, 1987, p. 417). Prior studies have found that familiarity is less related to prepurchase search of information (Bettman and Park, 1980). Moreover, expertise has been found to be positively related to personal (both internal searchBone’s past experiencesBand search by oneself in the store) (Bloch, 1986) but negatively related to interpersonal prepurchase search (e.g. taking the salespeople’s opinions) (Brucks, 1985). Thus, we expect the following:

Hypothesis 3a: Familiarity and product knowledge should be positively related to (a) personal search, (b) the ability to search for information, and (c ) negatively related to interpersonal search.

Another aspect of analysis is the type of attributes that consumers use to judge and select products. Research has suggested that more familiar and knowledgeable individuals rely more on functional, intrinsic and less on perceptual attributes (Dodds et al. 1991). This is consistent with the A&H’s (1987) proposition that experts are more able than novices to restrict acquisition to relevant attributes. Thus, the following hypothesis comes to mind:

Hypothesis 3b. Familiarity and product knowledge should be positively related to relevant choice attributes

Finally, less familiar and knowledgeable individuals are more likely to rely on schema-based inferences. As a result, they should make more stereotyped (erroneous) judgments when evaluating a product (A&H, 1987). Thus, we hypothesize that:

Hypothesis 3c. Familiarity and product knowledge should be negatively related to the propensity to make stereotyped judgments.

Memory

Familiarity and product knowledge increase the ability to perform non-aided, memory-related tasks (A&H, 1987). Thus, we expect the following:

Hypothesis 4: Familiarity and product knowledge should be related positively to the number of attributes an individual can memorize.

2. METHOD

Product class, Population and Sample Selection

The study was conducted in France in November 1996. As in France wines are more distinguished according to their certificate of origin(e.g. AOC, Appellation d’Origine Contr(lTe), region of production (e.g. Bordeaux, Bourgogne) and Estate name (e.g. Mouton de Rothschild), brand is not a meaningful level of analysis. Thus, knowledge was operationalized at the product category level. The sample was composed of 212 students representative of the population of a large Parisian University (not students of the authors). Quota samples were selected by controlling for gender, residence area, year in the program (i.e. from the first to fifth year after high school) and type of studies (e.g. economics, literature, law). Students were interviewed face to face outside their classrooms, in regular conditions

Consumer knowlege measures development

Familiarity

It was operationalized in terms of accumulated product-related experiences (Alba and Hutchinson 1987). As it has been done in most of the studies, we measured only the depth of familiarity as the accumulated experience with a particular type of product. To do so, we defined six (6) varieties of wines: three colors (white, pink, red) combined with two levels of quality (regular, high). We then measured the consumption frequencies of the six corresponding varieties using a 6-point scale (from "never" up to "every day"). We also used a list of 6 main usage situations [These situations were identified in a previous research using the product x usage situation methodology developed by Day and al. (1979).] where people usually consume wine and measured the frequency of wine consumption in each situation, on the same scale. Five measures were therefore computed: (1) the mean consumption frequency of the six varieties of wine, (2) the highest frequency across the six varieties, (3) the mean consumption frequency of drinking wine in the six usage situations, (4) the highest of the six, and (5) the number of years since the consumer has been a wine drinker. The latter was eliminated for low reliability.

Product knowledge

As shown in Table 1, four operationalizations of knowledge were considered: objective and subjective product class information, objective and subjective expertise. Objective product class information was operationalized as the quantity of information stored in memory regardless of its verifiability (Punj and Staelin, 1983; Fiske et al. 1994). Three indicators were developed: (1) the number of brands / names (up to 10), (2) the number of attributes / features (up to 20) and (3)the definition of four basic categories of wines from their respective initials (for instance "do you know what AOC means ?"). For the last indicator, answers were coded true or false and summated to get a composite indicator.

Objective expertise was operationalized as follows. A 20 item multiple choice battery was developed with two wine professional experts and then cross validated with another one. Seven dimensions of expertise were originally identified: (1) geographic aspects of production (areas of production), (2) agricultural activities (varietals and sickness of vineyard, methods of pruning), (3) production of wine (how to produce different varieties, colors of wine), (4) consumption (how to consume different varieties of wine), (5) maturation of wine in a cellar, (6) tasting (evaluation of sensorial characteristics) and (7) choosing. Tasting and choosing were abandoned because of the difficulty to deal with them in a face to face interview and also because they are intrinsically subjective. Twenty items (four items were developed for each of the five remaining dimensions, with four possible answers for each with a unique exact one).

It was important to know if objective product class information and objective expertise are separable measures of product knowledge (it is difficult to capture procedural skills on the basis of a questionnaire which, by nature, is expected to tape declarative knowledge). Thus, for each of the five dimensions of expertise that were retained, we summed the corresponding 4 questions in order to form a specific score of objective expertise (expertise on production techniques, for example).

These five specific objective expertise scores and the three scores of objective product class information were then factor analyzed. The three measures of product class information loaded highly on the first factor, demonstrating their discriminant validity from expertise. They were summed to obtain a global score of objective product class information. The five scores of expertise were also summed to obtain a global score of objective expertise (see Maheswaran et al., 1996; Park et al., 1994).

Subjective product class information and subjective expertise

Four items (5-point Likert scale) of subjective product class information were developed with the aim of capturing the "feeling of knowing" facts about wine. Six items (5-point Likert scale) were developed with the aim of capturing different facets of the consumer subjective expertise: global feeling of expertise, expertise relative to others, expertise regarding choice, consumption, and to the ability to advise other buyers.

TABLE 2

THE STUDIED CONSEQUENCES OF CONSUMER KNOWLEDGE

Our theoretical framework suggests that product class information and expertise might be separate concepts at the subjective level. Confirmatory factor analysis showed that it is impossible to separate them and we had to conclude that the feeling of knowing is similar to the feeling of expertise.

Cognitive consequences of consumer knowledge

Cognitive effort

It was measured in terms of self perceived difficulty to answer objective product category questions (one 5-point Likert item).

Cognitive structure

In France, the color of wine (red, white, pink) serves as the basic level of categorization. To capture this basic level of categorization, respondents were presented a color photo featuring an actual bottle of wine (medium quality) and prompted to describe it by using the three most important characteristics coming to mind (non aided). Color was then coded as follow: 3 if cited first, 2 if cited second, 1 if cited third, 0 if not cited.

As for graded structure, a cardboard with eight actual wine color photos was presented to the respondents who had to provide, for each bottle, a family resemblance judgment (semantic scales). To control for the cognitive category, when showing the photos, interviewers indicated that all were AOC wines (AOC is in France the main legal certification of origin, a synonymous for good wine quality). Then, for each respondent, we computed the variance of its family resemblance judgments across the eight wines. This indicator, was supposed to capture his degree of "graded structure" and to be positively correlated to consumer knowledge.

Analysis

Prepurchase information search was measured with eight 5-point Likert items relative to "the last time you bought wine for a good dinner with friends or family" were developed. Three factors were characterized using exploratory factor analysis: (1) the perceived importance of search by the consumer himself (number of visited stores, effort to find information in magazines, 4 items, Cronbach coefficient=0.69), (2) importance of in-store comparisons (2 items) and (3) interpersonal search (2 items).

Choice attributes. A list of 13 attributes has been constituted during the qualitative phase of the research with experts and consumers. Respondents were shown this list during data collection, by decreasing order of importance, and asked to indicate the criteria (a maximum of 3) they might use to choose a wine in a predefined context ("for a dinner with friends and family"). For each respondent, each attribute was then coded as follow: 3 if cited first, 2 if cited second, 1 if cited third, 0 if not cited. After data collection, the list was examined separately by 4 regular consumers and 2 wine experts who had to decide, a priori, for each attribute, if it is an expert- or a novice- specific criterion, when choosing wine. When experts and consumers disagreed, the corresponding attribute was eliminated (this was also the case when the number of respondents who used this attribute was too small, i.e. less than 30). We finally retained color, price and design of the bottle as a priori "novice" choice attributes (they correspond to extrinsic, non-functional and perceptual attributes), and certificate of origin, region of production, name of the estate, bottled at the estate, vintage, matured in barrel (oak cask) and erceived quality, as a priori "expert" attributes. Global scores of propensity to use "expert" ("novice") attributes were computed by simple summation of the scores of the "expert" ("novice") attributes used by the respondent and hypothesized to be positively (negatively) associated with knowledge.

Inferences and Stereotyping. To measure the tendency to stereotype during product judgments, a six-item battery was developed with experts, with 3 possible answers (objective / false / stereotyped). The total number of stereotyped answers (going from 0 to 6) was then used as a measure of the tendency to stereotype.

Memory

After respondents were presented the color photo of an actual moderate quality bottle of wine, the photo was hidden and respondents were prompted to answer 8 non-aided questions relative to the description of the bottle (color, name of the estate, region of production, vintage, etc.). The responses were then coded as true (1) or false (0) and summated to constitute a global memorization score. [We could not however categorize these recalled attributes in terms of expert versus novice-specific attributes because, unfortunately, the questionnaire did not actually retain the same number of attributes for this task.]

TABLE 3

TEST OF THE MEASUREMENT MODEL WITH CFA

TABLE 4

CFA THEORETICAL CORRELATIONS BETWEEN CONSTRUCTS

3. THE FINDINGS

Structure of the consumer knowledge of wine

A confirmatory factor analysis (CFA) was performed with LISREL8 (J÷reskog and S÷rbom, 1993) to test our four-component consumer knowledge structure (familiarity, objective product category information, objective expertise, subjective expertise). [Both objective product information and objective expertise were included as constructs without measurement error, due to their objective nature.]

Examination of sample-free fit indices (see Marsh, et al. 1996: NNFI; CFI) in Table 3 suggests that the model fits the data quite well and that our constructs can be considered as unidimensional . Table 4 shows the correlation among the theoretical constructs of interest. The first result is that all the components of consumer knowledge are significantly correlated. This is consistent with previous research findings by Brucks (1985), Selnes et Gronhaug (1986), Park et al. (1994). The other interesting result is that subjective expertise is the only construct to be highly correlated with all the others (mean correlation =0.47). Finally we can observe that objective expertise and objective product category knowledge are separate concepts (r=0.28) and their correlation is lower than their respective correlation with subjective expertise.

Links between Consumer Knowledge and its cognitive consequences

Structural models were estimated to examine the links between the four dimensions of knowledge and the cognitive consequences (one model for each cognitive consequence). Based an our previous discussion in section 1 (see also Aurier and Ngobo, 1998), familiarity was included as an antecedent of PCI, objective and subjective expertise; PCI was in turn considered as an antecedent of objective and subjective expertise and finally, objective expertise was considered as an antecedent of subjective expertise. Further, all these components were related to each cognitive consequence. Table 5 presents the estimate results (direct and total effects) of this model for each dependent variable. Research hypotheses (formulated in terms of simple associations) are tested on the basis of total effects because they reflect the effort to reproduce the observed correlation matrix.

As expected, familiarity has positive effects on PCI, objective and subjective expertise; PCI has an effect on both objective and subjective expertise and finally objective expertise has a positive effect on subjective expertise.

Testing the cognitive effort and automaticity hypothesis

Table 5 shows that perceived cognitive effort is negatively and significantly affected by the four omponents of consumer knowledge. The strongest effects are with subjective expertise and product category information.

Testing the cognitive structure hypothesis

H2a cannot be accepted. None of the effects is significant. Knowledge was also hypothesized to positively affect graded structures. This hypothesis is partly accepted. Only familiarity and objective product category information positively affect our indicator of graded structure. Thus, graded structure seems to be essentially a function of cumulated experience and quantity of information stored.

Testing the analysis hypotheses

H3a cannot be rejected. Consumer knowledge positively affects personal search (objective product category information and subjective expertise) and in-store comparisons (objective product category information). It negatively affects personal and interpersonal search (familiarity). These results are consistent with previous studies: (1) negative effects are obtained with familiarity (Bennett and Mendell, 1969, More and Lehmann,1980), (2) positive effects are obtained with measures of product category knowledge (Brucks, 1985). Specifically, experienced consumers have knowledge about attributes and see no utility for further search while knowledgeable consumers have both the motivation and ability to seek information (Bettman and Park 1980). However, given that we did not measure search in terms of intrinsic and extrinsic attributes examined by individuals, we did not tested the U-shaped relationship (Rao and Sieben 1992).

H3b cannot be rejected. The global propensity to use novice attributes is negatively influenced by objective product class information and subjective expertise. Reciprocally, the propensity to use expert-specific attributes is positively affected by objective product class information, objective expertise and subjective expertise. It is interesting to note that the propensity to use a priori expert attributes is, apparently, more influenced by subjective expertise than the other, more objective, components of consumer knowledge.

H3c cannot be rejected. The propensity to stereotype is negatively affected by familiarity. Actual experience seems to be the best way to have an objective perception of wine.

Testing for the memory hypothesis

Table 5 shows that non-aided recall of the characteristics of a bottle is significantly affected by familiarity, objective product category information and subjective expertise. It is interesting to note that subjective expertise more than objective expertise is significantly correlated with an objective task.

TABLE 5

ESTIMATE RESULTS OF THE STRUCTURAL MODELS RELATING CONSUMER KNOWLEDGE COMPONENTS AND COGNITIVE CONSEQUENCES

DISCUSSIONS AND CONCLUSIONS

The structure of Consumer Knowledge. Two major results emerge from this research. First, we identified and validated the existence of four components of consumer knowledge: familiarity, objective product category information, objective and subjective expertise. This is consistent with prior studies which have shown that knowledge is a multidimensional construct (Brucks 1986, Park et al. 1994, Mitchell and Dacin, 1996). However, in relation with the Park et al. (1994) study, we find that objective expertise, subjective expertise and product category information are separate constructs. Second, it was not possible to separate the feeling of knowing (subjective product class information) from the feeling of having skills (subjective expertise).

The consequences of Consumer Knowledge

First, as suggested by A&H (1987; Bettman et al. 1991, Brucks 1986, Fiske et al. 1994), different components f consumer knowledge may have different effects on cognitive tasks; not all the components are correlated with wine related tasks. However, except for personal search, when these components are related, they never exhibit opposites relationships.

Second, as far as each component is concerned, we found that every time objective expertise is correlated with wine related tasks, objective product class information (PCI) is also correlated. However, the reverse is not true: PCI is correlated with more tasks than objective expertise. This implies, for researchers, that a measure of objective expertise seems to be useless when one has also measured PCI, insofar as it is more difficult to operationalize objective expertise (one has to develop product category specific multi item questionnaires). Conversely, objective product category information significantly affects almost all the cognitive consequences of interest in this research.

Similarly, every time subjective expertise is correlated with wine related tasks, objective product class information (PCI) is also correlated. Objective product class information seem to be a good substitute for subjective expertise.

Familiarity seems to play a complementary role in relation with objective product class information. Most of the time, when objective product class information is not correlated with a given task, familiarity is.

Limits

Our results have to be considered with some cautions. First, only one product category was studied. Even though, wine is an appropriate product for the study of consumer knowledge, a cross-category investigation is still necessary before any generalization of the findings. Second, situation, which is an important variable in the study of knowledge (A&H 1987), was only verbally controlled during the interviews. Finally, one can wonder if PCI, as measured in this research, is not a correlate of the efficiency of the consumer working memory.

REFERENCES

Alba Joseph w. and J. Wesley Hutchinson ( 1987), "Dimensions of consumer expertise" Journal of Consumer Research, vol.13 (March), 411- 454;

Aurier, Philippe and Paul Ngobo (1998), A re-examination of the consumer knowledge-information search: it depends on the components, Proceedings of the 27th EMAC Conference, Research and Practice, Editor Peter Andersson, 277-296.

Beatty, S. E. and S. M. Smith (1987), "External search effort: An investigation across several product categories" Journal of Consumer Research, vol. 14, 83-95;

Bennett, P. D. and R.K. Mandell (1969), "Prepurchase information seeking behavior of new car purchasers: the learning hypothesis" Journal of Marketing Research, vol. 6 (November), 430-433;

Bettman, James 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, vol.7 (December ), 234-248;

Bettman, James R., Eric J. Johnson and John W. Payne (1991), "Consumer decision making" In Handbook of Consumer Behavior (eds.) Thomas S. Robertson et Harold H. Kassarijan, Englewood Cliffs, New Jersey: Prentice-Hall, 50-84;

Brucks, Merrie ( 1985), "The effects of product class knowledge on information search behavior" Journal of Consumer Research, vol.12 (June ), 1-16 ;

Cole, Catherine, Gary Gaeth, Goutham Chakraborty, and Irwin Levin (1992), "Exploring the relationship among self-reported knowledge, product usage and consumer decision making" Advances in Consumer Research, 1991;

Day George Shocker Allan and Srivastava Rajendra (1979), "Customer Oriented Approaches to Identifying Products-Markets," Journal of Marketing, vol. 43, January, 8-19

Dodds, William B., Kent Monroe, and David Grewal (1991), "Effects of price, brand store information on buyers’ product evaluation" Journal of Consumer Research, vol. 28, 307-19;

Fiske, Carol A., Luebbehusen L. A., Miyazaki A. D., and Joel E. Urbany (1994), "The relationship between knowledge and search: It depends" Advances in Consumer Research, vol. 21, 43-49;

Green Alison J. K. and K. J. Gilhooly (1992), "Empirical advances in expertise research" In (eds.), M. T. Keane et K. J. Gilhooly, Advances in the Psychology of Thinking, vol. 1, New York: Harvester Wheatsheaf, 45-70;

J÷reskog Karl G. and DSg S÷rbom (1993), LISREL8: User’s reference guide. Mooreville, IN: Scientific Software;

Maheswaran Durairaj, Sternthal Brian and Zeynap Gnrhan (1996), "Acquisition and Impact of Consumer Expertise" Journal of Consumer Psychology, vol. 5 (2), 115-133;

Marsh Herbert W., John R. Balla and Kit-Tai Hau (1996), "Evaluation of incremental Fit Indices: A clarification of mathematical and empirical properties" In Advanced Structural Equation Modeling: Issues and Techniques, (eds.) George A. Marcoulides and Randall E. Schumacker, 315-353;

Mitchell Andrew A. and Peter A. Dacin (1996), "The assessment of alternative measures of consumer expertise" Journal of Consumer Research, vol. 23 (December), 219-239;

Moore, William and Donald R. Lehmann (1980), "Individual differences in search behaviour for a non durable" Journal of Consumer Research, vol.7 (December), 296-308;

Park, C. Whan, D. L. Motherbaugh and L. Feick (1994), "Consumer knowledge assessment+ Journal of Consumer Research, vol. 21 (June), 71- 82;

Punj, N. Girish and Richard Staelin (1983), "A model of consumer information search behaviour for new automobiles" Journal of Consumer Research, vol.9 (March), 366-380 ;

Rao Akshay and W. A. Sieben (1992), "The effect of prior knowledge on price acceptability and the type of information examined" Journal of Consumer Research, vol. 19 (September), 256-270;

Selnes F. and K. Gronhaug (1986), "Subjective and objective measures of product knowledge contrasted" Advances in Consumer Research, vol. 13, 67-71;

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