Consumer Knowledge Assessment: How Product Experience and Knowledge of Brands, Attributes, and Features Affects What We Think We Know

C. Whan Park, University of Pittsburgh
Lawrence Feick, University of Pittsburgh
David L. Mothersbaugh, University of Pittsburgh
ABSTRACT - Results of a study conducted to understand the knowledge assessment process indicate that information ranging from specific knowledge about the product (e.g., brand names, attributes, and features), to memory for experience events (e.g., ownership, usage, and search), to statements about involvement with the product category are used in making knowledge assessment judgements. In addition, experience-related information (ownership, usage, search) was found to dominate consumers= responses in the knowledge assessment process. Finally, tentative results suggest that attribute and brand information may not be used as a cue to infer actual knowledge even when this information is available in memory. Implications of these results for future research in knowledge assessment are discussed.
[ to cite ]:
C. Whan Park, Lawrence Feick, and David L. Mothersbaugh (1992) ,"Consumer Knowledge Assessment: How Product Experience and Knowledge of Brands, Attributes, and Features Affects What We Think We Know", in NA - Advances in Consumer Research Volume 19, eds. John F. Sherry, Jr. and Brian Sternthal, Provo, UT : Association for Consumer Research, Pages: 193-198.

Advances in Consumer Research Volume 19, 1992      Pages 193-198

CONSUMER KNOWLEDGE ASSESSMENT: HOW PRODUCT EXPERIENCE AND KNOWLEDGE OF BRANDS, ATTRIBUTES, AND FEATURES AFFECTS WHAT WE THINK WE KNOW

C. Whan Park, University of Pittsburgh

Lawrence Feick, University of Pittsburgh

David L. Mothersbaugh, University of Pittsburgh

ABSTRACT -

Results of a study conducted to understand the knowledge assessment process indicate that information ranging from specific knowledge about the product (e.g., brand names, attributes, and features), to memory for experience events (e.g., ownership, usage, and search), to statements about involvement with the product category are used in making knowledge assessment judgements. In addition, experience-related information (ownership, usage, search) was found to dominate consumers= responses in the knowledge assessment process. Finally, tentative results suggest that attribute and brand information may not be used as a cue to infer actual knowledge even when this information is available in memory. Implications of these results for future research in knowledge assessment are discussed.

INTRODUCTION AND BACKGROUND

In consumer behavior, researchers have examined the effects of knowledge on behaviors such as the extent of information search (Brucks 1985; Moore and Lehmann 1980; Punj and Staelin 1983), the efficiency of information search (Brucks 1985; Punj and Staelin 1983), the order of information acquisition (Siminson, Huber, and Payne 1988), the learning of new product information (Johnson and Russo 1984), choice of evaluation strategies (Sujan 1985), and decision processes (Bettman and Park 1980).

However, across studies, there has been little consistency in the definition or operationalization of consumer knowledge. Specifically, while some research has defined and operationalized knowledge as the nature and amount of information about a specific domain that is stored in long-term memory (i.e., actual knowledge (AK)), other research has defined and operationalized knowledge as consumers= perceptions of what or how much they know (i.e., self-assessed knowledge (SAK)).

Despite the apparent multidimensional nature of the consumer knowledge construct (e.g., AK vs. SAK), it is often treated as a unidimensional construct, with results obtained using AK or SAK measures of knowledge interpreted as representing a "knowledge effect.? However, research indicates that (1) what people think they know and what they actually know often differ (DeNisi and Shaw 1977; Lichtenstein and Fischoff 1977; , Fischoff, Slovic, and Lichtenstein 1977; Schacter 1983; Nelson, Leonesio, Shimamura, and Landwehr 1982; Park 1982), and (2) actual knowledge and self-assessed knowledge may have different effects on consumer search and decision processes (Bransford 1979; Park, Gardner, and Thukral 1988; Rudell 1979).

Actual vs. Self-Assessed Knowledge

The research by DeNisi and Shaw (1977) suggests that at "moderate? levels of actual knowledge, what people know and what they think they know do not correspond well. Research by Lichtenstein and Fischoff (1977), and Fischoff, Slovic, and Lichtenstein (1977) had subjects provide answers to dichotomous questions and then judge how certain they were that the answer they gave was the correct answer (i.e., they provided a confidence rating). Results suggest that even when subjects were very confident in their responses (e.g., a 100 percent probability that the chosen answer was the correct one), these responses were only objectively correct 20 to 30 percent of the time (Fischoff, Slovic, and Lichtenstein 1977).

Further indications that what people know and what people think they know may differ comes from Schacter (1983), Nelson et al. (1982), and others studying the "feeling-of-knowing? phenomenon. Schacter reasoned that people who lacked the information needed to recall an item may still know enough to indicate whether or not they would be able to recognize it if they saw it. This mode of expressing knowledge about unrecalled information is termed the "feeling-of knowing? effect. Although such feelings appear to predict recognition of formerly unrecallable items at an above chance level, the effect is modest. Specifically, subjects fail to recognize many items which they feel they know (Schacter 1983). Nelson, et al. studied how extent of learning affects the accuracy of feeling of knowing judgements (Extent of learning was operationalized as the number of times an item was recalled after an initial exposure). They found that when learning was low, the correlation between feeling-of-knowing (analogous to SAK) and recognition (analogous to AK) was zero. Even when the extent of learning was high, the correlation was modest (r=+.31).

Finally, there is also research in consumer behavior that supports the potential discrepancy between AK and SAK. For example, Park (1982) found that consumers think they know more about their spouses= home preferences than they actually do. Specifically, their perceived knowledge of their spouses= home preferences did not accurately reflect their spouses= actual home preferences in terms of salient features, and preferred levels on these features.

The Differential Effects of AK and SAK

Because AK is an objective measure of factual memory content, and SAK is a perceptual measure of memory content, it seems reasonable to view AK as an ability factor, and SAK as a motivational factor. Specifically, AK may provide raw material for problem solving, and lead to increased efficiency in search, and, hence, better accuracy in problem solving. Alternatively, SAK may provide the motivation (or lack of motivation) to search for and process task-relevant information. Research support for SAK as a motivational factor is provided by Bransford (1979), Park, Gardner, and Thukral (1988), and Rudell (1979).

Bransford (1979) has suggested that metacognition (or awareness of a deficiency in knowledge needed to complete a task) is one of the most important means to enhance the motivation to learn. Given the close conceptual connection between SAK and metacognition (both require self-assessments of knowledge), it is likely that SAK consists of a motivational component as well. In addition, Park, Gardner and Thukral (1988) found that independent of AK, SAK affected inference generation, and receptivity of subjects towards updating prior information and beliefs based on new information. Specifically, individuals with low SAK generated more inferences and were more receptive to updating prior information based on the new information than those with high SAK. Rudell (1979) found that while higher levels of AK increased subjects ability to use new information, higher levels of SAK increased subjects reliance on internal information or memory (i.e., it appeared to decrease their motivation for external search).

An additional finding on the differential effects of AK and SAK is provided by Brucks (1985). Brucks found that SAK can have an effect on search strategy that is unrelated to AK. Specifically, Brucks found that SAK was negatively related to use of expert advice (dealer evaluations), while AK was not related to the use of these evaluations.

These results have several implications for consumer behavior research. First, given that SAK acts as a motivational factor, it may be important in determining whether or not search is initiated, and the types of information that receive attention and subsequently get processed by a consumer. Second, since SAK is often an inaccurate indicator of AK, this inaccuracy may lead to biased or suboptimal consumer behaviors.

Given the potential importance of SAK in consumer behavior, and the potential for SAK to be a biased indicator of AK, it is important to understand how SAK judgements are made. Understanding this assessment process, and information that is used by consumers in making SAK judgements, will yield a better understanding of the SAK construct This, in turn, should help researchers understand (1) when and why discrepancies will exist between AK and SAK, and (2) when and why SAK will have effects on behaviors and decision processes that are different from the effects of actual knowledge.

The purpose of this study is to gain a better understanding of the knowledge assessment process and the information that is used by consumers in making these knowledge assessments.

KNOWLEDGE ASSESSMENT

Consumer knowledge assessment can be seen as a judgement task in which contents of internal memory are scanned for information relevant to the judgement. Specifically, assessing one=s own knowledge involves self-perception (Bem 1972) based on internal information, or memory cues.

A wide range of potential memory cues could be used in making an SAK judgement. For example, one set of cues that could be used is specific product-related information (or awareness of the lack of such knowledge) such as brand names that are known, attributes, features, and differences among brands on a given feature or attribute. Specifically, when asked how much they know about a given product, consumers might use the fact that they know many brand names or attributes of the product, and therefore infer that they know quite a bit about the product.

Alternatively, SAK judgements could be made based on memory for events or experiences (or lack of experience) related to that product. For example, memory of past experience with a product in the form of ownership, usage, or information search may be used to make SAK judgements. Specifically, when asked how much they know about a given product, consumers might utilize the fact that they own or use the product or that they have searched for information about the product, and therefore infer that they know quite a bit about the product. In addition, they might distinguish between different sources of information search such as media vs. personal sources. In fact, given that event memory is less likely to decay over time (Carlston 1980), memory of product-related experiences may have an impact on SAK judgements that is largely independent of the amount of product-related information that is available.

Other memory cues that could be used include (1) relative amount of knowledge compared to others (e.g., I know a lot compared to my friends), (2) relative amount of knowledge in one domain as compared to knowledge in another domain (e.g., Compared to my knowledge of Product X, I know very little about Product Y),and (3) level of interest/involvement in the product and product class (e.g., I am very interested in Product X).

METHOD

Overview

In order to understand what information consumers use when making knowledge assessments, we conducted an exploratory study. The product used in this study was CD players, since a pretest indicated that there was sufficient across-individual variation in SAK on CD players to allow for a meaningful study of the knowledge assessment process.

First, participants were asked to indicate how much they felt they knew about compact disc (CD) players. Next we collected cognitive response data by asking participants to write down why they felt they knew as much about CD players as they indicated on the first question. Later in the questionnaire, respondents were asked to recall as many (1) brands, and (2) attributes and features of CD players as they could.

Two questions were of interest in this study. First we were interested in identifying the types and frequency of usage of information or internal memory cues used in making knowledge assessments. Second we were interested in understanding the effect of the use of various memory cues on knowledge assessments.

Sample

Self-administered questionnaires were completed by a convenience sample of 93 MBA students at The University of Pittsburgh who were paid to participate. The average age in this sample was 26 years, 54 percent of the respondents were male, and 39 percent of the respondents owned a CD player at the time of questioning.

Measurement

Self-Assessed Knowledge - Self-assessed knowledge (SAK) was measured using a 9-point single item question. This question asked "How much do you feel you know about CD players?" Response categories ranged from (1) "very little," to (9) "very much." This SAK measure is a very general type of SAK question. It includes (1) no mention of a comparison standard (e.g., How much do you know about CD players as compared to the average person), and (2) no mention of the dimension on which knowledge assessment should be based (e.g., How much do you know about the features of CD players?, or How much do you know about the brands of CD players?) Although other research on SAK has utilized both comparison standards and/or dimension statements (Brucks 1985; Spreng and Olshavsky 1990), we wanted to impose as little outside structure on the task as possible, and therefore did not utilize these types of structuring statements.

Cognitive Responses - Immediately after responding to the general SAK question, respondents turned the page and were asked "On the lines below, please list the reasons why you felt you knew as much as you indicated in question 1." Coding of the responses was done by two coders, one of whom was not involved in the study design. Interjudge agreement was 85 percent, and disagreements were resolved by discussion. Next, general categories were developed for the analysis presented in this paper.

Table 1 describes these general categories which include: (1) knowledge-based responses such as (a) brand name knowledge (Brands), (b) knowledge of the attributes and features of CD players (Attributes), and (c) knowledge of technology (Technology), (2) experience-based responses such as (a) ownership of a CD player (Ownership), (b) usage of CD players (Usage), and (c) search for CD player information (Search), and (3) responses indicating involvement with CD players (Involvement).

For each of the seven response categories, each respondent was categorized into one of three groups as follows: (1) Positive mention - respondents with one or more positive responses in that category (e.g., I own a CD player), (2) Negative mention - respondents with one or more negative responses in that category (e.g., I don't own a CD player), and (3) No mention - respondents who made no statements that fell within that category. Because of the coding scheme, it was possible for a respondent to be categorized into both the positive and negative mention groups. In the few cases when this occurred, the respondent was not included in that particular analysis.

As an aside, although previously we suggested that comparisons with other people and comparisons with knowledge in other categories might be used as cues to infer SAK, no statements of this nature were made by the respondents.

Knowledge of CD Brand Names - Knowledge of brands of CD players was measured using a free-recall question that asked respondents to list as many brands of CD player as they could. The total number of brand names listed was used as a measure of the number of brands of CD players available in memory.

Knowledge of CD Attributes - Knowledge of CD attributes and features was measured using a free-recall question that asked respondents to list as many attributes and features of CD players as they could. Examples of attributes and features from other product domains were given to ensure that respondents understood the meaning of these terms. Again, the total number of attributes and features listed was used as a measure of the number of attributes and features of CD players available in memory.

ANALYSIS AND RESULTS

To answer the general questions of interest in this study, we conducted two sets of analyses that utilized the SAK measure and the cognitive response categories shown in Table 1.

The Relative Importance of Different Internal Memory Cues

In the first analysis, we examined the relative importance of each cognitive response category in making knowledge assessments. This analysis was completed by computing the percentage of each participant's total responses that fell into each of the seven categories, and then averaging these across all 93 respondents. These results are shown in Table 1. The results indicate the dominance of experience-based cognitive responses. Almost seventy percent of the cognitive responses were linked to product experience. On the other hand, less than thirty percent of the responses related to knowledge about CD players. The single most frequently mentioned category was ownership (28 percent) and the least frequently mentioned was knowledge of brands (2 percent).

The Effect of Internal Memory Cues on SAK Judgements

In our discussion of knowledge assessment, we proposed that knowledge assessment is a judgment process based on internal memory cues. Therefore, the cues used by a respondent in making a knowledge assessment should affect the level of SAK indicated by the respondent. Specifically, those in the positive mention group (POSM) of a given category should be expected to have higher levels of SAK than those who make no mention (NOM), while those in the negative mention group (NEGM) should have lower levels of SAK than those in the no mention group.

TABLE 1

DESCRIPTION OF THE COGNITIVE RESPONSE CATEGORIES

In order to determine which internal cues had an effect on SAK judgements, we performed a one-way ANOVA for each cue category, using the general SAK measure as the response variable, and group membership (e.g., negative mention, positive mention, and no mention) as the explanatory variable. In addition, Tukey's pairwise comparisons were performed on the groups to determine which groups differed in terms of average SAK. These results are presented in Table 2.

Several results from Table 2 are worth noting. First, with one exception, the means within each cognitive response category exhibit the pattern NEGM < NOM < POSM, i.e., those in the positive mention group had higher SAK scores than those in the no mention group, and those in the no mention group had higher SAK scores than those in the negative mention group. Even in the case of the exception (usage), the positive mention group had higher levels of SAK than the negative. Differences were significant in all response categories except the two in which there were very small n=s in one cell (brands and attributes). For all other response categories, SAK was higher in the positive then in the negative mention group.

Table 2 also includes the percentage of respondents who mentioned each of the response categories. The results indicate that more respondents gave experience-based than knowledge-based responses. For example, while brands and attributes were mentioned by 8 and 33 percent of respondents, respectively, ownership, usage, and search were mentioned by 73, 50, and 44 percent of respondents. These results led us to wonder if lack of positive mention of brands and attributes was due to a lack of knowledge of brands and attributes. To check for this possibility, we performed further analyses using the free-recall measures of brand-name and attribute knowledge. In these analyses, we examined the number of brands and attributes listed (free-recall) by whether or not they mentioned brands or attributes as a reason for their SAK.

The number of brands listed by respondents ranged from 0 to 16, with an average of about 6. The number of attributes listed ranged from 0 to 20, with an average of 5.5. Because of the small number of respondents who provided positive mention of brands, the results for brands are likely to be unreliable and therefore no statistical tests were performed. Still, in the no mention group, the average number of brands listed was about 6, while the positive mention group listed 7.

TABLE 2

SELF-ASSESSED KNOWLEDGE SCORES BY RESPONSE CATEGORY

For attributes, the mean number of attributes listed in the no mention group (NOM) was 5.2, while the mean in the positive mention group (POSM) was 6.5, and the difference is not significant (t = 1.5, p=.136). For both brands and attributes, these results suggest that those in the no mention group were as able to list brands and attributes as those in the positive mention group. The results suggest that brand and attribute information was available in memory to these respondents but not used in making knowledge assessments.

DISCUSSION

This study was conducted to better understand the knowledge assessment process, and the kinds of information or internal memory cues used by consumers in making knowledge assessments. Preliminary analyses indicate several interesting results. First, the types of internal memory cues used by consumers in making knowledge assessments range from specific knowledge about the product (e.g., attributes and features), to memory for experience events (e.g., ownership, usage, and search), to statements about involvement with the product category.

Second, experience-based cues were found to dominate consumers= responses in the knowledge assessment process. On average, 68 percent of a participants responses fell into the experience-based category. This is contrasted with only 28 percent in the knowledge-based category.

Third, in addition to being mentioned more frequently, experience-based responses also seemed to have a greater effect on the level of SAK reported by respondents. There were significant differences in SAK across the mention groups for ownership, usage, and search: all of the experience-based responses. On the other hand, of the knowledge-based responses, only technology responses resulted in significant differences in the level of SAK reported by respondents. One implication of these results is that to the extent that there is not a one-to-one correspondence between experience and actual knowledge (Brucks 1985), the use of experience as a heuristic to infer actual knowledge will lead to biased knowledge assessments.

Fourth, although the results are tentative because of small sample problems, the results indicate that the presence of a cue in memory does not necessarily lead to its use in an SAK judgement. Those who did not mention attributes or brands as cues were found to be about as capable of listing attribute and brand information as those who did mention brands or attributes. It appears that memory for experiences is relatively more accessible than memory for specific product knowledge, and therefore is more likely to be used in making knowledge assessments.

Several areas of future research in consumer knowledge need further exploration. First, further research is needed to better understand the interrelationships between experience, actual knowledge, and self-assessed knowledge. Although product experience is likely to result in actual knowledge (Alba and Hutchinson 1987; Brucks 1985), the relative effects of experience and actual knowledge on self-assessed knowledge are less clear.

A second and related issue is the role of actual knowledge and experience in causing over and underconfidence in SAK judgements. Specifically, while much research has examined factors leading to overconfidence, little research has examined factors such as lack of product-related experience that might result in underconfidence in knowledge assessments. Finally, more research is needed to better understand the different effects that actual and self-assessed knowledge may have on consumer search, processing, and decision processes.

REFERENCES

Alba, Joseph W. and J. Wesley Hutchinson (1987), "Dimensions of Consumer Expertise," Journal of Consumer Research, 13 (March), 411-454.

Bem, Daryl J. (1972), "Self-Perception Theory," in L. Berkowitz (ed.), Advances in Experimental Social Psychology, 14, 23-31.

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.

Bransford, John D. (1979), Human Cognition: Learning, Understanding and Remembering, Belmont, CA: Wadsworth Publishing Co.

Brucks, Merrie (1985), "The Effects of Product Class Knowledge on Information Search Behavior," Journal of Consumer Research, 12 (June), 1-16.

Carlston, Donal E. (1980), "Events, Inferences, and Impression Formation," in R. Hastie, T. Ostrom, E. Ebbesen, R. Wyer, D. Hamilton, and D. Carlston (eds.), Person Memory: The Cognitive Basis of Social Perception, Hillsdale, NJ: Lawrence Erlbaum Associates.

DeNisi, Angelo S. and James B. Shaw (1977), "Investigation of the Uses of Self-reports of Abilities," Journal of Applied Psychology, 62 (October), 641-644.

Fischoff, Baruch, Paul Slovic, and Sarah Lichtenstein (1977), "Knowing with Certainty: The Appropriateness of Extreme Confidence," Journal of Experimental Psychology: Human Perception and Performance, 3 (November), 552-564.

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

Lichtenstein, Sarah and Baruch Fischoff (1977), "Do Those Who Know More also Know More About how Much They Know? The Calibration of Probability Judgements," Organizational Behavior and Human Performance, 20 (December), 159-183.

Moore, William L. and Donald R. Lehmann (1980), "Individual Differences in Search Behavior For a Nondurable," Journal of Consumer Research, 7 (December), 296-307.

Nelson, Thomas O., R. Jacob Leonesio, Arthur P. Shimamura, and Robert F. Landwehr (1982), "Overlearning and the Feeling of Knowing," Journal of Experimental Psychology: Learning, Memory, and Cognition, 8 (July), 279-288.

Park, C. Whan (1982), "Joint Decisions in Home Purchasing: A Muddling-Through Process," Journal of Consumer Research, 9 (September), 151-162.

Park, C. Whan, Meryl P. Gardner, and Vinod K. Thukral (1988), "Self-Perceived Knowledge: Some Effects on Information Processing for a Choice Task," American Journal of Psychology, 101 (Fall), 401-424.

Punj, Girish N. and Richard Staelin (1983), "A Model of Consumer Information Search Behavior for New Automobiles," 9 (March), 366-380.

Rudell, Fredrica (1979), Consumer Food Selection and Nutrition Information, New York: Praeger.

Schacter, Daniel. L. (1983), "Feeling of Knowing in Episodic Memory," Journal of Experimental Psychology: Learning, Memory, and Cognition, 9 (January), 39-54.

Siminson, Itimar, Joel Huber, and John Payne (1988), "The Relationship Between Prior Brand Knowledge and Information Acquisition Order," Journal of Consumer Research, 14 (March), 566-578.

Spreng, Richard A. and Richard W. Olshavsky (1990), "The Impact of Standard of Comparison and Knowledge Domain on the Measurement of Subjective Knowledge," in A. Parasuraman and William Bearden (eds.), 1990 AMA Educators' Proceedings, American Marketing Association, 44-48.

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

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