Consumers' Perceptions of Attributes and Behavioral Intentions: an Extended Comparison-Level Model

J. Bradley Barbeau, The University of Michigan
William J. Qualls, The University of Michigan
ABSTRACT - Researchers of consumer satisfaction have typically focused their efforts on examining determinants and outcomes of the satisfaction/dissatisfaction process. Generally, it has been held that satisfaction results from disconfirmation of expectations, and it is considered to be a critical factor in determining consumers' future purchase behavior or behavioral intentions. The majority of these studies have assumed satisfaction to be an affective response; only a few have attempted to examine consumer satisfaction as a cognitive process. The present study examines two major theoretical paradigms, (1) disconfirmation of expectations and (2) comparison-level theory, and evaluates their contribution to the explanation and prediction of satisfaction. e paper presents an extended comparison-level model of satisfaction based on a cognitive framework. The model vertically integrates consumers' perceptions of product attributes with their behavioral intentions.
[ to cite ]:
J. Bradley Barbeau and William J. Qualls (1984) ,"Consumers' Perceptions of Attributes and Behavioral Intentions: an Extended Comparison-Level Model", in NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT : Association for Consumer Research, Pages: 143-147.

Advances in Consumer Research Volume 11, 1984      Pages 143-147

CONSUMERS' PERCEPTIONS OF ATTRIBUTES AND BEHAVIORAL INTENTIONS: AN EXTENDED COMPARISON-LEVEL MODEL

J. Bradley Barbeau, The University of Michigan

William J. Qualls, The University of Michigan

ABSTRACT -

Researchers of consumer satisfaction have typically focused their efforts on examining determinants and outcomes of the satisfaction/dissatisfaction process. Generally, it has been held that satisfaction results from disconfirmation of expectations, and it is considered to be a critical factor in determining consumers' future purchase behavior or behavioral intentions. The majority of these studies have assumed satisfaction to be an affective response; only a few have attempted to examine consumer satisfaction as a cognitive process. The present study examines two major theoretical paradigms, (1) disconfirmation of expectations and (2) comparison-level theory, and evaluates their contribution to the explanation and prediction of satisfaction. e paper presents an extended comparison-level model of satisfaction based on a cognitive framework. The model vertically integrates consumers' perceptions of product attributes with their behavioral intentions.

INTRODUCTION

Interest in consumer postpurchase behavior has grown tremendously among practitioners and researchers during the last decade. As the body of studies continues to accumulate, so do the number of different conceptualizations of how consumer satisfaction/dissatisfaction (CS/D) affects decision behavior (Anderson 1973; Swan and Combs 1976; Oliver 1979, 1980; Latour and Peat 1979, 1980; Westbrook and Cote 1980). Despite the proliferation of CS/D models, we still understand very little about the process of consumer satisfaction and its ultimate impact upon decision making and purchase behavior. It appears that every researcher has his or her own definition of satisfaction, and that there is little agreement about the meaning of the term.

Overwhelmingly, the focus of CS/D research has been on determinants and outcomes of satisfaction. Investigations into determinants of consumer satisfaction have typically been grounded in two theoretical paradigms: (1) Disconfirmation of expectations (Olshavsky and Miller 1972, Anderson 1973. Oliver 1977, 1980, 1981, Churchill and Surprenant 1982) and (2) comparison-level theories (Latour and Peat 1979, 1980, Swan and Martin 1981). While each of the theoretical approaches has contributed to our understanding of CS/D, they also point to a significant gap in our knowledge. Specifically, researchers have examined the nature of satisfaction as an affective phenomenon, but have not attempted to incorporate the process of satisfaction into cognitive models of consumer decision/purchase behavior. It is our contention that to fully understand the process of CS/D, current models must be extended to include a cognitive base from which affective and cognitive processes can be developed.

This paper briefly examines the disconfirmation-of-expectations and comparison-level paradigms. Critical reviews of these studies are available elsewhere (Latour and Peat 1979, Swan and Martin 1981). The purpose of this paper is to suggest an alternative, more comprehensive conceptualization of the CS/D process, by extending the comparison-level paradigm to include a cognitive framework for examining consumer perceptions of satisfaction and their resultant impact upon consumer decision and purchase behavior.

Disconfirmation of Expectations

Disconfirmation models of satisfaction have been based on the hypothesis that consumers form expectations about how a product will perform and that the consumer's resulting satisfaction/dissatisfaction with the product is directly related to whether the product meets, exceeds, or falls short of those expectations. Disconfirmation of expectations has been shown to be positively related to satisfaction (Oliver 1977, 1980; Trawick and Swan 1980; ChurchilL and Surprenant 1982). Despite the positive relationship, however, the critical variables of satisfaction and disconfirmation of expectations have been conceptuaLized and operationalized in numerous ways, much to the detriment of the CS/D theoretical framework.

Expectations have been defined in the literature as "predictions of affect" regarding, the product (Oliver and Linda 1981), "user anticipation of product performance" (Swan and Martin 1981), and expected degree of satisfaction (Trawick and Swan 1980), creating the impression that it is the research study that determines the definition of expectation. e proliferation of conceptual definitions of expectations has led to a lack of researcher agreement, inability to compare findings, and a weakening of the basic CS/o theoreticaL framework. Whether satisfaction responses to product outcomes are related to the difference between expected and obtained attribute levels or to the difference between expected and obtained levels of affect (satisfaction) associated with the attributes is still another point of contention that has yet to be resolved by disconfirmation proponents.

More recently, Oliver (1980) has conceptualized expectations on the basis of Helson's (1964) adaptation-level theory, as a means of justifying the disconfirmation model. according to Helson (1964), the adaptation level of an individual determines how s/he will perceive a stimulus and is the result of: (1) the qualities of the stimulus (perceived discrepancy), (2) the environment tn which the stimulus is perceived (contextual variables), and (3) the physiological, psychological, and background traits of the perceiver (individual characteristics). Examined in the framework of the disconfirmation paradigm, an individual's adaptation level represents the operationalization of consumer expectations about a product, and the degree of satisfaction is determined by whether the product meets, exceeds, or falls short of the individual's adaptation level. While Oliver's (1980) framework is intuitively appealing, his contention that adaptation occurs at the cognitive level is weakened, in that the way in which expectations are defined remains affectively oriented.

While disconfirmation-of-expectations models formed the early basis of research into and knowLedge of consumer satisfaction, their ability to predict and explain the nature of consumer satisfaction has been minimal. The major criticism may be that these models are based on the relationships between affective responses to outcomes and satisfaction, completely ignoring the nature of the relationship between product attributes and the determinants of the affective responses to those outcomes. Churchill and Surprenant (1989) provide a clue to the problem in their study, which manipulated product "expectations" and performance by varying the information about product attributes. They found that the manipulation of information affected consumer expectations and their disconfirmation, demonstration that a connection exists between attribute levels and levels of affect. The causal nature of the relationship remains to be explained.

Another problem with the disconfirmation model is that it is intuitively tautological in nature. A model based on t e premise that a person will be more satisfied when a product turns out to be better' than expected, and less satisfied when it turns out to be worse than expected, adds little to an ability to predict (1) how levels of affect are arrived at or (2) what absolute levels of satisfaction will occur in a given situation. As conceptualized, the result of the disconfirmation theorem (satisfaction) is assumed in the statement of the theory itself.

Comparison-Level Models

An alternative to the disconfirmation and adaptation-level models of consumer satisfaction is the comparison-level paradigm. As presented by Latour and Peat (1979), satisfaction is based on an evaluation of the difference between obtained attribute levels and a reference set of attribute levels labeled the comparison level (CL). Discrepancies between the obtained level for each attribute (CLi) and the comparison level for each attribute (CLi) are combined additively, and then weighted by the importance of each attribute to the consumer, to determine over-overall satisfaction with the product. a e basis for the comparison level of an attribute for a consumer has been defined by Latour and Peat (1979) as the average levels of past experience with the attribute, levels of the attribute that tile consumer believes to nave been obtained by others, and information about obtainable levels provided b; the manufacturer, salesperson, or other information source. Thus, the CL may be seen as the totality of direct and indirect information about the attribute available to the consumer weighted by tile salience of the information.

While the contribution of each al tribute to the overall satisfaction level is proportional, a key question is whether the (Ai - CLi) discrepancies would be weighted by an absolute level of importance (e.R., 7 on a scale of 1 to 10), or by relative importance (e.g., Ii = .4 and S li = 1). The implication in the case of the absolute scale would be that for products containing attributes with low (high) levels of importance, only low (high) levels of satisfaction!dissatisfaction would be attainable.

In addition, it is uncLear whether the importance of the variable mediates the effect of the evaluation of the discrepancies (Ai - CLi) on satisfaction or affects the individual comparison level. For a "more is better" attribute such as gasoline mileage, high importance might affect the level of the attribute required for the satisfaction response, or might increase the degree of the response associated with a given discrepancy.

While the disconfirmation, adaptation, and comparison-level models all view satisfaction as an evaluative response to the discrepancy between product outcomes and consumer expectations, the nature of the constructs employed in the models is different. The CL model relates satisfaction to cognitive perceptions of attribute levels, while disconfirmation models relate satisfaction to perceptions of affective response to attribute levels. While one advantage of comparison-level models is that they explicitly incorporate different sources of expectations, a disadvantage is that they Ho not explain the determinants of consumers' perceived attribute levels. The results of Latour and Peat's (1980) investigation showed that prior experience and the experiences of others influence individuals' perceptions of attribute levels. Since the CL model depends upon consumers' comparisons of the perceived attribute Levels and their CL, the lack of an explicit theory about the determination of the perceived attribute levels is a serious flaw. In addition, the Latour and Peat model views satisfaction as being completely determined by the degree of the discrepancies between the attribute levels and the comparison level, weighted by the importance of each attribute. This does not allow for the contextual and idiosyncratic factors found by Westbrook (1980) to affect satisfaction.

AN EXTENDED MODEL OF CONSUMER SATISFACTION

Utilizing the strong points of the disconfirmation, adaptation, and comparison-level models of consumer satisfaction, a more comprehensive model may be derived which is able to explain the process of consumer satisfaction/dissatisfaction (CS/D). The model we propose attempts to overcome some of the conceptual difficulties associated with earlier CS/D models and to extend the process to actual behavior or behavioral intentions.

Briefly, the model proposes that the sum total of a consumer's direct and indirect experiences with product attributes plus contextual and individual factors determines the consumer's adaptation level and comparison level for product attributes. It is suggested that these two processes, rather than occurring at the affective level of consumer decision making, are more cognitive in nature. The resulting combination of the discrepancies between adaptation levels (as mediated by perceived attribute levels) and comparison levels, evaluated in light of contextual and individual characteristics, determines the degree of satisfaction (an affective response). At this point the model is extended to include the impact of satisfaction on behavioral intentions, as explained by Thibaut and Kelley's (1959) "Comparison Level for Alternatives." The difference between the perceived level of product attributes and the comparison level for alternatives, as mediated by contextual and individual characteristics, determines the expected satisfaction from alternatives. The resulting interaction between expected satisfaction from alternatives and (obtained) satisfaction determines the behavioral intention (e.g., continue search, switch brands. complain).

Figure 1 illustrates in detail the proposed model. The posited functional form of the model is based on Proposition 1, as follows:

Proposition 1: Satisfaction (SAT) is determined by the discrepancy (D) between perceived attribute level (PL) and the comparison level (CL), and contextual (S) and idiosyncratic (Ci) factors.

The combination of discrepancy and contextual and idiosyncratic effects to for: the affective evaluation (satisfaction) should be modeled well by a weighted additive functional form. The proposed model can be formally expressed as:

EQUATION    (1)

where

D = discrepancy level between perceived attribute level and comparison level,

S = contextual factors,

ci = idiosyncratic factors, and

c1 + c2 + c3 = weighting factors.

FIGURE 1

AN EXTENDED MODEL OF CONSUMER SATISFACTION

Support for an additive functional form can be found in studies by Oliver (1980) and Westbrook (1980). While Oliver (1980) found support for the additive effects of the disconfirmation level, Westbrook (1980) investigated the additive relationships of contextual and idiosyncratic factors. Although Westbrook found the relationships to he approximately linear, nonlinear effects could be modeled easily by utilizing a power weighting factor of the form:

EQUATION    (2)

This model, like that of Latour and Peat (1979), applies to product attributes with positive infinite ideal points.

A key relationship in the model is the method by which the adaptation and comparison levels for attributes are determined. As suggested earlier, it is believed that these processes occur at the cognitive level and are determined by a variety of factors. Proposition 2 states:

Proposition 2: The adaptation and comparison levels for product attributes are determined by both direct and indirect experiences with a specific attribute and contextual and individual factors.

More formally, the posited functional form would be:

EQUATION   (3)  and   (4)

where

ALi = adaptation level for attribute i,

CLi = comparison level for attribute i,

Pi = personally experienced levels of attribute i,

Oi = information about attribute i obtained from others,

Ci = individual factors,

Si = contextual factors,

a1 a2, a3 = weighting factors, and

b1, b2, b3, b4 = weighting factors.

The proposed model above combines Helson's (1964) adaptation-level theory with tile comparison-level theory of Thibaut and Kelley (1959) and Latour and Peat (1979) This perceived attribute level (PL) is based on the consumer's adaptation level.

The final link in the model is that between the qualities of the product and the discrepancy level. The discrepancy level (D) is the sum of the individual's comparison between the perceived level (PL) of each attribute and the CL, weighted by the importance of each attribute to the individual. This is the Latour and Peat (1979) equation:

EQUATION    (5)

where

Ii = importance of attribute i

PLi = perceived level of attribute i.

The perceived level of each attribute will be a function of the level of the attribute which exists in the produce and the individual's adaptation level (ALi). For example, if an individual were to test an unfamiliar sub-subcompact model whose interior was of a similar size to that of other subcompacts experienced by the individual, s/he would judge the interior to be of "normal" size. However, if the individual were to test a full-sized automobile with the same interior space as the subcompact, s/he would probably judge the interior to be "smaller" or "much smaller" than normal, because of the difference in adaptation levels for the two sizes of automobile.

While the previous propositions teal with the antecedents of consumer satisfaction, it is appropriate that a process view of satisfaction/dissatisfaction examine the consequences of this affective response to product attributes. The outcomes of the satisfaction process are the consumer's behavior or behavioral intentions with respect to that product or decision. Too few studies have attempted to investigate what appears to be the key concern among marketing practitioners: namely, how does consumer satisfaction/dissatisfaction affect future behavior? While satisfaction with the present alternative will be one determinant, the availability of other choices which will yield greater satisfaction must be evaluated. This lay be conceptualized as a process in which the consumer fores perceptions of the outcomes that could be expected from purchasing competing products/brands. Based on Thibaut and Kelley's (1959) earlier work on comparison level theory, allowances were conceptualized for the above situation in that a comparison level for alternatives (CLALT) was developed

While the Ct is the standard against which consumers evaluate the attractiveness of the relationship or how satisfactory an alternative is, the CLALT is the standard the member uses in deciding whether to remain in or to leave the relationship- (Thibaut and Kelley 1959, p. 21). With respect to the present model, the CLALT represents the set of outcomes that the consumer believes would be obtained if s/he were to switch to a different alternative. The CLALT ultimately impacts upon the expected level of satisfaction from alternatives. As such, our third proposition states:

Proposition 3: The expected level of satisfaction with alternative products in relation to the level of satisfaction with the present product will affect the consumer's behavioral intentions.

As illustrated in the model, the CLALT is a function of the consumer's past experience with similar produces, information obtained from others, and individual characteristics. The CLALT is compared to the perceived attribute level of the present product, resulting in a measure of the relative advantage of switching to another product alternative. m is discrepancy level for alternatives, in relation to contextual and individual factors, will determine the expected level of satisfaction of the various alternatives. The functional fores of these relationships are:

EQUATION (6) ,  (7) ,  (8)

where

CLALT = comparison level for alternatives for attribute i,

DLALT = discrepancy level for alternatives,

EXP SATALT = expected satisfaction with alternatives, and

d, e = weighting factors.

DISCUSSION

The implications of the comparison-level-for-alternatives model are that an individual may find him/herself in the position of being dissatisfied with a product but, because of the lack of satisfactory alternatives, might still repurchase the product. In this situation, where the comparison level is above both the perceived attribute level for the product and the CLALT, we would expect to observe extensive search behavior for important products and possibly no purchase of any brand for unimportant products. For the case in which the perceived attribute level is above the CL but below the CLALT for at least one alternative, we would expect to observe brand switching, depending upon the degree of importance of the product and the effort required to switch brands.

For example, a consumer considering the purchase of an automobile might find her/himself in one of several situations. If the automobile that the individual presently owns is considered satisfactory or better, and newer models of the same automobile are perceived as having the sane attributes, the individual could be expected to purchase the same automobile unless others are perceived to be better. One would expect the search for alternatives to be limited, with the objective of confirming the expected choice. On the other hand, if the owner considers his/her present model to be unsatisfactory, one would expect to observe extensive search behavior, although brand switching would not be expected to occur unless a model which was perceived to be better was found.

CONCLUSION

In recent gears the literature has been moving toward the development of theoretical frameworks for understanding CS/D. We have reviewed two major theoretical paradigms of CS/D, the disconfirmation-of-expectations paradigm and the comparison-level paradigm. We believe that the comparison-level theory as presented by Latour and Peat (1979, 1980) provides a richer framework within which to develop cognitive models of satisfaction. We have extended their model both backward to include consumers' perceptions of attributes and forward to include behavioral intentions. The resulting model allows for the investigation of a broad range of effects on CS/D, including contextual and individual differences, varying attribute levels, varying choice environments, and manipulation of available information. Future research with the model is needed to test the strength of its hypothesized relationships as well as the relative contributions of marketer-dominated and nonmarketer-dominated variables to consumer satisfaction.

REFERENCES

Anderson, Rolph E. (1973), "Consumer Dissatisfaction: The Effect of Disconfirmed Expectancy on Perceived Product Performance," Journal of Marketing Research 10 (February), 460-469.

Churchill, Gilbert A., and Carol Surprenant (1982), "An Investigation into the Determinants of Customer Satisfaction, Journal of Marketing Research 19 (November). 491-504.

Day, Ralph L. (1980), "How Satisfactory is Research on Consumer Satisfaction?, In Advances in Consumer Research, Vol 7, Jerry C. Olson, ed. Ann Arbor, Mich.: Association for Consumer Research.

Helson, Harry (1964), Adaptation-level Theory, New York: Harper and Row.

Hunt, H. Keith (1977), "CS/D: Bits and Pieces," In Consumer Satisfaction, Dissatisfaction and Complaining Behavior, Ralph L. Day, ed. 38-41.

Kelley, Harold H., and John W. Thibaut (1978), Interpersonal Relations, New York: John Wiley and Sons.

Latour, Stephen A., and Nancy C. Peat (1979), "Conceptual and Methodological Issues in Consumer Satisfaction Research," In Advances in Consumer Research, Vol. 6, William L. Wilkie, ed. Ann Arbor, Mich.: Association for Consumer Research, 431-40.

Latour, Stephen A., and (1980), "The Role of Situationally-Produced Expectations, Others' Experiences, and Prior Experience in Determining Consumer Satisfaction,' In Advances in Consumer Research, Vol. 7, Jerry C. Olson, ed. Ann Arbor, Mich. Association for Consumer Research, 588-592.

Oliver, Richard L. (1977), Effect of Expectation and Disconfirmation on Post-exposure Product Evaluations: An Alternative Approach," Journal of Applied Psychology 62, 480-486.

Oliver, Richard L. (1980), A Cognitive Model of the Antecedents and Consequences of Satisfaction Decision," Journal of Marketing Research 17 (November), 460-469.

Oliver, Richard L. (1981), "Measurement and Evaluation of Satisfaction Processes in Retail Settings," Journal of Re- tailing 57 (3), 25-48.

Oliver, Richard L., and Gerald Linda (1981), "Effect of Satisfaction and Its Antecedents on Consumer Preference and Intention," In Advances in Consumer Research, Vol. 8, Monroe, ed. Ann Arbor, MI: Association for Consumer Research, 88-93.

Olshavsky, Richard U., and John A. Miller (1972), "Consumer Expectations, Product Performance, and Perceived Product Quality," Journal of Marketing Research 9 (February), 19-21.

Olson, Jerry C., and Phillip A. Dover (1976), "A Disconfirmation of Consumer Expectations through Product Trial," Journal of Applied Psychology 64, 179-189.

Swan, John E., and Linda Jones Combs (1976), "Product Performance and Consumer Satisfaction: A New Concept," Journal of Marketing 40 (April), 25-33.

Swan, John E., and Warren S. Martin (1981), "Testing Comparison LeveL and Predictive Expectation Models of Satisfaction, in Advances in Consumer Research, Kent B. Monroe, ed., 77-82.

Swan, John E., and I. Frederick Trawick (1 981), "Disconfirmation of Expectations and Satisfaction with a Retail Service," Journal of Retailing 57 (Fall), 48-67.

Thibaut, John W., and Harold H. Kelley (1959), The Social Psychology of Groups, New York: John Wiley and Sons.

Trawick, I. Frederick, and John E. Swan (1980), "Inferred and Perceived Disconfirmation in Consumer Satisfaction," In Marketing in the 80's, Richard P. Bagozzi et al., eds. Chicago: American Marketing Association, 97-100.

Westbrook, Robert A. (1980), "Intrapersonal Affective Influences on Consumer Satisfaction with Products, Journal of Consumer Research 7 (June), 49-54.

Westbrook, Robert A., and Joseph A. Cote, Jr. (1980), "An Exploratory Study of Non-Product-Related Influences Upon Consumer Satisfaction, In Advances in Consumer Research, Vol. 7, Jerry C. Olson, ed Ann Arbor, Mich.: Association for Consumer Research.

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