Towards a Model of Consumer Post-Choice Response Behavior

ABSTRACT - A process model of consumer post-choice response behavior is developed from four components which include the type of post-choice response, as the dependent variable, and as the independent variables, attribution of the cause of satisfaction/ dissatisfaction, the subjective probability of a successful response, the expected consequences of a response, and the characteristics of individual consumers.


J. D. Forbes, David S. Tse, and Shirley Taylor (1986) ,"Towards a Model of Consumer Post-Choice Response Behavior", in NA - Advances in Consumer Research Volume 13, eds. Richard J. Lutz, Provo, UT : Association for Consumer Research, Pages: 658-661.

Advances in Consumer Research Volume 13, 1986      Pages 658-661


J. D. Forbes, University of British Columbia

David S. Tse, University of British Columbia

Shirley Taylor, University of British Columbia


A process model of consumer post-choice response behavior is developed from four components which include the type of post-choice response, as the dependent variable, and as the independent variables, attribution of the cause of satisfaction/ dissatisfaction, the subjective probability of a successful response, the expected consequences of a response, and the characteristics of individual consumers.


Two distinct approaches to model consumer post-choice processes have been attempted. The first, called comprehensive models in this paper (Hunt 1979, Wilton and Tse 1983), conceptualizes consumer post-choice processes through flow chart diagrams. This approach describes post-choice activities and their feedback loops, but it fails to specify how determinants of post-choice response interact to affect the consumer's choice of one or more post-choice responses.

The second approach, pioneered by Day and Landon (1977) and subsequently modified by Day (1980, 1984), attempts to specify how post-choice determinants interact through mathematical models. These attempts to model post-choice processes mathematically, plus a desire to use consumer satisfaction/dissatisfaction research and related concepts to help in the development of consumer policy, stimulated our project. The model proposed in this paper represents a first step towards this goal.


The behavioral processes included in this model encompass more of the total behavior and experience of the subject's performance of psychological and behavioral activities than previous formulations. We view the process as the psychological and behavioral activities the subject performs on an object within an environment to attain some specific goals. The activities involve costs and effort, generate feedback and take time. In short, behavioral processes consist of a subject, an object, an environment, the activities, the motives and goals underlying the activities, the costs incurred and the benefits derived, the feedback generated and the time elapsed in the attainment of those goals. The major elements of the model developed in this paper is a vector of post-choice responses as the dependent variable, the attribution of the source of satisfaction or dissatisfaction, the subjective probability of a successful response, the expected consequences of a particular-response, and an individual's predisposition to respond. A formulation of the model is as follows:


The relationships specified in the model include the multiplicative nature of the standard expectancy model of subjective probability and expected consequences of a response in the two middle terms. We are not sure of the nature of the functional form of the attribution part of the model. And the individual differences portion is similarly unspecified as discussed near the end of the paper. The subcomponents of the expected consequences component are standard economic theory relationships, with the addition of psychological and social costs and benefits. While not especially wedded to the functional forms, we wanted to include hypothesized relationships where the theory would seem to support them.

The model differs from existing conceptualizations in three ways; it includes responses of both satisfied and dissatisfied consumers, while previous models are generally of complaint behavior by dissatisfied individuals; it is a process model which incorporates feedback from consumers' previous choice and post-choice experiences; and it incorporates consumer's expected psychological, monetary, time and physical costs and benefits as variables in predicting the choice of a variety of possible post-choice responses. Each component, and ways to measure and validate them, are discussed below.


The model's first component, its dependent side, contains a vector of consumers' psychological and behavioral postchoice response alternatives. It includes the responses of both satisfied and dissatisfied consumers ranging from expressing one's thanks for a good meal at a restaurant by thanking the chef or leaving a tip, to taking a defective product back to the store for an exchange, to bringing suit for a million dollars in damages caused by the faulting engineering design of an automobile. These responses may be complex. They may be engaged in simultaneously, or they may be sequential in nature, one being used after a previous response has been ineffective. Three complexities of this component are discussed below.

First, the literature indicates that satisfied and dissatisfied consumers may engage in a wide variety of psychological and behavioral responses subsequent to their consumption experience. The psychological responses may include changing one's overall belief, attitude, product attribute salience, and repurchase intent towards the product. Behavioral responses include private actions, legal actions, and compensatory behavior.

Second, the choice of a specific post-choice response is not discrete. Dissatisfied consumers frequently engage in number of post-choice responses rather than a single activity (Wilton and Tse 1983, Day and Ash 1979). For example, we know that some dissatisfied consumers have and continue to lower their evaluation of the brand, AND to tell heir friends about it, AND complain to the retailer AND to be manufacturer.

Third, Nicosia's (1981) study of how consumers cope with product failure shows that consumers may possess a "hidden agenda" of post-choice responses. Some actions, especially he less extreme activities such as complaining to the ales clerk, are taken first. If the complaints are not favorably received, then other activities are likely to be undertaken. The number of previous studies of post-choice behavior is skewed towards complaint behavior for obvious reasons. However, recent findings in the TARP (1984) study for the Coca-Cola Company suggest that satisfied consumers engage in psychological and behavioral responses towards the company that are different from the responses of dissatisfied consumers. Further, previous studies tend to concentrate on predicting specific post-choice activities while little or no effort has been expended on predicting the choice of specific sets of responses or the possible sequence development of an agenda of response activities.


Measures of post-choice responses are well established. Psychological responses can be measured by changes in belief, product attribute salience, brand attitude, and repurchase intention. Behavioral responses can be assessed through measuring the activities and the responses in which consumers engage.

This model component is probably best studied by longitudinal consumer surveys, consumer panels and process design experiments in which consumer post-choice responses are simulated to occur in predetermined sequence (Tse 1984). In such designs, consumer's beliefs, attitudes, and repurchase intentions can be measured before and after their consumption experience. In addition, the post-experience questionnaires should include intention measures of various post-choice behavioral activities. Follow-up to validate the behavioral activities are also indicated.


Consumers have expectations about their consumption experiences. The satisfaction/dissatisfaction literature hypothesizes that if these expectations are either positively or negatively disconfirmed, consumers may feel confused and stressed (stress is defined as a general state of arousal (Selye 1976)). The emergent stress is likely to motivate consumers to engage in an attribution process. Attribution studies in psychology suggest that consumers may engage in attribution to reduce their dissonance in addition to conventional dissonance reduction strategies. Models of consumer post-choice behavior (Tse 1984) suggest that consumers who experience dissonance may engage in attribution before adopting other dissonance reduction strategies. We believe that further research on this area will be a part of our task. That is, the research process is directed toward explaining the underlying causes for the disconfirming experiences and how consumers cope with them.

Consumers engage in the attribution process for three major reasons. First, they wish to restore control over their confused environment. Knowing the underlying causes of disconfirmation is a beginning step in this direction. Second, knowledge of the underlying causes of dissatisfaction offers important guidance for subsequent activities. Third, disconfirmed consumers may use the attribution result to reduce their stress, such as reducing their responsibility for the dissatisfaction (Krishnan and Valle 1979, and Tse 1984).

Previous studies of attribution in post-choice processes investigated the concepts of locus - who should be responsible (Valle and Wallendorf 1977), stability - the dynamics of the locus (Folkes 1983), level of specificity by chance or if the experience can be generalized (Lawther, Krishnan and Valle 1979), and controllability - whether or not the locus is in control of the outcome (Folkes 1984).

Attribution is a process where consumers act as intuitive scientists who develop and test hypotheses about the underlying causes of their experience from the available information. They actively seek evidence to validate their hypotheses and/or revise their prior hypotheses. Feedback from attribution activities are likely to change the consumers' beliefs, attitudes and repurchase intentions. The attribution result has sometimes appeared to be unstable (Folkes 1983) and seldom does there appear to be a clean dichotomy where the attribution is completely recognized to

be internal to the individual or the fault of manufacturers, sellers or other external factors (Tse 1984). Because of the process nature of attribution, to fully understand measurement results, researchers need to know if consumers are still in a stage of the attribution process or whether they have already decided on how to attribute a situation.


The four different attribution dimensions listed above may be measured by either surveys or experiments. Measurement scales may also be developed. Longitudinal surveys or process design experiments are needed to study the notion of an attribution process, that is, to observe how disconfirmed consumers are simulated to engage in attribution, to measure the attribution activities undertaken and the results of those activities, to identify the factors that terminate the attribution process, and, last and most important to our model, to determine how the attribution result influences the choice of subsequent post-choice responses.


The second determinant in the model is the subjective probability of success the consumer associates with a specific post-choice response. While this construct has not been incorporated formally into models of post-choice behavior, except in Day (1983), there is evidence in previous studies that this construct is important in consumer's decisions. For example, in complaint behavior surveys, dissatisfied consumers explained that they did not engage in specific form of redress because: "it wouldn't do any good", (Warland, Herman and Willits 1975); they "didn't think it would make any difference" (Leigh and Day 1979); or they "didn't think anything they could do would make any difference" (Day and Ash 1979). It is likely that consumers develop these subjective probabilities from information and/or prior experience.


In a recent paper,Day (1984), developed some measures of the consumer's subjective estimates of the possible consequences of specific post-choice responses. His measures, for example, the chances of recovering out of pocket costs of complaining, seem to combine both the subjective probability assessment and the consequences of a particular redress activity. Our model is designed to isolate the consumers' subjective estimates of the probability of the success of a particular post-choice response.

Timing of measuring this component is an important issue. Ideally, the measure needs to be administered before the consumer engages in a post-choice response activity if that activity is not to bias the results. Finding an adequate sample of dissatisfied consumers, given that satisfaction/dissatisfaction studies generally have shown that consumers are satisfied with most purchases, appears to be a non-trivial problem. Being able to find and measure their psychological state before any response has been made appears even more difficult.


The third determinant in the model is designed to test our hypothesis that the choice of post-choice response is a function of the consumers' expected consequences of that particular response. The model adopts an "holistic" approach to expected costs and benefits by including the effects of the current choice and all previous post-choice activities. The model overtly recognizes the possible importance of feedback from previous responses on the present response. We think this component is the model's most important contribution. It differs from previous and more conventional approaches which have used only the effects of the current product/service experience on postchoice decisions.

For example, a person may bring suit for an apparently trivial matter which on its face seems illogical. However, the sales clerk was rude and unhelpful when they tried to return a faulty product to the department store. A hot argument with the store manager concluded with the manager's parting comment casting doubt on the consumer's parentage. A telephone call to the manufacturer's consumer affair department was put on hold for fifteen minutes and then disconnected. And to add insult to injury, the company ignored their letters of complaint. While this and similar scenarios are infrequent, they have happened. The post-choice response to sue in court is likely the result of the totality of the results of the unfortunate experiences and we feel an attempt should be made to overtly include and accumulate such matters in the model.

The expected consequences component of the model is the net result of summing the expected benefits of a response and subtracting both its expected costs as well as the accrued costs of all previous post-choice responses. The problems associated with these concepts are discussed below.


This component has been considered by others. Different costs and benefits include time (Day 1980), money (Richins 1979), and psychological costs and benefits (Bearden and Teel 1980). We attempt to include within this component responses beyond complaining behavior which include things such as physical costs, e.g., the costs consumers may incur trying to repair their broken appliances, the travel costs to return merchandise and similar costs seldom considered in previous studies. The major challenge in this element is not that of conceptualization, since the ideas are the concepts basic to micro economics. The challenge is how to measure and combine these disparate types of costs into a few measures that are mathematically tractable in the model. A further problem is how to accrue such costs and benefits, as perceived by the consumer, over a number of time-periods.


Measures of each type of cost and benefit has been developed to some degree (Richins 1979). As stated above, what needs to be developed is a method to combine them. Nicosia's (1981) coping study pioneered the use of conjoint analysis in this area. This technique merits further study. Economists have been using tradeoff techniques to measures relative utility of two or more goods or services. We are in the process of investigating whether these techniques, similar in some respects to conjoint analysis, may provide some help in measuring this component. Measuring, incorporating and combining values of psychological costs and benefits with more conventional monetary and time costs presents an additional challenge that will have to be addressed.


This model component contains the key difference of the proposed model from existing models of consumer complaint behavior. Accrued costs and benefits are defined as the accumulation of the time, monetary, psychological and physical costs and benefits of consumers from all their previous post-choice responses. The construct is meant to be a "reservoir" or "storage tank". That is, it is a system that captures and records the feedback of all previous post-choice activities and attacks the problem in other models which do not overtly recognize that few consumers entertransactions without previous experience and without existing attitudes towards the exchange transaction. We expect to have to include a forgetting (decay) function in this component as there is much evidence that the salience of both positive and negative feelings about product/service choice and use experience decline over time (Oliver 1981).


We have not found previous attempts to measure consumer feedback systems of post-choice response experiences. In this relative void of empirical measurement we presently plan to proceed as follows.

Measures designed to understand behavioral processes must be able to capture consumers' activities and psychological states at different time slices. Having obtained these time series data, we may be able to make inferences about the process(es) through which consumers have gone. Previous studies and discussions on how consumers may change after product experiences suggest studying the following variables in feedback mechanisms. They include attitudes (Nicosia 1966), beliefs (Engel, Kollat, and Blackwell 1978) confidence (Howard and Sheth 1969), salience (Tybout and Yalch 1980), and choice heuristics (Bettman 1979). The above constructs are likely not to be exhaustive. Nonetheless, the idea that there exists a feedback system within consumers which enable them to remember and be changed by their experiences is intuitively realistic. Studying these feedback processes is crucial to understanding and modelling post-choice response behavior.


This component of the model will either become an error term or will be the pool of data from which one or more characteristics specific to individuals or to groups of consumers will be found that can help in explaining post choice response. Given the frequent failure of both psychological and consumer behavior research to identify individual differences capable of explaining a significant proportion of variation among individuals, we do not expect much from this variable and expect it will be mainly the error term. However, we could be surprised.


A process model to explain and predict consumer choice of post-choice behavior in response to satisfied and dissatisfied consumption experiences was outlined. The authors are presently engaged in finding existing data bases to test some portions of the model. A series of research designs are in the planning process to test other aspects of the model.


Ash, S., (1978) "A Comprehensive Study of Consumer Satisfaction with Durable Products", in K. Hunt, ed., Advances in Consumer Research, Vol. V, p. 254-262.

Bearden, W. and Teel, J. E., (1980) "An Investigation of Personal Influences on Consumer Complaining". Journal of Retailing, Vol. 56, No. 3, p. 320.

Bettman, J. R., (1979) An Information Processing Theory of Consumer Choice, Reading, MA: Addison-Wesley.

Day, R., (1980) "Extending the Concept of Consumer Satisfaction", in J. Olson, ed., Advances in Consumer Research, Vol. VII, p. 149-154.

Day, R., (1984) 'Modeling Choices Among Alternative Responses to Dissatisfaction", in T. C. Kinnear ed., Advances in Consumer Research, Vol. XI, p. 496-499.

Day, R. and Ash, S., (1979) "Consumer Response to Dissatisfaction with Durable Products", in W. Wilkie, ed., Advances in Consumer Research, Vol. VI, p. 438-444.

Day, R., and Landon, E. (1978) "Towards a Theory of Consumer Complaining Behavior", in A. Woodside, J. Sheth, and P. Bennett, eds., Consumer and Organization Buying Behavior, New York: North Holland, p. 425-437.

Day, R., Schaetzle, T., Grabicke, K., and Staubach, F., (1981) "The Hidden Agenda of Consumer Complaining", Journal of Retailing, Vol. 57, No. 3, p. 86-106.

Engel, J., Kollat, D. and Blackwell, R., (1978) Consumer Behavior, 2nd ed. New York: Holt, Rinehart and Winston.

Folkes, V., (1984) "An Attributional Approach to Postpurchase Conflict Between Buyers and Sellers", in T. C. Kinnear, et., Advances in Consumer Research, Vol. XI, p. 500-503.

Howard, J. A. and Sheth, J. N. (1969) The Theory of Buyer Behavior, New York: Wiley.

Hunt, J., (1979) "Consumer Response to Inequitable Economic Exchange: An Attribution Theoretical Analysis", in K. Hunt and R. Day, eds., Redefining Concepts and Measures of Consumer Satisfaction and Complaining Behavior, Bloomington: University of Indiana, Division of Research, p. 23-28.

Lawther, K., Krishna, S. and Valle, V., (1979) "The Consumer Complaint Process: Directions for Theoretical Development", in R. Day and K. Hunt eds. New Dimensions of Consumer Satisfaction and Complaining Behavior, Bloomington: University of Indiana. Division of Research. p. 10-15.

Leigh, T. and Day, R., (1979) "Satisfaction, Dissatisfaction and Complaint Behavior", in R. Day ed., Consumer Satisfaction, Dissatisfaction and Complaining Behavior, Bloomington: Indiana University, Division of Research, p. 91-94.

Nicosia, F., (1966) Consumer Decision Processes, New York: Prentice-Hall Inc.

Nicosia, F., (1981) Final Report on Consumer Technology and Public Policy, Parts I and II, Berkeley: University of California

Richins, M., (1979) "Consumer Complaining Process: A Comprehensive Model", in R. Day and K. Hunt eds., New Dimensions of Consumer Satisfaction and Complaining Behavior, Bloomington: School of Business, Indiana University, p.30-34.

Selye, H., (1976) Stress in Health and Disease, Reading: Butterworths.

TARP (1984) The Bottom Line Benefits to Consumer Education, Washington, DC: Technical Assistance Research Programs.

Tse, D., (1984) "A Model of Consumer Post Choice Processes", Unpublished Doctoral Dissertation, Berkeley: University of California.

Tybout, A. and Yalch, R., (1980) "The Effect of Experience: A Matter of Salience?", Journal of Consumer Research, Vol. 3, p. 406-413.

Valle, V. and Wallendorf, M., (1977) "Consumer Attributions of the Cause of Their Product Satisfaction and Dissatisfaction:, in R. Day ed., Consumer Satisfaction, Dissatisfaction and Complaining Behavior, Bloomington: Indiana University.

Wilton, P. and Tse, D., (1983) "A Model of Consumer Response to Communication and Product Experiences, A. Woodside and L. Percy, eds., Advertising and Consumer Psychology, Lexington, MA: Lexington Books, p. 315-332.

Warland, R. H., Hermann, R. O., and Willits, J., (1975) "Dissatisfied Consumers: Who Gets Upset and Who Takes Action," Journal of Consumer Affairs, Vol. 9, p. 148-163.



J. D. Forbes, University of British Columbia
David S. Tse, University of British Columbia
Shirley Taylor, University of British Columbia


NA - Advances in Consumer Research Volume 13 | 1986

Share Proceeding

Featured papers

See More


Meaningfulness in New Products: Conceptualization and Measurement

Maria Sääksjärvi, Delft University of Technology, The Netherlands
Katarina Hellén, Univeristy of Vaasa

Read More


Vicarious Pride: When Gift Customization Increases Recipients’ Appreciation of the Gift

Marta Pizzetti, Università della Svizzera Italiana
Michael Gibbert, Università della Svizzera Italiana

Read More


“I Desire A Brand When I See How They are Different from Me”: Differential Effects of Blatant and Subtle Brand Rejection

Jun Yan, University of Manitoba, Canada
Fang Wan, University of Manitoba, Canada
Americus Reed, University of Pennsylvania, USA

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.