To Choose Or Not to Choose: This Is the Question

Ravi Dhar, University of California, Berkeley
ABSTRACT - The question of how people choose from a set of alternatives has been extensively examined at the expense of addressing another interesting and related issue: whether or not to choose from a given set of alternatives. The theme underlying normative theory of consumer choice emphasizes that consumers maximize outcome utility. The assumption is that people arrive at an overall value (utility) by combining a vector of attributes or outcome sequences for each option. For example, when selecting from multiattribute alternatives, the vector of product attributes is reduced to a scalar called utility that is maximized. Similarly, consumers selecting between extended sequences choose the plan that maximizes the discounted value of the component outcomes. The findings presented here argue against the adequacy of the classical utilitarian view that decisions are evaluated solely on the basis of their ends rather than their means. We test this general premise for two consumer decision making scenarios: preference for not choosing and choice between extended sequences. This session shows that the process of deciding influences choices in a manner such that normatively equivalent outcomes result in different preference orders. This paper summarizes the session by providing a brief overview, and by describing representative ideas from the four papers comprising the session.
[ to cite ]:
Ravi Dhar (1992) ,"To Choose Or Not to Choose: This Is the Question", in NA - Advances in Consumer Research Volume 19, eds. John F. Sherry, Jr. and Brian Sternthal, Provo, UT : Association for Consumer Research, Pages: 735-738.

Advances in Consumer Research Volume 19, 1992      Pages 735-738

TO CHOOSE OR NOT TO CHOOSE: THIS IS THE QUESTION

Ravi Dhar, University of California, Berkeley

ABSTRACT -

The question of how people choose from a set of alternatives has been extensively examined at the expense of addressing another interesting and related issue: whether or not to choose from a given set of alternatives. The theme underlying normative theory of consumer choice emphasizes that consumers maximize outcome utility. The assumption is that people arrive at an overall value (utility) by combining a vector of attributes or outcome sequences for each option. For example, when selecting from multiattribute alternatives, the vector of product attributes is reduced to a scalar called utility that is maximized. Similarly, consumers selecting between extended sequences choose the plan that maximizes the discounted value of the component outcomes. The findings presented here argue against the adequacy of the classical utilitarian view that decisions are evaluated solely on the basis of their ends rather than their means. We test this general premise for two consumer decision making scenarios: preference for not choosing and choice between extended sequences. This session shows that the process of deciding influences choices in a manner such that normatively equivalent outcomes result in different preference orders. This paper summarizes the session by providing a brief overview, and by describing representative ideas from the four papers comprising the session.

OVERVIEW

Marketers have extensively examined the decision rules that consumers use to choose from a given set of alternatives (Bettman 1979). One aspect of consumer choice that has received little attention pertains to the question of how consumers decide whether they are ready to choose or even more fundamentally, if they want to make a choice. As we know from our everyday experience, a question that consumers often face is not which alternative to choose but whether or not to make a choice. For example, a potential vacationer may choose between vacation spots, collect more information, or simply decide to postpone the holiday (Corbin 1980). It is somewhat surprising how little empirical attention has been devoted to this issue.

Consumers may sometimes choose to delay choice if the alternatives are unfamiliar or undesirable. Greenleaf and Lehmann (1991) identified five general factors that causes delay in consumer decision making. Delay arising from many of these factors, such as uncertainty, task avoidance, and risk of product performance can be justified on normative grounds. In such cases, the option of not choosing can be rationalized by a more complete inventory of individual preferences, one that captures the utility of not choosing. Such a specification of not choosing can be incorporated into the general framework of multiattribute utility theory (MAUT).

One case of preference for not choosing that is problematic for normative theory to handle arises when consumers are less likely to choose when a choice situation involves more than one attractive alternative. The cornerstone of economic viewpoint is that providing consumers with additional attractive alternatives can not decrease utility, and therefore, can not decrease the preference for choosing. In other words, the choice utility cannot be less than the utility of the most attractive option in that set (Meyer 1981). This leads to the obvious conclusion that having fewer options should never be more attractive than having more options. The first two papers question this premise and empirically show that people may prefer a restricted choice set to a larger one. The papers show that adding alternatives to increase choice may affect the willingness as well as the ability to choose. The resulting violation of the normative theory can not be accommodated within the classical framework.

One reason why consumers prefer to not choose when the choice set is expanded to include more attractive options is conflict. The increase in the number of available alternatives increases conflict (Berlyne 1957). Conflict arises because choices typically involve tradeoffs between attributes that appeal to different values. The increased conflict due to more number of attractive alternatives induces greater indecision, lowering the likelihood of choosing (Tversky and Shafir 1991). The difficulty of deciding between alternatives is not captured by the classical framework (e.g., the weighted additive utility model) which assumes that people have well defined attribute weights or preferences that are mapped onto the particular choice situation (see Shugan 1980 for a modified framework).

A second reason why people do not prefer to have a choice is when both alternatives are valued and there appears to be no clear reason to prefer one over the other. The rejection of the non chosen alternatives potentially induce feelings of responsibility, not considered in the classical framework. The feelings of regret are further moderated by the ease of imagining that the foregone alternative could have been selected (Kahneman and Miller 1986). When these emotions influence decisions before any consequences actually materialize, they are referred to as anticipatory regret. Thaler (1980) cites an instance of restricting choice even when restricting choice clearly has economic costs attached to it. For instance, members of the Israeli Army display a resistance to trading patrol assignments, even when it would be convenient for both individuals to do so, presumably due to anticipatory regret. The present view contends that a choice criteria in addition to the choice rule needs to be developed to determine if a choice is made. Furthermore, since consumers may not choose given attractive alternatives, cut off based strategies (e.g., conjunctive rules) allow us to infer when people will not choose, rather than when they will choose.

Since in usual choice situations consumers always have the options to search for more alternatives, an interesting issue that arises is whether consumers may sometimes voluntarily restrict the set of available options. Some evidence is found in the observation made by Thompson et al (1990) that participants in their study sometimes experienced a greater sense of freedom when placed in a restricted situation and gained a feeling of control by giving up control.

The second issue addressed in this session concerns the decision involving choosing between extended outcomes. Most economic analyses of preference between outcome sequences relies on the discounted utility model with a positive discounting function. According to this perspective, people should prefer desirable outcomes that occur as soon as possible and delay undesirable outcomes whenever feasible. However, Loewenstein (1985) showed that preferences for outcomes can also be affected by the utility from their anticipation. The utility of anticipation , positive for gains (savoring) and negative (dread) for losses, has been shown to result in preferences for delaying desirable outcomes and accelerating undesirable outcomes.

An interesting extension addressed in this session is its application to the domain of risky choice. One could argue that for anticipation to kick in requires that the outcome occur with a high degree of certainty. Furthermore, since probabilities of outcomes can be described for the occurrence or non occurrence of an outcome, it is possible that savoring and dread are susceptible to framing effects. For example, a probabilistic desirable outcome could be described in terms of its probability of success or failure. The frame describing the probability of success would generate positive utility of anticipation whereas the frame describing the probability of failure will generate negative utility of anticipation. Some preliminary evidence for this phenomena in the domain of risky choice.

The utility of an extended sequence is not only evaluated by its discounted worth but is also influenced by the sequence frame. The discounted utility model computes the value of outcomes extended over time to arrive at a single value. An independent factor that may distort the discounted utility in reality is the degree of saliency. The outcomes that are concentrated would be more salient, and as a result, have a greater impact on computed value of a sequence, than outcomes that are diffused over a longer period. This allows us to understand why certain borrowing and saving plans are preferred over others even though their discounted utility as predicted by the normative model may be equal or even lower.

SUMMARY OF ABSTRACT - S

The first paper "Choosing Not To Choose: Context and Task Effects on Indecision," by Ravi Dhar argues that an interesting phenomena that has been largely ignored in empirical studies of consumer choice is the topic of undecidability. The author explicitly incorporates the option of not choosing in the decision problems executed in an experimental setting, just as consumers often do. The author argues that consumers often delay choices even when the available alternatives are desirable. The factors that influence the level of indecision, such as the degree of anticipated regret and conflict, are shown to moderate the willingness to choose. The degree of conflict and regret is manipulated by expanding the choice context. The choice context imparts utility to choosing, either positive or negative, thereby influencing the preference for choosing. These results violate normative theory which states that consumer preference should not be context dependent. The results also support the premise that the process of choosing influences the utility experienced from the outcomes.

An instance of indecisiveness occurs when consumers are unable to discriminate sufficiently between options. The author manipulates the degree of indecision by changing the choice context and the nature of the task. This is consistent with the viewpoint that choice is often based on reasons that are generated in support of one option over another (Slovic 1975). Normatively, this should not influence consumer willingness to choose as relative advantage has little implication for instrumental effect on the attractiveness of the alternative. The context effect on indecision is however consistent with the notion that consumer preferences are constructed in the process of choice (Payne, Bettman and Johnson in press).

The choice context is manipulated by increasing the number of attractive alternatives or by varying the difficulty of making attribute tradeoffs among available alternatives. The experiments show that willingness to choose differs across contexts. Furthermore, the effect of indecision is reduced if the task is changed to allow subjects to choose more than one option. The observed violations in axiomatic choice theory are explained in terms of the assumption that the two psychologically different states of indifference and undecidability map onto the same behavioral response. We discuss the implication of these results for marketers and outline other factors that may moderate indecision, and are to be investigated in future research, such as accountability and positive affect.

The second paper "Decision Seeking and Decision Aversion," by Jane Beattie, Jonathan Baron, John Hershey, and Mark Spranca questions the normative models of economic choice that assumes that people would always rather have more choices than fewer choices. They forward two constructs: decision seeking and decision aversion and identify the major determinants of the two constructs.

Decision aversion occurs when an individual prefers to give up the more preferred items to avoid making a choice. An example of decision aversion would be to prefer either of your children alone to be educated at a good private school, but to send both to the worse public school because you could only afford the tuition for one child and were unwilling to choose which child would be sent to the good school. The causes for decision aversion are examined and a strong arguments is obtained based on subjects' fear of blame for poor outcomes as the major determinant.

Decision seeking involves a preference for the choice between two items that are not individually preferred to the single item. The main causes of decision seeking appears to be love of self-determination and desire to appear to cause good outcomes. The hypothetical role of the subject in the decision situation (e.g., doctor or a lay person) also affect the level of decision aversion and decision seeking, as do stable individual differences. These findings, the authors contend, argues against the descriptive adequacy of the classical utilitarian view that decisions should be evaluated solely on the basis of their ends rather than their means.

The third paper "Impact of sequence frames on borrowing and saving programs," by Drazen Prelec and George Loewenstein builds on the premise that intertemporal choice is strongly affected by the sequence frame, i.e., whether the relevant items are considered in isolation or whether they are perceived as part of a larger sequence. A central finding of this research is that when events are placed into a sequence, then the preferences of the representative individual become strongly oriented towards the future.

The present paper explores two hypotheses the sequence frame on consumer preference over borrowing and saving schedules. Decisions concerning purchase timing reveal consumer's utility for money which can then be used to predict choices in other settings. Whether this utility is experienced in a concentrated fashion or is diffuse over the lifetime of a durable is not considered fundamental to the analysis. Normative theory assumes that the pleasure or pain can be reduced the single dimension of utility.

The present research presents descriptive evidence that argues against the above premise. The first hypothesis states that the preference for different types of borrowing and saving programs depend on the degree to which the cash flows are psychologically coupled to the specific products. For savings programs, they predict a preference for a highly coupled arrangement, such as savings for a college tuition, rather than more loosely coupled ones, such as saving for some general category of expenditure. However, with borrowing programs, the loans that weaken the sequence frame and hence a preference for loose coupling would be more attractive. A loan taken out specifically to finance a recent vacation, for instance, should be unattractive whereas loans taken on credit card would be more attractive.

A number of explanations are explored for the observed patterns of preferences. For instance, the role of the sequence frame is affected by the temporal concentration (diffuseness) of the relevant costs and benefits (e.g., a vacation as opposed to consumer durable). The authors expect to show that concentration of benefits will increase the attractiveness of saving up for an item, and decrease the attractiveness of paying back for it. The reverse would be expected for consumer durables only if the loan program starts at the end of the life of the product, an unusual case.

The fourth paper "The Consumption of Uncertainty," by Dan Lovallo and Daniel Kahneman investigates the preference for the resolution of uncertainty surrounding uncertain outcomes, that are not immediately resolved. The delayed resolution of uncertainty is the norm in decisions ranging from new product introductions to marriage. The authors draw together research from both risky choice and intertemporal choice, and describe how the experience of unresolved uncertainty systematically affects the attractiveness of a prospect.

The authors build on the concept of savoring and dread (Loewenstein 1985) that have been empirically demonstrated to accompany future riskless prospects into the domain of risky choices. The anticipations has been shown to affect the utility for the outcome in a consistent manner: dread for losses and savoring for gains. The authors argue that both disappointment and elation are influenced by prior expectations and only become relevant for sufficiently high probability events.

Drawing form the principle of loss aversion, the authors argue that disappointment and dread dominate savoring and elation. The anticipated utility from a positive but uncertain event depends on the outcome valence as well as the probability of occurrence. For example, dreaming about a 10% chance to win a tour of europe may be fun (savoring), whereas a 90% chance of winning the same trip is reframed as a 10% chance of missing out (dread). These phenomena have implications for the timing of resolution for the uncertainty. The results show that people prefer immediate resolution of the uncertainty for the vacation when the probability of winning the vacation is high. However, when the probability of winning is small, people prefer to defer the timing of resolution. These results are replicated in the domain of losses. Furthermore, consistent with loss aversion, resolution timing for good outcomes are driven by disappointment whereas the timing of bad outcomes are largely determined by feelings of dread.

CONCLUSION

The research issues discussed here are consistent within the constructivist framework that argues that preferences or attitudes are not revealed but elicited in the process of making a judgment or choice. In the same way that preferences are ambiguous, consumer goals may also be imprecise and constructed in a given situation. Thus, the assumption made in hierarchical models of consumer choice, that the decision to choose is independent of which alternative to choose, may not always be valid. On the contrary, consumers may often decide whether or not to choose depending on the context. Furthermore, consumers may shift their preferences in the direction of alternatives that minimize conflict. This would provide a competitive advantage for the niche brand simply from the consumers manner of processing information (Carpenter, Glazer, and Nakamoto 1991).

REFERENCES

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

Berlyne, D. E. (1957), "Uncertainty And Conflict: A Point of Contact Between Information-Theory And Behavior-Theory Concepts," Psychological Review, 66, 329-339.

Carpenter, G., R. Glazer, and K. Nakamoto (1991), "Meaningful Brands from Meaningless Differentiation: The Dependence on Irrelevant Attributes," Working Paper, University of California, Berkeley.

Corbin, R. M., (1980), "Decisions that might not get made,", In T. S. Wallsten (Ed.), Cognitive Processes in choice and decision behavior, Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.

Festinger, L. (1964), Conflict, decision and dissonance, Stanford University Press.

Greenleaf, E., and D. Lehmann (1991), "Causes of Delay in Consumer Decision Making: An Exploratory Study," in Advances in Consumer Research, Vol. 18, ed. Michael R. Solomon and Rebecca H. Holman, IL: Association for Consumer Research.

Janis, I. L. (1977), Decision Making, New York: Free Press.

Kahneman, D. and D. T. Miller (1986), "Norm theory: Comparing reality to its alternatives," Psychological Review, 93, 136-53.

Levi, I. (1986), Hard Choices, Cambridge University Press.

Loewenstein, G. (1987), "Anticipation and the Valuation of Delayed Consumption," Economic Journal, 97 (September), 666-84.

Loewenstein, G., and D. Prelec (1990), "Preferences Over Outcome Sequences," Working Paper, Harvard Business School.

March, J. G. (1978), "Bounded rationality, ambiguity, and the engineering of choice," Bell Journal of Economics, 9, 587-608.

Meyer, R. J. (1981), "A Descriptive Model of Consumer Information Search Behavior," Marketing Science, 1, 93-121.

Payne, J. W., J. R. Bettman, and Eric J. Johnson (in press), "Behavioral Decision Research: A Constructive Processing Perspective," Annual Review of Psychology, vol. 43.

Prelec, D., and R. J. Herrnstein (1991), "Preferences or Principles:Alternative Guidelines for Choice," In R. Zeckhauser (Ed.), Strategic Reflections on Human Behavior, MIT Press.

Shugan, S. M. (1980), "The Cost of Thinking," Journal of Consumer Research, 7 (September), 99-111.

Slovic, P. (1975), "Choice between equally valued alternatives," Journal of Experimental Psychology: Human Perception and Performance, 1, 280-87.

Thompson, C. J., W. B. Locander, and H. R. Pollio (1990), "The Lived Meaning of Free Choice: An Existential-Phenomenological Description of Everyday Consumer Experiences of Contemporary Married Women," Journal of Consumer Research, 17 (December), 346-361.

Tversky, A., and E. Shafir (1991), "Decision Under Conflict: An Analysis of Choice Aversion," Working Paper, Stanford University.

Wright, P. L., (1975), "Consumer Choice Strategies: Simplifying Versus Optimizing," Journal of Marketing Research, 11(1), 60-67.

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