Consumer decision making within a goal-driven framework

23
Consumer Decision Making within a Goal-Driven Framework Robert Lawson William Paterson College ABSTRACT Research and theory in consumer decision making has been dominated by a perspective that assumes that a consumer knows what product category he or she needs. This limited view has resulted in equating consumer decisions with brand choice. The purpose of this article is to provide a framework for considering a wider range of consumer decision-making processes by linking them to different goals that consumers might pursue. A hierarchical goal structure consisting of four levels of goals; abstract principles or values, action programs, more concrete product acquisition, and brand acquisition goals, is proposed as the theoretical construct which ties together a wide range of consumer decision making phenomena. An experiment in which consumers think out loud in contemplating two levels of more familiar or unfamiliar goals reveals that their thoughts are constrained by the proposed goal hierarchy. The goals provide a useful framework for understanding decision- making processes involving product level consideration, set generation, and the evaluation of those self-generated sets. ©1997 John Wiley & Sons, Inc. Understanding how consumers make purchase decisions has long been a central issue among consumer behavior researchers. Over the years, research and theory on the consumer decision-making (CDM) process has focused squarely on the problem of choosing a most preferred brand from a set of several alternative brands that all belong to the Psychology & Marketing Vol. 14(5):427 – 449 (August 1997) © 1997 John Wiley & Sons, Inc. CCC 0742-6046/97/050427-23 427

Transcript of Consumer decision making within a goal-driven framework

Page 1: Consumer decision making within a goal-driven framework

Consumer Decision Makingwithin a Goal-DrivenFrameworkRobert LawsonWilliam Paterson College

ABSTRACT

Research and theory in consumer decision making has beendominated by a perspective that assumes that a consumer knowswhat product category he or she needs. This limited view hasresulted in equating consumer decisions with brand choice. Thepurpose of this article is to provide a framework for considering awider range of consumer decision-making processes by linking themto different goals that consumers might pursue. A hierarchical goalstructure consisting of four levels of goals; abstract principles orvalues, action programs, more concrete product acquisition, andbrand acquisition goals, is proposed as the theoretical constructwhich ties together a wide range of consumer decision makingphenomena. An experiment in which consumers think out loud incontemplating two levels of more familiar or unfamiliar goals revealsthat their thoughts are constrained by the proposed goal hierarchy.The goals provide a useful framework for understanding decision-making processes involving product level consideration, setgeneration, and the evaluation of those self-generated sets. ©1997John Wiley & Sons, Inc.

Understanding how consumers make purchase decisions has long beena central issue among consumer behavior researchers. Over the years,research and theory on the consumer decision-making (CDM) processhas focused squarely on the problem of choosing a most preferredbrand from a set of several alternative brands that all belong to the

Psychology & Marketing Vol. 14(5):427–449 (August 1997)© 1997 John Wiley & Sons, Inc. CCC 0742-6046/97/050427-23

427

Page 2: Consumer decision making within a goal-driven framework

428 LAWSON

same product category. Although advances in understanding thisbrand choice problem have been made (e.g., Bettman, 1979; Bettman,Johnson, & Payne, 1991), this article adopts the perspective that theCDM process is truly multifaceted and encompasses various particulardecision situations, including, but certainly not limited to, the brandchoice situation.

The primary theoretical construct that makes it possible to take sucha broadened view is goal hierarchy. People who sometimes but not al-ways function as consumers can be viewed as being capable of behavingin purposive, goal-directed, and self-regulating ways. When they do, it isappropriate to ask about the nature of the goals that are behind suchthoughts and actions. The basic idea to be developed in this article is thenotion that much (but not all) of consumer decision making takes placein the context of pursuing goals. These goals, which range in abstract-ness from general principles and life themes to concrete purchase inten-tions, function to shape the nature of the decision-making task.

For instance, the general understanding that brand choice resultsfrom an alternative evaluation process, in which consumers directly com-pare the alternatives in their consideration sets on their various attrib-utes, may be limited by the context provided by the brand choiceproblem. In a typical brand choice problem, the experimenter providessubjects with a set of brands from the same product category. The sub-jects’ task is to determine which brand from the set they most prefer.Typically, subjects are not given much in the way of background informa-tion concerning their task, such as what their motivation might be orwhy some brands and not others are to be considered. Rather, the unspo-ken assumption is that the subject finds him- or herself in need of theproduct category represented by the brands and must figure out whichone is best. Given that alternative-driven situation, it is hard to imaginethe consumer adopting any strategy other than one that compares thealternatives directly with one another on the basis of their attributes.

However, this alternative-driven laboratory situation may not cap-ture essential features of the real-life consumer decision-making con-text. One glaring limitation of this approach is that it disconnects theconsumer from her or his original motivation. At first glance, the tradi-tional CDM model that appears in most consumer behavior textbooksseems to provide an explanatory framework that suggests a more top-down, goal-driven perspective (e.g., Engel, Blackwell, & Miniard,1993). In this linear, multistage model, the consumer begins theprocess by recognizing a need, which is followed by the stages ofsearching for information and constructing consideration sets of alter-natives, evaluating the alternatives, making the purchase, and evalu-ating the consumptive experience. However, the need that isrecognized is assumed to be a needed product, some category of goodor service for which various brands (or types, or styles) exist. The CDMmodel reduces to a brand choice model precisely because of that as-

Page 3: Consumer decision making within a goal-driven framework

sumption. Once a product category is specified as the recognized need,the underlying motivation for having that need in the first place be-comes irrelevant, or at least less important, for decision-making pur-poses. Although the separation of more abstract goals from thedecision-making process describes most CDM studies, and probablysituations in which undecided consumers are confronted with a storeshelf containing different brands, other situations require the exami-nation of goals.

In terms of theory, it should be clear that need recognition does nothave to occur at the product level. A person could just as easily recog-nize that he or she needs to lose weight, or would like to become moreattractive, as to recognize the need for an exercise bike. Once the prod-uct assumption is removed as the outcome of the need-recognitionstage, the CDM model can begin to accommodate issues other thanthose concerned with brand choice. Primarily, the role of consumergoals is enhanced, and a view of the consumer as a self-regulating de-cision maker begins to emerge.

THE ROLE OF GOALS IN CONSUMERDECISION MAKING

According to the traditional CDM model, the recognition of a need ini-tiates the decision-making process, and makes it possible for the con-sumer to then generate a consideration set of alternatives, which inturn allows for selection. Although the alternative-driven view gener-ally has downplayed the role of goals in the process, some studies havestressed the goal-related nature of these stages. The few researchersthat have been concerned with the process of alternative generationhave viewed it as a matter of categorization, whether it occurs at thebrand level (Gutman, 1982; Narayana & Markin, 1975), or the productlevel (Klenosky & Rethans, 1988; Ratneshwar & Shocker, 1988). Fur-thermore, Barsalou’s (1983, 1985) work on goal-related categories hasmade it clear that ad hoc goals can function as categories insofar assets of examples are generated as possible means to reach these goals.In consumer research it has been demonstrated that consumer goalscan function as categories for the generation of product-level consider-ation sets (Klenosky & Rethans, 1988), and that consumers vary whatproducts they generate within the same broad product category ac-cording to specific usage goals (Ratneshwar & Shocker, 1991).

A goal-driven perspective is also implicated in the alternative evalu-ation process when the alternatives are not all different brands fromthe same product category. Johnson’s (1984) conclusion that choiceamong noncomparable alternatives (alternatives that cut across prod-uct category boundaries) is a matter of comparing the alternatives di-rectly on the basis of abstracted attributes that they have in common,

429GOAL-DRIVEN CONSUMER DECISION MAKING

Page 4: Consumer decision making within a goal-driven framework

430 LAWSON

has been questioned. Bettman and Sujan (1987) argued that Johnson’swithin-attribute, abstraction strategy was used because the typicalchoice task that was employed did not make goals or decision criteriareadily available, rather than because of the nature of the considera-tion set, per se. Park and Smith (1989) found further evidence of goal-driven (top-down) processing in a product-level choice task in asituation in which subjects were presented with a goal of satisfyingtheir needs for entertainment, and were provided with a considerationset of different product categories.

Together these studies point to the necessity of adopting some goal-related perspective when the focus of decision making deviates fromthe traditional brand choice task. The categories that serve as anchorsfor generating alternatives can be consumer goals, and alternativeevaluation processes appear to be goal related when the alternativesare product categories rather than brands within the same productcategory.

More than two decades ago, Wright (1973) clearly made the distinc-tion between alternative-driven and goal-driven modes of processingby describing two variants of the general linear compensatory attitudemodel. The Class I model begins by having consumers consider the at-tributes of products and their degree of liking for each attribute. Thisalternative-driven processing mode assumes that consumers have al-ready related the attributes to their separate needs in their value sys-tems along with the relative importance of those needs. Class IImodels, on the other hand, begin by focusing on that assumption, andthen provide an overall evaluation of how well the product as a wholesatisfies the need. According to Wright, these (goal-driven) models as-sume that consumers have already considered their beliefs about prod-uct attributes and how they combine. The purpose of this article is todevelop a CDM model that incorporates both modes of processing, andto present experimental data that examine implications of the model.The basic viewpoint taken here is that consumers are people who es-sentially are self-regulating and volitional decision makers (seeBagozzi, 1993), and who actively use a whole spectrum of goals to con-trol their own behaviors, including their consumer-related behaviors(see Huffman, Ratneshwar, & Mick, 1997).

A CONTROL PROCESS CONSUMERDECISION-MAKING MODEL

Control Process Theory

A goal-driven perspective of decision-making processes sets consumerdecision making within the broadest domain of purposive human be-havior. Even in our materialistic society, trying to achieve such goalsas happiness or a sense of inner peace may not immediately stimulate

Page 5: Consumer decision making within a goal-driven framework

thoughts about purchasing products, although these goals ultimatelymay have consumer implications. Such abstract goals, when operative,can exert a regulatory influence on our actions and related decisionmaking. Control process theory represents a general approach to un-derstanding how goals work in controlling behavior (Carver & Scheier,1982; Fiske & Taylor, 1991; Ford, 1987; Locke & Latham, 1990).

In research that is particularly relevant to understanding goals thatvary in abstractness, Carver and Scheier (1982, 1990) propose an em-bedded system of feedback loop mechanisms in which abstract goalsserve as reference values that have the role of monitoring actions.These goals do not necessarily become directly involved in routine,day-to-day activities, but may when the outcome of some action cre-ates a large enough deviation from what is consistent with higher-level goals. For example, a father’s values of peace and nonviolence,which have been dormant in the context of day-to-day life, may becomeactivated and salient when his son includes Power Ranger action fig-ures on his holiday gift list. This negative, discrepancy-reducing feed-back loop is the central concept of control process mechanisms, inwhich people can self-regulate their goal-directed behavior. The view-point taken here proposes that such a top-down executive level of reg-ulation is crucial for understanding a broad range of consumerdecision-making phenomena, from the lower-level, alternative-drivenbrand choice situation to the achievement of more abstract, higher-level goals.

According to the Carver and Scheier (1990) approach, relevant goalsmay be activated at different levels of abstraction. They propose athree-level hierarchy of feedback loops, incorporating the (highest) sys-tem concept, the principle and the program levels, which are interre-lated in that the output from the higher level becomes the referencevalue of the next lower level. In their example, the idealized self-image(system concept) may be realized in the principle of “being kind,”which may, in turn, find expression in the program of “shoveling snowoff neighbors’ walks.” The important theoretical point is that any con-crete act or decision can provide feedback for not only the immediate,lower-level (more concrete) goal, but for any relevant higher-level(more abstract) goal as well. This goal hierarchy provides the motiva-tion for a consumer to purchase a snowblower as well as to find a wayto be kind.

By allowing for the operation of goals that vary in their abstract-ness, control process theory provides a potential framework by whichconsumer decision making and action can be understood in a contextother than that of brand choice. Outside the laboratory, consumers arenot only under the influence of the immediate purchase situation, inwhich they may be confronted with a ready-made set of brands fromwhich to choose, but they themselves operate in a purposive, self-regu-lating way in which goals must play a prominent role.

431GOAL-DRIVEN CONSUMER DECISION MAKING

Page 6: Consumer decision making within a goal-driven framework

432 LAWSON

A Proposed Model

The central objective in developing this model of the CDM process is toaccount for both alternative-driven and goal-driven modes in the sameframework. Figure 1 presents the proposed theoretical framework.This theory is based on the Carver and Scheier (1990) embedded loopcontrol structure, but is extended to incorporate traditional consumerdecision-making issues. The theory includes four levels of goals thatserve to guide and exert executive control over action and consumerdecision making.

Figure 1 An outline of consumer decision-making processes within a hierarchicalgoal structure.

Page 7: Consumer decision making within a goal-driven framework

The two top levels, which are the principle and program levels de-scribed above, are directly operative in controlling action that is notnecessarily consumer related. The two lower levels of control, labeledproduct acquisition and brand acquisition, represent goals that arerelevant to traditional consumer decision making and purchase behav-ior in that the motivation for these processes is focused on the selec-tion of the best alternative. Goals realized at the product acquisitionlevel identify a needed product category, whereas brand acquisitiongoals identify the brand itself. Although action and consumer decisionmaking are viewed as separate subsystems, the abstract levels of con-trol may become linked with more traditional CDM processes in one oftwo ways. First, in a top-down, or feed-forward manner, after decidingon a certain program of action, the consumer is ready to enter a prod-uct-level decision-making process. Alternatively, by using a feedbackmechanism sensitive to discrepancies between the aftermath of pur-chases and the more abstract reference values, the consumer can eval-uate those purchase decisions against any higher level of referencevalue.

Traditional brand-level consumer decision-making processes mayproceed disconnected from the more abstract levels of control, or theymay tie into them. In the former case, brand-level consumer decisionmaking can be understood more accurately as a bottom-up, alterna-tive-driven process, whereas in the latter, it can be viewed as inte-grated into a top-down process. It is also important to note thatproduct-level consumer decision activities are understood as flowingfrom the program level of control, thus specifying that this level of con-sumer decision making is a goal-driven process. In fact, the model fur-ther implies that product-level choice cannot be understood as purelya matter of comparing product alternatives without first connecting tohigher-level reference values.

In addition, the model makes the point that a person or person-as-consumer can enter this self-regulating decision-making process atvarious levels, depending upon the outcome of the need recognitionprocess. Needs begin to be translated into actions when they are real-ized as goals, which themselves can be understood as reflecting vari-ous degrees of abstractness. Goals that express abstract valuesactivate the principle-level reference value, and more concrete goalsthat are related to performing activities directly activate the program-level reference value. Neither type of goal is inherently a consumergoal, in that it primarily functions to control action; but either may be-come connected to CDM processes in a top-down fashion when con-sumption-related means to achieve goals are brought into the process.The assumption behind the traditional CDM model has been that theoutcome of the need-recognition process is the goal to acquire a partic-ular product. This model incorporates that assumption by identifyingproduct-acquisition-level goals, which then lead to brand-choice

433GOAL-DRIVEN CONSUMER DECISION MAKING

Page 8: Consumer decision making within a goal-driven framework

434 LAWSON

processes. Finally, to complete the description of the range of goalsthat may be realized, it is possible to arrive at the goal of acquiring aparticular brand, in which case any level of consumer decision makingis rendered unnecessary.

The hierarchical structure of this framework implies that the clear-est view of the goal-driven CDM process is revealed when principle-level reference values become salient in a consumer-related context.The embedded feedback loop structure of the theory suggests that avery abstract goal, when serving as a controlling reference value, leadsto a decision-making process that considers various courses of actionor programs as alternatives because the program represents both theresult of the decision making at that level and the reference value forthe next lower level of control. Less abstract goals that are realized atthe program level in terms of a task to accomplish naturally lead theconsumer to consider various products that may be employed in theservice of that activity or task. It is further implied that variousbrands will be brought to mind only when a product has been identi-fied as the controlling consumer goal.

Summarizing, the current model views the CDM process as one thatfits into the basic multistage framework of the traditional model (e.g.,Engel et al., 1993), but goes beyond it by proposing that the initiatingneed-recognition stage can be realized at different levels of abstract-ness. This inclusion has the effect of converting the traditional CDMapproach from a relatively simplistic process that focuses attention onthe brand choice problem into a richer, more complex system that canincorporate a wide range of consumption-related activities and deter-minations. The basic theoretical mechanism that unites such diverseconsumer issues as brand-level choice, product-level choice, the gener-ation of alternatives, activity determination, and the role of abstractgoals and values is the Carver and Scheier (1990) control process sys-tem.

The purpose of this experiment is to tap into the upstream regionsof the proposed model by presenting subjects with goals that do notspecify or imply a particular product so that they can make consumer-related decisions as means to reach these goals. The goals vary in theirlevel of abstractness (principle and program levels) and in their famil-iarity.

Goal abstractness is conceived to be the primary independent vari-able because it maps into the control process hierarchy. The more ab-stract goals represent values and are designed to correspond to theprinciple level, while the more concrete goals specify activities and areakin to the program level of control. According to the proposed model,it is anticipated that the initial reaction to the concrete goals would beto generate product-level consideration sets, whereas the abstractgoals would tend to be associated with thoughts of goal construal andactivity generation.

Page 9: Consumer decision making within a goal-driven framework

Goal familiarity, on the other hand, should not affect one’s entry intoa particular level on the control process hierarchy, but rather affectthe ease with which product alternatives are retrieved from memory.Familiar goals are expected to lead to larger and more easily retrievedconsideration sets than are unfamiliar goals.

Both goal variables are expected to exert their primary influence onthe process of consideration set generation. By using goals in whichproducts are not yet specified, it is anticipated that the considerationsets predominantly will be identified as product category names ratherthan as brand names. Once the product-level sets are generated, theprocess of alternative evaluation is expected to be an across-attributeone in which the alternatives are evaluated according to their fit withthe goal, rather than one that directly compares the alternatives witheach other. There is no expected impact of goal type on alternativeevaluation strategy.

METHOD

Overview and Design

This experiment is designed to gather think-out-loud protocol data ofsubjects’ strategies as they are guided through three phases of a goal-initiated decision-making task. In Phase 1 they are allowed simply toreact to the presented goal. Phase 2 asks specifically for the generationof consumer product alternatives to be used in the construction of aconsideration set if they have not already spontaneously done so inPhase 1. Then, in Phase 3, subjects are asked to choose a single, mostpreferred alternative from their self-generated list.

A mixed design was used with goal abstractness and familiarityserving as between-subject factors, and with repeated measures ongoals. Each subject provided verbal protocols to two goals in one of theconcrete familiar (CF), concrete unfamiliar (CU), value-level familiar(VF), or value-level unfamiliar (VU) goal conditions. In the analysesthat follow, the primary concern is to assess the effects of goal ab-stractness and familiarity. Indications of the consistency of responsesto both goals in a goal category become evident by considering interac-tion effects involving the within-subjects factor.

Goal Selection

Eight goals, two in each of the four goal categories, were chosen to beused in the experiment. They are presented in Table 1. All goals werestated in such a way as to leave open the possibility that various prod-uct categories could be instrumental in achieving the goals. On the ab-stractness dimension, it was intended that the more abstract goals

435GOAL-DRIVEN CONSUMER DECISION MAKING

Page 10: Consumer decision making within a goal-driven framework

436 LAWSON

correspond to general positive values, and that the less abstract goalsrefer to particular concrete occasions for action. Goal familiarity wasconceived to reflect the extent of one’s past experience with thinkingabout a stated goal as a goal.

Pretests in which subjects rated the goals on relevant abstractnessand familiarity scales, confirmed the expected differences betweengoal categories and similarities within goal categories, with one excep-tion: the VU goals were rated as being more familiar than the CUgoals. After several attempts to create value-level goals that weremore unfamiliar failed, it was concluded that because people haveheard of the abstract goals, and perceive them as desirable, they areinclined to rate them as somewhat familiar, even though they mayhave lacked personal experience with them.

Subjects

Subjects were 58 full-time MBA students from a large eastern univer-sity. They were between 21 and 36 years old, with a median age of 25.They served in individual sessions lasting 30–45 minutes and werepaid $5 for their participation. They were assigned to one of the fourgoal conditions in a predetermined random order. Data from one sub-ject in the VF condition were eliminated from analyses involving pro-tocols due to an inaudible tape recording. This resulted in there being15 respondents (7 male) in the CF condition, 15 (6 male) in the VUcondition, 14 (6 male) in the CU condition, and 13 (7 male) in the VFcondition for the main protocol analyses.

Procedure

First subjects were given an instruction sheet that stated that the pur-pose of the study was to learn more about how consumers thought

Table 1. Goals Used in Experiment.

Concrete-Familiar (CF)G1 To eat something nutritious for breakfast.G2 To keep cool on a hot night.

Concrete-Unfamiliar (CU)G3 To prevent birds from crashing into your house window.G4 To remove a poisonous snake from your child’s outdoor play area.

Value-Familiar (VF)G5 To be happy.G6 To live a comfortable life.

Value-Unfamiliar (VU)G7 To increase your sense of inner harmony.G8 To give the impression of being a sensitive person.

Page 11: Consumer decision making within a goal-driven framework

about various goals they might have, and that they would be asked tothink out loud as they considered two different goals. The tenor of theinstructions emphasized goal achievement rather than consumer deci-sion making. Subjects were next given a practice goal at the same ab-stractness level as their assigned condition.

The first of two goals were then presented, and the experimenterguided the subject through the three phases of the task. In Phase 1, ifsubjects spontaneously generated alternatives, Phase 2 either wasskipped or began with a prompt for additional alternatives. Phase 2ended when the subject felt satisfied with the extent of his or her con-sideration. The experimenter initiated Phase 3 by reviewing the gener-ated set and asking the subject to choose the most preferred setmember. After selection was completed, and subjects were given a 1–2min break, the second goal was presented. The presentation order ofthe two goals was counterbalanced.

Next, subjects completed a short posttask questionnaire, which con-sisted of a group of manipulation check items and a group of questionsconcerning their alternative generation and evaluation experience.The latter items were intended to supplement the protocol data. Fi-nally, subjects were debriefed, were asked not to divulge the details ofthe experiment to their classmates, and were given their payment.

Protocol Categorization Scheme

Subjects’ protocols were categorized according to the scheme that ispresented in Table 2. Consideration set generation (SG) statementscovered the first two phases of the procedure, and are organized intothree subgroupings. Goal construal (SG 1–3) statements were the sub-jects’ attempts to arrive at a personalized meaning of the goal. Activityidentifications (SG 4–5) cited potential actions that might be em-ployed as means to achieve the goal, and product identifications (SG6–9) were statements that included mentions of products that wouldbe considered for goal attainment. Alternative evaluation (AE) cate-gories were restricted to Phase 3 protocols. They identified strategiesinvolving the direct comparison of two or more alternatives (AE 1–4),the evaluation of one alternative at a time (AE 5–8), and the prelimi-nary elimination of some alternatives (AE 9–11). By rearranging thespecific categories, attribute-based (AE 1, 2, and 10) and goal-related(AE 4–7 and 9) strategies could be assessed. Decision criterion (DC)categories pinpointed subjects’ reasons for selecting their most pre-ferred alternative. The categories utilized focused on the decision crite-rion as a means to achieve a goal (DC 1–4), nongoal-related benefitsstemming from the preferred alternative (DC 5), and the comparisonof common attributes (DC 6). The latter two categories were adaptedfrom Park and Smith (1989) in an attempt to capture non-goal-relateddecision criteria.

437GOAL-DRIVEN CONSUMER DECISION MAKING

Page 12: Consumer decision making within a goal-driven framework

438 LAWSON

Table 2. Protocol Response Categorization Scheme.

Response Category Description

Consideration Set Generation (SG)Goal construal Attempts to understand the personal impli-

cations of a goalSG1 goal restatements/definitions Restating or defining to oneself the goal

as presented. (“Having a goal as beinghappy would be to . . . concentrate onjust happiness.”)

SG2 goal analysis Translating goal into an idea that haspersonal meaning (“I think I’m happywhen I’m busy”), establishing a subgoal(“I guess in living a comfortable life,my first goal would be to obtain em-ployment”), or stating a situationwhere the goal could be realized.

SG3 miscellaneous goal related Default category. Goal construal state-ment, but not SG1 or SG2.

Activity identification Actions that might be used as a means ofachieving the goal

SG4 first mention of activity First mention of a mental (“Think aboutthings I enjoy doing.”—Happy) orphysical (“I would take a shower.”—Keep Cool) action, contingent upon goalcomprehension.

SG5 miscellaneous activity Combined category including elabora-related tions of previously cited activities or

identifications of activities to avoid doing.Product identification Consumer products that might be used as a

means of achieving the goalSG6 product bursts Two or more products mentioned in a

string.SG7 identification of single First mention of a product identified at

product the product category level.SG8 identification of single brand First mention of a product identified at

the brand level.SG9 miscellaneous product- Related to products identified without

related actually naming new products. In-cludes elaborations, products to avoid,and introspective remarks.

Alternative Evaluation (AE) StatementsAlternatives directly compared Evaluation involving two or more alterna-

tives in a comparison processAE1 concrete attributes in Comparison based on relatively concrete

common attributes. (“It would probably be muchmore expensive, but it would be worthit—to have the a/c.”)

AE2 abstract attributes in Comparisons based on relatively abstractcommon attributes. (“They all combine, the

clothes, make-up, and accessories. Theyall have to do with what your look is.”—Sensitive Person)

(continued)

Page 13: Consumer decision making within a goal-driven framework

439GOAL-DRIVEN CONSUMER DECISION MAKING

Table 2. (Continued from previous page).

Response Category Description

Alternative Evaluation (AE) StatementsAE3 basis not revealed Comparison without comment. (“It’s be-

tween grill and mesh.”—Birds)AE4 goal-related basis Comparison based on how well the alter-

natives achieve the goal. (“Like a bigdose of inner harmony instead of lotsof little doses.”—Harmony: on compar-ing the merits of a vacation vs. multiplemanicures)

Evaluation of individual alternatives Evaluations involving one alternative at atime

AE5 fit with given goal Alternative evaluated as a means toachieve given goal. (“I’d call the [ani-mal control] service first, and thenmaybe they could tell me how to keepit from happening again.”—Snake)

AE6 fit with subgoal Alternative evaluated as a means toachieve a subgoal of the given goal.(“I’d say easy-going clothing. Just sopeople aren’t scared off by me eitherbeing too liberal or too conservative.”—Sensitive Person)

AE7 fit with integrated goal The comparison goal combines aspects ofthe given goal with some other goal.(“French toast may be nutritious. I see‘nutritious’ as low calorie.”—Breakfast)

AE8 basis not goal related Alternative evaluated against nongoalcriteria, such as specific attributes orbenefits. (“Investments. Not having ajob at the moment, it’s a little difficultto do that.”—Sensitive Person)

Preliminary elimination of Eliminating one or more alternatives prioralternatives to additional evaluationAE9 goal-related basis Alternative(s) eliminated because of not

fitting with some goal. (“Bird houses—I don’t know if that would do it, to dis-tract him from crashing into mywindows.”)

AE10 attribute-related basis Elimination based on some nongoal-related product attribute. (“I doubt if Icould go to the Humane Society andpick up dogs and cats all that often,without having a zoo at some point.”—Sensitive person)

AE11 elimination basis not “I wouldn’t want the tapes or the incense,revealed so we can eliminate those.” (Harmony)

Decision Criteria (DC) for Final Product SelectionDC1 means to achieve given goal Given goal cited as reason for choice.DC2 means to achieve subgoal Some subgoal cited as reason for choice.

(continued)

Page 14: Consumer decision making within a goal-driven framework

440 LAWSON

Two coders independently classified more than 1,800 statementsfrom 57 subjects, working on a total of 114 goals. Encoding reliabilitywas assessed for the different major category groupings separately.Cohen’s kappa for the SG, AE, and DC statements were 0.94, 0.86, and0.84 respectively, indicating that for each analysis agreement betweenthe two observers is much stronger than if the classifications were sta-tistically independent.

RESULTS

Manipulation Check

Four items on the posttask questionnaire ask subjects to rate the twogoals that they considered on familiarity (frequency of thoughts andperceived knowledge) and abstractness (different types of activitiesand abstractness/concreteness). The mean scores on the 7-point ratingscale items are presented for each goal in Table 3. The same patternof results as in the pretest was found, including the finding that theCU goals were judged to be more unfamiliar than the VU goals. Inparticular, the familiar goals (M 5 4.98) were judged to be signifi-cantly higher in familiarity [F(1,54) 5 61.64, p , .001] than were theunfamiliar goals (M 5 3.02), and the value-level goals (M 5 5.57) weregiven higher abstractness ratings [F(1,54)572.58, p , .001] than themore concrete goals (M 5 3.66). However, the two VU goals (M 5 3.70)were rated higher on familiarity [F(1,27) 5 17.16, p , .001] than werethe two CU goals (M 5 2.34). There appears to be a limit on how unfa-

Table 2. (Continued from previous page).

Response Category Description

Decision Criteria (DC) for Final Product SelectionDC3 means to achieve goal not

related to given goal Default category. Goal is cited, but notDC1, DC2 or DC4. (“I like the fruitbecause it has the most flavor, and it’seasy to eat.”—Breakfast)

DC4 means to achieve integrated Integrated goal cited as reason for choice.goal (“Drapes or blinds. It would be some-

thing that would be decorative inside. . . as well as to keep the birds fromcrashing.”)

DC5 product-specific usage Product attributes cited as reason forcondition choice. (“The health club . . . it’s im-

portant to me . . . and very attain-able.”—Harmony)

DC6 common benefit character- Alternative chosen because it scoredistics higher on specific attributes than other

alternatives. (Category not utilized.)

Page 15: Consumer decision making within a goal-driven framework

miliar an understandable abstract goal can be. Nonetheless, in termsof the goal manipulations employed here, the two VU goals are stillrated as less familiar [F(1,54) 5 18.04, p , .001] than the four famil-iar goals. In addition, the three-way interactions involving the within-subjects factor of goals were not significant [F(1,54) 5 2.88 and 0.08,for the familiarity and abstractness analyses, respectively], which in-dicated that both goals within a category were perceived similarly.

Alternative Generation Processes

The basic prediction from the control process CDM model is that thememory access to consideration set products is more direct for the con-crete goals than for the value-level goals, and that the initial reactionto the value-level goals will be thoughts about goal construal and goal-related activities. This outcome is tested by comparing the number ofproduct-level identification, activity identification, and goal analysisstatements made in Phase 1 of the task (see Table 4). A highly signifi-cant response category 3 abstractness interaction [F(2,106) 5 76.37,p , .001, MSE 5 1.64] reflected the clear tendency for subjects to nameproducts as part of their initial response to the more concrete goals(M 5 2.53) as opposed to the value-level goals (M 5 0.09), but to en-gage in more goal analysis in response to the value-level goals(M 5 1.75) as opposed to the concrete goals (M 5 0.02). However, con-trary to expectations, activities were mentioned just as frequently tothe concrete goals (M 5 1.22) as to the value goals (M 5 1.25). This re-sult stems from the fact that two of the concrete goals (the Keepingcool and Snakes goals), which apparently did not completely specifythe means by which to reach them, stimulated frequent thoughts

441GOAL-DRIVEN CONSUMER DECISION MAKING

Table 3. Goal Abstractness and Familiarity Manipulation Checks.

Abstractness Familiarity

Goal M sd M sd

Concrete-Familiar (N 5 15)G1 breakfast 4.27 0.90 5.00 1.49G2 keep cool 4.00 1.24 4.23 1.29

Concrete-Unfamiliar (N 5 14)G3 birds 3.18 1.35 2.43 1.09G4 snake 3.18 0.87 2.25 1.03

Value-Familiar (N 5 14)G5 happy 6.00 0.81 5.07 1.37G6 comfortable 5.21 1.35 5.61 0.86

Value-Unfamiliar (N 5 15)G7 harmony 5.70 0.94 3.83 1.37G8 sensitive person 5.37 0.88 3.57 1.16

Note. All means were based on 7-point rating scales. High scores correspond to greater abstractness andgreater familiarity.

Page 16: Consumer decision making within a goal-driven framework

442 LAWSON

about activities. The variability of response patterns of the two goalswithin a goal type is reflected by a significant three-way interaction[F(1,53) 5 6.01, p , .05, MSE 5 0.54]. Although response type stronglyinteracted with goal abstractness, it did not with familiarity[F(2,106) 5 2.67], thus establishing the uniqueness of abstractness ef-fects on initial problem-solving thoughts.

Goal familiarity, on the other hand, is predicted to affect the easewith which various products are brought to mind once the consumerunderstands that the task is to generate products. Two measuresthat reflect retrieval ease are the number of products generated andthe number of statements in which unbroken strings of two or moreproducts are produced. As predicted, there were significant main ef-fects of familiarity for both set size [F(1,53) 5 23.01, p , .001,MSE 5 3.33; M 5 6.39 vs. 4.14] and number of product bursts[F(1,53) 5 18.59, p , .001, MSE 5 0.80; M 5 1.29 vs. 0.57]. Con-versely, goal abstractness did not affect these two measures of re-trieval ease [F(1,53) 5 0.04 and 1.29, for set size and product bursts,respectively].

Another way in which increasing degrees of goal familiarity couldaffect the set-generation process is by making it more likely to re-trieve products at the brand level rather than at the product-classlevel. This could occur because specific memories tied to the familiargoals might include brand-level designations for the rememberedproducts. If such a retrieval process was at work for the familiargoals, it was not revealed in this task. Overall, brand names wereinfrequently produced, accounting for only 7.2% of product set mem-bers, and their retrieval was not affected by familiarity[F(1,53) 5 2.08, MSE 5 0.43; M 5 0.52 and 0.24, for familiar and un-familiar goals, respectively].

Table 4. Mean Number of Product Identification, Activity Identification,and Goal Analysis Statements for Each Goal in Phase 1.

Goal

ResponseCF CU VF VU

Category G1 G2 G3 G4 G5 G6 G7 G8

Product identificationM 3.33 2.67 3.00 1.07 0.08 0.23 0.07 0.00sd 2.69 1.76 1.52 0.73 0.28 0.44 0.26 0.00

Activity identificationM 0.13 2.53 0.64 1.57 1.31 0.46 1.20 1.93sd 0.35 1.60 0.75 0.94 1.49 0.88 1.37 1.34

Goal analysisM 0.07 0.00 0.00 0.00 1.23 2.85 1.53 1.47sd 0.26 0.00 0.00 0.00 0.83 1.77 0.92 0.92

Page 17: Consumer decision making within a goal-driven framework

Alternative Evaluation Processes

In Phase 3 of this experiment, subjects select their most preferred al-ternative from those that they themselves had generated in the firsttwo phases. It is anticipated that the process of evaluating the alterna-tives will be an across-attribute, goal-related one as opposed to awithin-attribute one. The tendency to engage in these two strategies isassessed by counting the number of Phase 3 strategy statements as-signed to these two categories. As expected, subjects clearly preferred[F(1,53) 5 130.26, p , .001, MSE 5 1.24] the goal-related strategy(M 5 2.58) to the attribute-based strategy (M 5 0.19). There were nosignificant effects involving abstractness or familiarity.

Separate encodings were made of the reasons subjects gave to ex-plain their final preference decisions. Consistent with the alternativeevaluation strategies described above, it is predicted that subjects willutilize decision criteria categories that are based on considerations ofthe product being used in achieving a goal as opposed to categoriesthat reflect assessing the product on the attributes that may be sharedby other set members. To test this hypothesis, the following decisioncriteria statement categories were compared: (1) target goal related,(2) nontarget goal related, and (3) nongoal related. (See Table 2 for amore detailed description of the protocol categories used.) Across allsubjects and all goals, of the 110 decision criteria statements assessed,all but 2 were classified as goal achieving, and only 6 of the 108 goal-achieving criteria were categorized as pertaining to some nontargetgoal [x2(2) 5 175.49, p , .001]. At the individual goal level, subjectsoverwhelmingly utilized a decision criterion that involved assessingthe chosen alternative with respect to an interpretation of the givengoal.

Finally, it is noted that a closer inspection of the alternative evalua-tion and decision criteria responses reveals an unexpected, but inter-esting pattern of evaluation for three of the concrete goals. With theexception of the Keeping cool goal, alternative evaluation for concretegoals, but not the value-level goals, involved greater use of the inte-grated goal category. This complex pattern is partially reflected by sig-nificant abstractness 3 response category interactions for alternativeevaluation statements AE 5–8 [F(3,159) 5 6.72, p , .001, MSE 5 0.86],and for the decision criteria categories DC 1–4 [F(3,159) 5 13.20,p , .001, MSE 5 0.41]. Often, the nutritious breakfast goal is inte-grated with tasting good, the crashing birds goal becomes merged withwindow decorating aesthetics, and the poisonous snake goal also stimu-lates more general thoughts about safety. It is as if the process of con-sidering various product alternatives had the effect of importingadditional attribute-related criteria into the decision-making process.Thus it was hard for subjects to evaluate breakfast-food alternativeswith respect to the given goal of nutrition without also evaluatingthem on taste. Although it is not clear why this creative type of

443GOAL-DRIVEN CONSUMER DECISION MAKING

Page 18: Consumer decision making within a goal-driven framework

444 LAWSON

processing would occur to only some goals, the fact that it happens at-tests to the importance of trying to understand consumer decisionmaking from a goal-driven perspective.

DISCUSSION

On the whole, the results of this study lend support to the idea thatknowledge of consumers’ goals is necessary for a more complete under-standing of the entire consumer decision-making process. When con-sumers are trying to achieve the more abstract goals that were used inthis study, their considerations led to the generation of sets of alterna-tives that consisted of different product categories rather than of dif-ferent brands within the same product category. Furthermore, whenchoice was based on these self-generated, product-level sets, process-ing turned out to be a matter of determining which product was bestsuited to help achieve the goal, rather than a matter of direct, at-tribute-based comparisons among the alternatives.

This experiment also demonstrated the separate effects of goal ab-stractness and familiarity on consideration-set-generation processes.Goal abstractness exerted its primary effect on consumer’s initial reac-tion to the goal. The concrete goals stimulated more immediate at-tempts to retrieve products that fit the goal, whereas value-level goalsled consumers to engage in the preliminary process of goal construal.Mentions of goal-related activities were tied to how completely thewording of the goal specified performing a single activity program. Forexample, the goal To keep cool on a hot night inspired a relatively highnumber of activity statements even though it is relatively concrete.Thus, reacting to the goal by mentioning possible activities cut acrossthe levels of abstractness.

The effects of goal familiarity on set generation seemed limited toretrieval processes once the consumer had actually begun generatingproducts. As predicted, greater familiarity led to easier retrieval ofproducts as reflected by larger set sizes and more statements contain-ing uninterrupted strings of products. However, the possible effect onthe number of brand names generated did not materialize. In fact,brands were hardly mentioned at all. This outcome points to the psy-chological reality of the hierarchical nature of the CDM process and itsunderlying goal structure. The generation of brand names dependsupon the prior identification of a product category as the outcome of aneed recognition stage. In this study, in which the goals did not specifythe identity of a single product, subjects understood their task to beone of product-level set generation.

Once the product-level consideration sets were generated, and sub-jects evaluated those alternatives, it was clear that they used aprocess very different from that described in the typical brand-choice

Page 19: Consumer decision making within a goal-driven framework

situation. Their strategy, which held across all four goal conditions,was to assess how well the various products functioned as means toachieve a goal, rather than directly to compare the alternatives ontheir attributes. In addition, the decision criterion itself was almost al-ways identified as a means to achieve the presented goal or a relatedgoal.

The proposed control process CDM model incorporates both the typ-ical brand choice results, which represents an alternative-driven modeof processing, and the current results, which stem from a goal-drivenmode. According to the model, the brand-choice situation is one inwhich the goal is activated at the product-acquisition level, withoutreference to any higher level of control. The decision-making task isone of determining the best brand alternative from the provided set, inwhich all the alternatives belong to the same product category. In thissituation, the most natural way for consumers to approach the task isto use the within-attribute strategy.

By way of contrast, the goals utilized in this study, which were spec-ified at the principle and program levels, succeeded in creating a goal-driven processing mode. The subtasks of understanding the goal,generating the alternatives, evaluating the alternatives, and selectingthe most preferred alternative were all accomplished with the goal inmind. In this situation, the generated product alternatives were differ-ent product categories, and the resulting product-level choice task wasexecuted in a manner very differently from the alternative-drivenbrand-choice task.

The above conclusions regarding the success of this study are lim-ited by the constraints associated with external validity problems in-herent in the experimental situation. The primary shortcoming in thisregard stems from the fact that the goals were imposed upon the par-ticipants, rather than provided by them. To some extent, the familiar-ity variable captured some of the awkwardness of dealing with outsidegoals, but even the more familiar goals lacked the element of personalgoal ownership. In attempting to deal with these goals, subjects re-vealed the strong influence of their personal goals by using them tocreate, and then think about achieving, integrated goals. Whether sub-jects decide to deal with the exact goal provided, some subgoal, inte-grated goal, or unrelated goal, they are able to generate a set ofproduct-level alternatives and evaluate them according to theirprospects of reaching that goal. A study in which consumers shed lighton decision-making processes regarding their own personal goals isleft to future research.

Compared to the traditional CDM model, the proposed model offers afar less restricted view of consumer decision-making processes. Notonly can the current model incorporate the traditional brand-choice sit-uation, it can provide a theoretical framework for studying such issuesas product-level choice and consideration-set-generation processes (as

445GOAL-DRIVEN CONSUMER DECISION MAKING

Page 20: Consumer decision making within a goal-driven framework

446 LAWSON

was done here), the role of activity programs, how values are involvedin the self-regulation of actions and decisions, and many more. As waspreviously described, the control-process CDM model provides linkagesbetween abstract levels of goals that reside within consumers as valuesor action programs and the choices that consumers make and actionsthat they do. For example, the value a concern for animal rights couldbe incorporated into a consumer-related, program-level goal such asavoiding buying and using products marketed by firms that are not sen-sitive to animal rights. Such a goal-driven approach could leave itsmark on consumer decision making in the form of actively boycottingcertain products or brands. On the other hand, a more passive stanceby a consumer holding the same value might not result in formulatingthe above goal, but that value could be manifested as a decision crite-rion in an alternative-driven context. For example, when confrontedwith a choice set of lipstick brands, such a consumer might evaluateBrand A less favorably than otherwise because of knowledge that thefirm engages in animal testing.

Another interesting question concerning brand choice that thisframework suggests is whether the brand-choice process in a goal-driven situation will be goal related or attribute based. Will the con-sumer who has generated a consideration set of cosmetics brandsbased on animal rights continue to evaluate the alternatives on thebasis of that goal, or use an attribute-based process, which considersthe full range of attributes, on the grounds that the animal-rights is-sue has already been resolved?

Although the control-process CDM model does increase the range ofissues to consider, it can be criticized for being too broad in scope. Un-derstanding how people use values to regulate their actions is a hugeissue that is not directly related to consumer behavior. However, byconsidering how values can be linked to action programs, and how ac-tion programs, in turn, can be linked to the world of consumer decisionmaking via product-level choice or a negative feedback loop, these ab-stract psychological issues become consumer-behavior issues. Themodel provides a framework for understanding a wide range of con-sumer decision-making issues by incorporating a hierarchical goalstructure. When higher-level goals are activated, consumer decisionmaking can be understood as a goal-driven process, while it takes oncharacteristics of an alternative-driven process when lower-level goalsare salient.

The model can provide insights to a wide range of marketing-related consumer decision-making issues. Situations that have beenunderstood as brand-choice problems may benefit by being framedat a higher goal level. For example, the problem of selecting arestaurant (from some consideration set) instead may be analyzedas stemming from the need to dine out because one is tired of cook-ing, which is linked to the program or activity goal level. The activa-

Page 21: Consumer decision making within a goal-driven framework

tion of this level of goal implies a product-level decision-makingtask, which may involve consideration of alternatives such asFrench, Chinese, Italian, fast food, et cetera. According to the model,once this determination is made, the consumer’s decision-makingprocess has progressed to the more concrete product level of goal.Not only is the consumer aware that he or she wants to dine out, butthat he or she wants to dine out at an Italian restaurant. At thispoint the brand-choice task of selecting the preferred Italian restau-rant can proceed.

Perhaps the most important applications of this model are relatedto situations in which the value level goals are activated. For exam-ple, an investor may arrive at a plan for purchasing stocks and mu-tual funds that takes the social responsibility record of firms intoaccount. In order to understand this consumer, instead of startingfrom a set of alternatives from which to choose, a more completeanalysis would begin with a consideration of the influence thatdeeply held values have on shaping one’s investment goals. Such aconcern would likely have an effect on the investor’s considerationset and decision-making process. The current model provides aframework for incorporating values into both action control and con-sumer decision making.

Indeed, much of the field of social marketing can be viewed as at-tempting to get people to adopt certain attitudes or do certain actionsthat are socially desirable. This model suggests that the most stablepositive outcomes are grounded in value-level goals. For example, thepersonal decision to boycott can be viewed as a consumer-related ac-tion program that is under the regulatory control of values. An inter-esting area for future research would be to use the current frameworkto explore the factors that inspire some consumers to take actionagainst certain companies. Do boycotters forego receiving the benefitsof certain products in order to belong to a social movement, to be in-strumental in actually changing company behavior, or simply to followtheir consciences?

Although marketers have long been aware of the impact of valueson consumer decision making, not much is known about how to influ-ence their operation. The means–end chain literature (see Peter & Olson, 1996) suggests that advertising is successful to the extent thatthe product and the consequences of its use can be connected to val-ues, but that approach does not explain how values can motivate andcontrol consumer behavior by initiating a goal-driven process. What isrequired are methods that work on the deeply held values of con-sumers. Ironically, from the social marketers’ perspective, perhaps themost effective example of such a marketing program is not some anti-drug-abuse effort, but the National Rifle Association’s campaign toconnect gun ownership with patriotism by invoking the SecondAmendment to the Constitution.

447GOAL-DRIVEN CONSUMER DECISION MAKING

Page 22: Consumer decision making within a goal-driven framework

448 LAWSON

REFERENCES

Bagozzi, R. P. (1993). On the neglect of volition in consumer research: A cri-tique and proposal. Psychology and Marketing, 10, 215–237.

Barsalou, L. W. (1983). Ad hoc categories. Memory and Cognition, 11,211–227.

Barsalou, L. W. (1985). Ideals, central tendency, and the frequency of instanti-ation as determinants of graded structure in categories. Journal of Experi-mental Psychology: Learning, Memory, and Cognition, 11, 629–654.

Bettman, J. R. (1979). An information processing theory of consumer choice.Reading, MA: Addison Wesley.

Bettman, J. R., Johnson, E. J., & Payne, J. W. (1991). Consumer decision mak-ing. In T. Robertson & H. Kassarjian (Eds.), Handbook of consumer behav-iour (pp. 50–84). Englewood Cliffs, NJ: Prentice-Hall.

Bettman, J. R., & Sujan, M. (1987). Effects of framing on evaluation of compa-rable and noncomparable alternatives by expert and novice consumers.Journal of Consumer Research, 14, 141–154.

Carver, C. S., & Scheier, M. F. (1982). Control theory: A useful conceptualframework for personality, social, clinical and health psychology. Psycholog-ical Bulletin, 92, 111–135.

Carver, C. S., & Scheier, M. F. (1990). Origins and functions of positive andnegative affect: A control-process view. Psychological Review, 97, 19–35.

Engel, J. F., Blackwell, R. D., & Miniard, P. W. (1993). Consumer behavior (7thed.). Ft. Worth, TX: Dryden Press.

Fiske, S. T., & Taylor, S. F. (1991). Social cognition (2nd ed.). New York: Mc-Graw-Hill.

Ford, D. H. (1987). Humans as self-constructing living systems: A developmen-tal perspective on behavior and personality. Hillsdale, NJ: Erlbaum.

Gutman, J. (1982). A means–end chain model based on consumer categoriza-tion processes. Journal of Marketing, 46, 60–72.

Huffman, C., Ratneshwar, S., & Mick, D. G. (1997). An integrative frameworkof consumer goals: Structure, goal determination processes, and applica-tions. Unpublished manuscript, University of Pennsylvania, Philadelphia.

Johnson, M. D. (1984). Consumer choice strategies for comparing noncompa-rable alternatives. Journal of Consumer Research, 11, 741–753.

Klenosky, D. B., & Rethans, A. J. (1988). The formation of consumer choicesets: A longitudinal investigation at the product class level. In M. Houston(Ed.), Advances in consumer research (Vol. 15, pp. 13–18). Provo, UT: Asso-ciation for Consumer Research.

Locke, E. A., & Latham, G. P. (1990). A theory of goal-setting and task per-formance. Englewood Cliffs, NJ: Prentice-Hall.

Narayana, C. L., & Markin, R. J. (1975). Consumer behavior and product per-formance: An alternative conception. Journal of Marketing, 39, 1–6.

Park, C. W., & Smith, D. C. (1989). Product-level choice: A top-down or bottom-up process? Journal of Consumer Research, 16, 289–299.

Peter, J. P., & Olson, J. C. (1996). Consumer behavior: Marketing strategy per-spectives (4th ed.). Chicago: Irwin.

Ratneshwar, S., & Shocker, A. D. (1988). The application of prototypes and cat-egorization theory in marketing: Some problems and alternative perspec-

Page 23: Consumer decision making within a goal-driven framework

tives. In M. Houston (Ed.), Advances in consumer research (Vol. 15, pp.280–285). Provo, UT: Association for Consumer Research.

Ratneshwar, S., & Shocker, A. D. (1991). Substitution in use and role of usagecontext in product category structures. Journal of Marketing Research, 28,281–295.

Wright, P. L. (1973). Use of consumer judgement models in promotion andplanning. Journal of Marketing, 37, 27–33.

This study is based on a doctoral dissertation submitted to the Katz GraduateSchool of Business of the University of Pittsburgh. The author gratefully ac-knowledges the help and patient support given by Larry Feick, Rajiv Grover,Tim Heath, Dick Moreland, Dan Smith, and, especially, C.W. Park.

Correspondence concerning this article should be sent to: Robert Lawson, De-partment of Marketing and Management, William Paterson College, Wayne,NJ 07470 ([email protected]).

449GOAL-DRIVEN CONSUMER DECISION MAKING