Advances in Consumer Research Volume 12, 1985 Pages 465-470
TOWARDS A MORE COMPREHENSIVE THEORY OF CHOICE
Richard W. Olshavsky, Indiana University
A comprehensive theory of choice should explain the composition and priority of goals within a consumer's life goal as well as the specificity of these goals. Likewise, the relationship, between goal formation and behaviors relating to the acquisition, consumption, and disposition of goods and the relationship of all these behaviors to obtaining information should be explained. A theoretical framework is presented that encompasses all of these behaviors and their relationships.
Within the last decade or so consumer researchers (e.g., Bettman 1979; Kassarjian 1978; Krugman 1965; Olson 1977; Olshavsky and Granbois 1979) have recognized that choice is determined in a much more varied fashion than was suggested by the early models. Unfortunately, theory development has not kept pace with these empirical findings. While there have been some efforts to construct theories capable of explaining some of these variations (e.g., Bettman 1979, 1982; Payne 1982), none have been comprehensive. Nor have any of these theories done an adequate job of explaining the relationship between preference and intention and the relationship between intention and acquisition. Further, the relationships between acquisition and the consumption/disposition of goods have largely been ignored (Jacoby 1976).
Much more theoretical work also needs to be done to clarify the relationship between behaviors directed toward goods and behaviors directed toward information (acknowledging that in some cases information is a good). The "low involvement" concept in particular has stimulated efforts to modify existing models of decision making (DM) (e.g., Engel and Blackwell 1982) or to modify Lavidge and Steiner's (1961) hierarchy-of-effects model (e.g., Ray 1973; Smith and Swinyard 1982) to accommodate this type of behavior; unfortunately, this concept has also generated a great deal of confusion.
The purpose of this paper is to present a new theoretical framework that encompasses these and several other important issues related to choice. Preference formation (PF) will be addressed first. Intention formation (IF) will be treated next. The manner in which goals are added/deleted to the life goal and prioritized will be discussed. Acquisition, consumption, and disposition of goods will be related to intentions. The goal-subgoal relation between these behaviors and the "obtaining" of information will be described. Finally some advantages of this framework will be briefly described.
THE THEORETICAL FRAMEWORK
The framework to be presented represents an adaptation and extension of Newell and Simon's (1972) information processing theory of human cognition. Very briefly, the important assumptions, elements, and concepts of the theory are as follows: (For a more detailed exposition, see Olshavsky 1975.)
Two assumptions are fundamental to the theory. 1) Man is an adaptive system. This means that man's behavior is shaped by an interaction, over time, between man and the task environment (TE). 2) Man is an information processing system (IPS). It is assumed that man has the ability to interpret and execute a small, but powerful set of elementary information processes (eip's), such as "compare two symbols," and "find the next symbol on a list.
One of the important concepts in Newell and Simon's theory is the "program." A program consists of an organized sequence of eip's. Man's behavior is guided by this program. It is further assumed that the appropriate program (production system) is evoked by the nature of the goal and the TE in which man finds himself pursuing his goal. Given the right program, man is capable of some very sophisticated behaviors (e.g., playing chess, proving logic theorems). Another important concept is the "problem space;" it is man's internal representation of the external task environment that is determinant.
Newell and Simon's theory can be extended and adapted to consumer choice by the crucial assumption that a consumer (here defined as a household) possesses a generalized PF "schema" (GPFS) (Abelson 1981) that guides the selection of a particular PF strategy on any particular PF occasion. (Due to space limits, intra-household behaviors, i.e., family DM, cannot be addressed; see Olshavsky and King (in press).) It must be emphasized that the GPFS (see Figure) is not a model of how choice itself occurs. Once selected, that strategy guides the behavior of the consumer, but it is the interaction of consumer and TE characteristics, over time, that determines behavior. For a more detailed description of the GPFS, see Olshavsky (1984).
THE GENERALIZED PREFERENCE FORMATION SCHEMA
Is Preference Formation Necessary? For many reasons, no PF strategy of any kind may be required on some PF occasions. For example, prior PF may have established a preference for a good (e.g., affect referral). In this case, the preferred alternative is the pre-existing preferred alternative. If PF is necessary, then the consumer has two fundamentally different types of PF strategies from which to choose: own-based or other-based.
Own-Based PF Strategies. The most prominent own-based PF strategy (in the literature) Is decision making. Several alternative decision making rules - lexicographic, conjunctive, additive (including expectancy-value models), additive difference, etc. - are assumed to be available to the typical consumer (Bettman 1979; Svenson 1979).
Once a particular rule is selected, the consumer attempts to execute that rule. The behavior of the consumer, at both the outcome and process level, is then determined by the interaction that occurs over time between the consumer and the TE, as that TE is internally represented by the consumer, i.e., the problem or choice space (evoked set and brand/store image are relevant here).
Another own-based PF strategy is the use of a cue(s) (e.g., price, country of origin) as an index of quality, if quality is the basis for preference (Olson 1977). The outcome of this PF strategy is determined by the value of each alternative on the cue(s) utilized.
Other cognitive processes such as judgment, concept identification, learning, and reasoning may also serve as the basis for own-based PF. For instance, judgment is implied whenever only one alternative exists. More research is required on these behaviors however.
Other-Based PF Strategies. Other-based PF strategies refer to those choice behaviors in which the consumer subcontracts the PF to another consumer or organization. The clearest example of this type of PF strategy is "following a recommendation" (Formisano, Olshavsky and Tapp 1982; Olshavsky and Rosen (in press).)
This PF strategy is qualitatively different from that usually studied under the rubric of "opinion leadership," "word of mouth," or "interpersonal influences" because the assumption is made in that literature (at least implicitly) that the consumer obtains information from others for use in an own-based PF strategy (see e.g., Engel and Kollat 1982).
Combination Strategies. Two or more PF strategies may be selected and utilized on any particular occasion. An illustrative example of a combined strategy is a "phased strategy;" here two or more decision making rules are executed in sequence, e.g., a lexicographic rule followed by an additive rule (Olshavsky 1979).
Affectional Processes in PF. It is assumed here that emotional responses (affect), of various types and degrees, are activated (simultaneously) by PF and other behaviors. For example, finding a brand that meets all of a consumer's evaluative criteria may evoke joy. This emotional response is assumed to interact in important ways with the purely cognitive based PF processes and outcomes; in some cases, preferences may be based exclusively on affect (Zajonc 1982).
Obtaining Information. (This term refers to several behaviors in addition to acquiring information as will be mate clear below.) As a precondition to the execution of any PF strategy the consumer must acquire information from internal ant/or external sources. Obtaining information therefore is a subgoal to the goal of PF (ant other behaviors to be described). The specific type of information acquired, sources contacted, and the manner in which the information is processed depend greatly upon the particular PF strategy evoked. Behaviors directed expressly toward obtaining information will be described in greater detail below.
Determinants of PF Strategy. If this framework is ever to have the status of a theory (Hunt 1983), the "if-then" relationships that determine which strategy or combination of strategies is evoked on any particular PF occasion must be specified. The following contingencies between strategy and the basic elements are merely illustrative. (See Olshavsky (1984) for a more detailed presentation.)
Consumer characteristics can influence the PF strategy evoked in several ways. For example, the relatively small capacity of short term memory has been found by several researchers to be a determinant of the strategy evoked in a variety of cognitive tasks, including consumer choice (Bettman 1979). (However, Olshavsky and Acito (1984) demonstrate that an external memory aid may offset this constraint.)
The TE can be subcategorized into three types of environments - the marketplace (MP), the social environment (SE), and the physical environment (PE). The MP refers to the goods available at a particular point in time. MPs are characterizable in terms such as the number of alternative brands/stores, their similarly, the nature of the alternatives (e.g., complexity, price), and the nature of information available (Nelson 1970). Any one or combination of these MP factors can influence the PF strategy selected by a consumer. For instance, if a large number of alternatives exist then a phased DM strategy may be used (Olshavsky 1979).
The SE refers to the groups, formal and informal, that exist within a particular society. Groups are characterizable in terms such as their size, objectives, strategies, and tactics. One or more of these groups may influence the PF strategy. For instance, if a business firm uses comparative ads, the consumer may use a DM strategy rather than affect referral (Bettman 1979; Wright 1976).
The PE refers to all remaining aspects of the TE, such as the structural and climatic features of a particular locale. One or more of these aspects can influence the PF strategy. For instance, an impassable highway or inclement weather may deter a consumer from utilizing a DM strategy that requires external search.
To explain the relationship between preference and intentions it is first necessary to describe in a little more detail the nature of consumer goals and the behaviors related to goal formation and change. (Goods are but one type of goal.)
A comprehensive theory of choice should explain: 1) how particular goals (or bundles of goals) are added to or deleted from a consumer's overall goal in life (i.e., life goal), and 2) how a priority among goals (or bundles of goals) included within the life goal is established and changed. PF is viewed as part of this more general process of goal formation (GF); but PF only increases the specificity of a goal (e.g., which brand/store).
Behaviors involving additions to a life goal relate to those behaviors traditionally studied in the "adoption of innovations" literature (Robertson 1971; Rogers 1962). And, changes in the priority of goals relate to research and theory on values (Rokeach 1968). Changes in both composition and priority of goals relate to the literature on family life cycle.
These two additional types of GF behaviors can be explained in a manner similar to that described for PF. A schema consisting of alternative strategies is assumed to exist and the same type of contingent relationship between strategy and the state of the basic elements is assumed.
Financial, temporal and other considerations typically preclude a consumer from acquiring all goals included in the life goal. A comprehensive theory of consumer choice should be able to explain how the specific goals (involving goods) that are to be sought within the next (or subsequent) purchase period are determined.
The composition, priority and preferences contained within a consumer's life goal, the price of goods (or expected price), and a consumer's disposable income (or expected income) and disposable time are among the factors assumed to play an important role in determining the outcome of the IF process. Traditional economic theory describes how consumers attempt to achieve all desired goods, in the desired amounts, subject to the budget constraint. But this is only one of several alternative IF strategies (Olshavsky and Granbois 1979). Research involving "budgeting and planning" (Ferber 1973) and purchase priorities are directly related to IF.
To explain IF, again a schema is postulated to exist that consists of several strategies and the particular IF strategy evoked on any particular occasion depends upon the state of the basic elements.
ACQUISITION, CONSUMPTION, AND DISPOSITION OF GOODS
An intention does not always result in an acquisition (purchase, choice). The acquisition of goods involves behaviors such as transporting the consumer to the marketplace, locating goods in the marketplace, exchanging money for goods, and transporting goods to the desired location (e.g., the consumer's home). Difficulties encountered during any one of these behaviors (e.g., inability to locate a store) can result in failure to acquire the good. A comprehensive theory of choice should be able to explain acquisition and its relation to other choice behaviors
Consumption refers to behaviors such as storing, preparing, and using goods. Consumption patterns also pertain to usage rates, timing of purchases, frequency of repurchase and certain unintended outcomes (e.g., automobile accidents). A comprehensive theory of choice should be able to explain consumption and its relation to other choice behaviors.
Some goods y are disposed of before another good of the same type can be acquired (e.g., houses, cars). Disposition of goods can take many different forms such as trashing, trading, and selling (Jacoby, Berning and Dietvorst 1977). Difficulties, actual or anticipated, encountered in the disposition of a good can have important implications for first time or repeat purchase. A comprehensive theory of choice should explain disposition and its relation to other choice behaviors.
The framework being presented here can be expanded to explain acquisition, consumption, and disposition by assuming that each of these behaviors is directed towards the achievement of a distinct goal and by assuming that associated with each goal is a separate generalized schema. Each generalized schema, in turn, is assumed to consist of a variety of alternative strategies for perForming each of these behaviors and that the particular strategy evoked on any particular occasion depends upon the state of the basic elements. For instance, there are many ways of preparing foods, the specific strategy (recipe) evoked on any particular occasion depends Upon a large number of consumer and TE factors.
The relationships among goal formation, acquisition, consumption, and disposition can be explained by the postulation of yet another more generalized schema (an "episode") that specifies the sequence in which these four behaviors typically occur (i.e., GF -> A -> C -> D) while recognizing that exceptions to this sequence occur under certain conditions (e.g., in the case of sampling, acquisition occurs without intentions).
Obtaining information is a subgoal in goal formation, acquisitions consumption and disposition. In the interest of parsimony, the assumption is made that achieving this subgoal involves the same four types of behavior - goal formation, acquisition, consumption and disposition; but, at this level these behaviors pertain to obtaining information. (It is recognized that some types of information are goods.)
Goal Formation (Regarding Information)
Before a consumer can attempt to acquire information, the type and amount of information desired must be specified. Also, the priority of the different types of information desired must be specified. And preferences must be established among specific sources of information.
Further, not all information desired will typically be translated into intentions to achieve that information due to financial, temporal and other constraints; thus, s separate stage of IF with respect to information is assumed to occur.
Considerable research has been done on GF with respect to information, although not under this terminology. for instance, an extensive literature exists on the effects of the "credibility of the communicator." But source credibility can be viewed as an attribute of a Source and as an evaluative criterion used by a consumer in PF with respect to alternative sources. GF with respect to information can be explained in a manner similar to that described above for GF with respect to goods.
Acquisition (Regarding Information)
After intentions with respect to information have been Formed, the consumer attempts to acquire that information. The same basic types of behaviors are assumed to be involved here as with the acquisition of goods; namely, transporting, locating, and exchanging with due recognition of the unique characteristics of information (e.g., information can be "transported" over telephone lines). It is to be emphasized that information can be acquired without intentions, as in influence attempts.
Considerable research has been performed on information Acquisition behaviors involving both internal sources (i.e., retrieval from long term memory) and external Sources (see Bettman 1979). Acquisition behaviors can be explained in a manner similar to that described above for the acquisition of goods.
Consumption (Regarding Information)
Once information has been acquired, this information is assumed to be consumed (processed) in some fashion. Consumption of information may result in changes to preC existing beliefs, attitudes, skills, affect, etc., which, in turn, may lead to a change in preference. The consumer's perception of the TE (i.e., the choice space) is assumed to be determined in this fashion.
Considerable research has been performed on the "consumption" of information. Indeed, all research on persuasive communications, from either personal or nonpersonal sources, can be viewed as studies of information consumption. This research has shown that information is consumed in ways ranging from simple learning to extended elaboration (e.g., Craik and Lockhart 1972; McGuire 1976; Olson, Toy and Dover 1977; Petty, Cacioppo, and Schumann 1983; Wright 1973, 1980). Also, research involving use/trial of a good (Bem 1972; Oliver 1980; Smith and Swinyard 1983) relates to information consumption. For instance, the satisfaction or dissatisfaction experienced during consumption of a good is viewed here as the outcome of the process of consuming information in the form of direct experience. Research on socialization, compliance, reactance, and conformity is also relevant here.
The proposed framework can accommodate all types of information consumption-behaviors by postulating, as before, the existence of a set of alternative strategies for the consumption of information. A detailed specification of these information consumption strategies would include basic processes already identified in one or more of the traditional theories on persuasion such as learning theory, attitude theory, consistency theory, attribution theory, social judgment theory and cognitive response theory.
The specific strategy evoked on any particular occasion is again assumed to depend upon the state of the four basic elements. Recent theoretical developments in the area of persuasive communications all seem to be headed in the direction of a similar, contingent view of information processing (Petty et al. 1983; McGuire 1978; Krugman 1966; Ray 1973; Smith and Swinyard 1982).
Variables identified in research on persuasive communication as being important determinants of effectiveness can all be conveniently classified as belonging to one of the four basic elements. For instance, consumer characteristics encompass variables such as the initial strength of the beliefs of the consumer. MP characteristics encompass variables such as the expertise/ credibility of the communicator (e.g., salesperson). SE characteristics subsume variables such as the characteristics of the message (e.g., sideness, vividness, type of appeal). and PE characteristics encompass variables such as the nature of the context in which the consumer attempts to consume information (e.g., loud. distracting noise).
Disposition (Regarding Information)
Information, once acquired and consumed, may be disposed of in various ways. Information stored in long term memory may be forgotten due to the consumer's failure to rehearse this information periodically. And, past copies, of Consumer Reports may be discarded or passed along to friends.
A great deal of research has been performed on learning and retrieval processes in psychology, but little is known specifically about the manner in which consumers dispose of information related to goods. It is expected however that these behaviors can be explained in the same manner described above for the disposition of goods.
The theoretical framework being proposed here has several distinct advantages over existing theories of choice. First, it can encompass all known types of PF strategies. Given the pre-occupation choice theorists have had with attitude as the sole mediator of intention and subsequent acquisition, the present framework should encourage future researchers to view attitude as but one of several bases for the establishment of a preference (i.e., preferences are not the same as attitudes). An important implication of this framework is that when an other-based PF strategy is evoked, advertising or other persuasive communications efforts may not affect preference even though it may produce changes in beliefs, attitudes, or affect. For instance, a consumer may prefer the brand recommended by a friend regardless of the impact advertising has had on his/her beliefs and attitudes toward that brand and toward competing brands (Formisano, et al. 1982).
Second, this framework encompasses behaviors and relationships involving goal formation (including PF and IF), acquisition, consumption and disposition. No theory of choice addresses all of these important aspects of consumer behavior (Jacoby 1976). The many ways in which PF is assumed to occur and the separate IF and acquisition stages that are assumed to intervene between PE and purchase may help to explain the low correlations so frequently observed between attitude and behavior.
Third, this framework provides a better perspective on the impact of advertising on choice. In particular, the manner in which a consumer processes information from ads (or any other source) has been shown here to depend greatly upon the specific context in which the desire for information arises (i.e., in a goal-subgoal relationship). The hierarchy-of-effects model ignores this situational dependence. Past attempts to explain the relation between advertising and choice based on the hierarchy-of-effects model share this problem. In addition, such models fail to maintain a clear distinction between behaviors relating to goods and behaviors relating to information. These models also treat attitude as the sole basis for choice.
By maintaining a clear separation between behavior directed towards goods and behavior directed towards information and by stressing the goal-subgoal relationship between these behaviors, the present framework brings greater clarity to this and other areas. For example, the great confusion that surrounds the low-involvement concept can be reduced somewhat by describing, within the proposed framework, some of the many different types of behaviors identified with this ambiguous concept: l) The priority of the good, as a goal within the life goal, may be low. (This interpretation appears to be closest to Krugman's definition of the term.) Little time may be allocated for GF for goods that are relatively unimportant. 2) At the level of GF involving goods, a consumer may not engage in any PF activity because the alternatives are judged to be so similar that any type of PF is unwarranted (e.g , commodities). 3) the level of GF involving information, a consumer may decide that the best source of information for a particular good is trial (Smith and Swinyard 1983). (Hence, "conation" can occur before "affect".) 4) At the level of consumption of information, a consumer may select a strategy that involves little or no processing of an advertisement. (Limited processing of an at is particularly likely to occur for ads acquired without intentions, as in watching television.) 5) At the level of consumption of information, the consumer may have such weak beliefs and attitudes (i.e., low "ego involvement") that little processing of the message occurs beyond simple learning.
In short, the low involvement concept has lead to great confusion because it refers to so many types of behavior and/or states of the consumer. The proposed theoretical framework adds precision to the description of the behaviors of interest; it also promises to provide an explanation for all of these behaviors.
The final advantage to be highlighted is that the assumptions expressed in this framework about the contingent relationships between strategies and the basic elements and between the consumer and the TE can be expressed in great detail and formalized in the form of computer program, such that predictions about the choice process and outcome can be made (Newell and Simon 1972). For an illustration of how part of the DM branch of the GPFS has been formalized see Olshavsky and Acito (1982). Clearly, insufficient detail has been provided in this paper to formalize the theory; however the task is less formidable than it might at first seem since much more research has been done to specify the relationships involved than could possibly be reviewed in this short PaPer.
Abelson, R. P. (1981), "Psychological Status of the Script Concept," American Psychologist, 36, 715-729.
Bem, D. J. (1972), "Self-Perception Theory," in Advances in ExPerimental Social Psychology, L. Berkowitz (ed.), 6, New York: Academic Press, 1-62.
Bettman, James R. (1979), An Information Processing Theory of Consumer Choice, Reading Mass.: Addison-Wesley Publishing Co.
Bettman, James R. (1983), "A Functional Analysis of the Role of Overall Evaluations in Choice Processes," in Advances in Consumer Research, Andrew Mitchell (ed.), 9, 87-93.
Craik, Fergus and Robert S. Lockhart (1972), "Levels-of-Processing: A Framework for Memory Research," Journal of Verbal Learning and Verbal Behavior, 11 (October), 671-684.
Engel, James F. and Roger D. Blackwell (1982), Consumer Behavior, Chicago: The Dryden Press.
Ferber, Robert (1973), "Family Decision Making and Economic Behavior: A Review," in Family Economic Behavior: Problems and Prospects, Kleanor B. Sheldon (ed.). Philadelphia: Lippincott. 29-61.
Formisano, Roger A., Richard W. Olshavsky, and Shelley Tapp (1982), "Purchase Strategy in a Very Difficult Task Environment," Journal of Consumer Research, 8, 474-479.
Hunt, S. D. (1983), Marketing Theory -- The Philosophy of Marketing Science, Homewood, IL: Richard D. Irwin.
Jacoby, Jacob (1976), "Consumer Psychology: An Octenium," in Annual Review of Psychology, Paul Mussen and Mark Rosenzweig (eds.), 27, 331-353.
Jacoby, Jacob, Carol K. Berning and Thomas Dietvorst (1977), "What About Disposition?" Journal of Marketing, 41, 22-28.
Kassarjian, Harold H. (1978), "Presidential Address 1977: Anthropomorphism and Parsimony," in Advances in Consumer Research, Vol. 5, B. Keith Hunt, (ed.), xii-xiv.
Krugman, B. E. (1965), "The Impact of Television Advertising: Learning Without Involvement," Public Opinion Quarterly, 29 (No. 3), 349-356.
Lavidge, R. J. and G. A. Steiner (1961), "A model for Predictive Measurement of Advertising Effectiveness," Journal of Marketing, 25 (October), 59-62.
McGuire, William J. (1976), "Some Internal Psychological Factors Influencing Consumer Choice," Journal of Consumer Research, 2 (March), 302-319.
Nelson, Phillip (1970), "Information and Consumer Behavior," Journal of Political Economy, 78, 311-329.
Newell, Allen and Herbert A. Simon (1972), Human Problem Solving, Engelwood Cliffs, NJ: Prentice-Hall, Inc.
Oliver, Richard (1980), "A Cognitive model of the Antecedents and Consequences of Satisfaction Decisions," Journal of Marketing Research, 17, 460-469.
Olshavsky, Richard W. (1975), "Implications of an Information Processing Theory of Consumer Behavior," in Combined Proceedings, Marketing in Turbulent Times and Marketing: The Challenge and the Opportunities, Edward M. Mazze (ed.), 151-155.
Olshavsky, Richard W. (1979), "Task Complexity and Contingent Processing in Decision Making: A Replication and Extension," Organizational Behavior and Human Performance, 24, 300-316.
Olshavsky, Richard W. (1984), "Perceived Quality in Consumer Decision Making: An Integrated Theoretical Perspective," in Consumer Perception of Merchandise and Store Quality, Jacob Jacoby and Jerry Olson (eds.), Lexington, MA: Lexington Books.
Olshavsky, Richard W. and Frank Acito (1980), "An Information Processing Probe into Conjoint Analysis," Decision Sciences, 11, 451-470.
Olshavsky, Richard W. and Frank Acito (1984), "The Effect of Memory Aids on Choice Strategy," in Developments in Marketing Science, Jay D. Lindquist (ed.), 7, 90-94.
Olshavsky, Richard W. and Donald B. Granbois (1979), "Consumer Decision Making - Fact or Fiction?" Journal of Consumer Research, 6, 93-100.
Olshavsky, Richard W. and Maryon King (in press), "The Role of Children in Household Decision Making: Application of a New Taxonomy of Family Role Structure," in Marketing to the Changing Household: Management and Research Implications, Ballinger Publishing Co.
Olshavsky, Richard W. and Dennis Rosen (in press), "Use of Product-Testing Organizations' Recommendations as a Strategy for Choice Simplification," Journal of Consumer Affairs.
Olson, Jerry G. (1977), "Price as an Information Cue: Effects in Product Evaluations," in Consumer and Industrial Buying Behavior, Arch G. Woodside, Jagdish N. Sheth, and Peter D. Bennett (eds.), NY: North-Holland. 267-286.
Olson, Jerry G., D. R. Toy, and P. A. Dover (1977), "Do Cognitive Responses Mediate the Effects of Advertising on Cognitive Structure," Journal of Consumer Research, 9 (December), 245-262.
Payne, John W. (1982), "Contingent Decision Behavior," Psychological Bulletin, 92 (No. 2), 382-402.
Petty, R. E., J. R. Cacioppo, and D. Schumann (1983), "Central and Peripheral Routes to Advertising Effectiveness: The Moderating Role of Involvement," Journal of Consumer Research, 10 (September), 135-147.
Ray, M. L. (1973) "Marketing Communications and the Hierarchy of Effects," in New Models for Mass Communication Research, P. Clarke (ed.), Sage Publishing.