The Development of the Hierarchy of Effects:
An Historical Perspective
Thomas E. Barry
Just what constitutes advertising effectiveness is often a topic of heated
debate. "Bottom-liners" contend that advertising is effective only when it sells.
Communication advocates, however, suggest that there is a series of stages
between the point of unawareness of a product and/or brand and the ultimate
purchase/sale of a particular brand. Proponents in each of these two camps
rarely agree on a middle road. The body of work in the marketing and advertis-
ing literature that relates to this advertising effectiveness controversy is called
the hierarchy of effects and has been accorded theoretical status by many
advertising and marketing practitioners and academics. However, as Tauber
[1982--83] noted, there are several different classes of advertising them-y and, in ""
light of this fact, researchers really must,
. . . take a step back to review theories about how advertising works. If we
had a thorough review, we might see that the appropriate research meas-
ures flowed from the objectives of each advertising situation (p. 7).
On the other hand, Ramond [1976] has argued that advertising has no general
theory that is widely accepted but rather the field is a collection of
"pseudo-theories" whose genesis is introspection. He argues that the more
mysterious and complex the field, the more systematic must be the approach to
understanding that body of work. The purpose of this article is to provide the
reader with a clear perspective of how the hierarchy of effects literature has
developed over time. This historical perspective will provide researchers and
practitioners with a thorough glimpse of this most important body of work and
The author appreciates and wishes to acknowledge the financial support of the Research and
Development Council and the Center for Marketing Management Studies of th(> Edwin L. Cox
School of Business at SM U as well as the constructive comments of Bonnie Blytilc and Professors
Mary GilIv, Bill Havlena, and Dan Eloward.
251
-
1922 Edward K. Strong, Jr., The
Psychology of Selling Life
Insurance
1923 Daniel Starch, Principles of
Advertising
1925 Edward K. Strong, Jr.,
"Theories of Selling, " The
Journal of Applied Psychology
1938 Edward K. Strong, Jr.,
Psychological Aspect of
Business
1940 Clyde Bedell, How To Write
Advertising That Sells
1956 Merrill, DeVoe, Effective
Advertising Copy
Referred to the popularity of Sheldon's AIDAS but suggested
outline was faulty; aim of advertising tcrget immediate
action all the time
SRBRA-Seen, Remembered, Believed, Read, Acted LTpon
Reviewed the theories to date; 3 main theories: AIDAS,
Situation-Response, and Appeal-Response (for the adver-
tising, not selling, world)
Refers to N. Hawkins' three different stages of attention
and interest; a sales prospect must be taken through these
six stages (3 each) before desire can occur; Synthesizes the
three theories of selling discussed in 1925 piece; one of
those is AID(W) C(S) PS--Attention, Interest, Desire (Want),
Conviction (Solution), Purchase, Satisfaction
Refers to "Proved Selling Stratagems, " of AIDCA--
Attention, Interest, Desire, Conviction, Action
Refers to different sequences in constructing ads; the
psychological sequence is popular and has been around for a long
time, e.g., AIDCA--Attention, Interest, Desire, Conviction
and Action; and AIDMA--Attention, Interest, Desire,
Memory, Action; no references to these models
Nw
I L
,. ..,., ,
,$$
YEAR
1961
1961
1961
1962
1962
1962
1964
1967
hlODERN DEVELOP\lENT PHASE
AUTHOR(S) FORhl/DESCRIPTION OF N1ODEL
Robert J. Lavidge and Gary
A. Steiner, "A i'vlodel for
Predictive Measurements of
Advertising Effectiveness, "
Journal of Marketing
Russell H. Coney, DAGhlAR
--Defining Advertising Goals
for Measured Advertising
Results
Advertising Research
Foundation
Harry D. \Volfe, James K.
Bro\vn and G. Clark Thompson
?vIeasuring Advertising Results
Everett M. Rogers, Diffusion
of Innovation
Harold Menclelsohn, "Measuring
the Process of Communications
ERect, " Public Opinion
Quarterly
Leo V. Aspinwall, "Consunler
Acceptance Theory"
Sanclage and Fryburger
HIERARCHY OF EFFECTS--Awareness, Knowledge, Liking,
Preference, Conviction, Purchase
ACCA--.Awareness, Comprehension, Conviction, Action
EPC(K)C(A)A--Exposure, Perception, Colnmunication
(Knowledge), Communication (Attitude), Action
AAPIS--Awareness, Acceptance, Preference, Intention,
Provocation of Sale
AIETA--Awareness, Interest, Evaluation, Trial, Adoption;
relates to new product adoption process
Looks at three types of responses: rudimentary response
(feeling), and active response (learning to aicl decision
making); developed an Active Response Scale similar to
Cuttnl& Sc:de -
API--.Acceptance,
EPIA--Exposure,
Preference, Insistence
perception, Integration, Action
.-.-.
1969 David A. Schwartz, "Measuring EARACP-Exposure, Attention, Retentioni
the Effectiveness of Your Attitude Change, Purchase
Company's Advertising, "
Journal of Advertising
1969 John Howard and jagdish Sheth, AC.41P--Attention, Comprehension, Attitude, Intention,
The Theory of Buyer Behavior Purchase
1969 William J. McGuire, "An PACYRB--Preselltatioll, Attention, Comprehension, Yielding
Information - Processing Retention, Behavior
hfodel of Advertising
Effectiveness"
19'71 Thomas S. Robertson, ACALTA--Awarelless, Comprehension, Attitude, Legitimation,
Innovative Behavior and Trial, Adoption
Communication
1971 Kenneth A. Longman, EAPCBMA-Ex~30sure, Attention, Perception, Comprehension,
Advertising Belief, Motivation, Action
1974 Andrew S. C. Ehrenberg, ATR--.Aw~arel~ess, Trial, Reinforcelllent hlodel; hierarchy
"Repetitive Advertising ant] models don't tell LIS enough; advertising's role is to
the Consumer, " Journal of facilitate trial purchase among the consumer's repertoire of
Advertising Research brands
1975 hlorris B. Holbrook, "A APMAI-Attel~tiol~, Perception, hlemory, Attitude, Intention
Review of Advertising
Research"
.
YEAR
1973
1979
1980
1981
1992
CHALLENGES AND DEFENSES
AUTHOR(S) FORN1/DESCRIPTION
Michael L. Ray, et al.
Marketing Communications and the
Hierarchy of Effects"
Robert L. Anderson and Thomas E.
Barry, Advertising Management:
Text and Cases
Richard Vaughn, "How
Advertising Works: A
Planning Model," Journal of
Advertising Research
Nlichael L. Rothschild and
William C. Gaidis, "Behavioral
Learning Theory: Its Relevance
to Marketing and Promotions, "
Journal of hlarketing
Robert E. Smith and William R.
Slvinyard "Infornlation Response
Models An Integrated Approacl~, "
Journal of Nfarketing
There may be a multiple of hierarchy models operating
under different involvement situations
Added
Stated
model
added
Brand Loyalty to the hierarchy of effects models
that "thinking lmodel"--the traditional hierarchy
of cognition, afl'ect, conation-not adequate;
three additional models in different sequencing,-.
e.g. affect-cognition-conation; conation-cognltlon -anec[;
conation-:lffect-cognition
For low involvement purchases; .r~dvertising is stimulus for
awareness and knowledge; leads to trial; product becomes
stimulus and reinforces and satisfies leading to in-
creased probability of repeat purchasing behavior
The learning hierarchy may not be ap~xopriate for low
in~olvemei~t purchase situations; look at high and lo\v
order belief and affect; suggest three lnoclels: tradi-
tional of cognition-tiect-conlmitnlellt; low involvement of
cognition -trial-:lffect-conlmitment; and brand switching of
cognition-trial-trial-trial
l
lc)h~
1983
1983
1984
1986
I\an L. Preston, "TheAsso-
ciation \fodel of the Aclvertis-
ing Communication Process, "
Journal of Advertising
Ivan L. Preston and Esther
Thorson, "Challenges to the
Use of Hierarchy hlodels in
Predicting Advertising Effec-
tiveness, >
' AAA Proceedings
Sandra Ernst Nloriarty, "Beyond
the Hierarchy of Effects: .4
Conceptual Fran~e\vork," Current
Issues and Research in
Advertising
Ivan L. Preston and Esther
Thorson, "The Expancled Asso-
tion Llodel: Keeping the Hier-
Hierarchy Concept Alive, "
Journal of Advertising Research
Richard Vauglm `<
How' Advertising
~Vorks: A Pl&ning hlodel Revis- "
ited, " Journal of Advertising
Research
g}d ...
Presents more comprehensive consumer information process-
ing model stating lack of this is fault of previous
models: Distribution, Vehicle Exposure, Ad Exposure, Ad
Awareness, Ad Elements Awareness, Association Evaluation,
Product Perception, Integrate] Perception, Product Evalua-
tion, Prior Evaluation, hitegrated Evaluation, Product
Stimulation, Prior Stimulation, Integrated Stimulation,
Action; states traditional hierarchy valid in spite of
low involvement theory
Defend traditional hierarchy models and add to the Asso-
ciation h loclel of 1982 \vitl]: Search, Search Perception,
Search E\ 'aluation, Search Stimulation, Trial, Trial Per-
ception, Trial Stimulation, Adoption, Adoption Perception,
Adoption Evaluation and Adoption Stimulation
Reviews ancl challenges traditional hierarchy models; pre-
sents a Continuum Domain hlodel with the domains being per-
ception (no a~vareness to recall), education (learning to
generalization and cliscrimination), persuasion (reinforce-
old attitudes to changing old ones) and behavior (inquiry
to repurchase)
In effect combines 1982 and 1983 articles and presents
the complete, expanded Association hlodel including the
addition of the three Action Steps (Search, Trial,
Adoption)
Continues contention in earlier article that there are
multiple hierarchies used in responding to advertising
messages
tow02
260 R(?c)ictm Of :.'l'lcctd AI-CUS
Among the more behaviorally oriented works was Adams' [1916] book about
the mental laws of advertising. Adams generally supported the traditional
hierarchy work of Lewis and Sheldon, noting that in order to produce effective
advertising an advertisement must lead to some action by arresting and holding
attention, creating a favorable impression, and having it remembered, Adams
focused on the "sensation complexes" of the audience-their receiving equip-
ment such as end and sense organs. According to Adams, these sensation
complexes are modified by and grouped with memories to form perceptions.
To Adams, attention was the basis of all voluntary actions, and incoming
stimuli, such as advertisements, had to link up with other ideas associated with
past experiences or hereditary responses. Associations could then be made and
tested by the "laws of recall. " These laws included communication variables
such as contiguity, similarity and contrast, primacy and reccncy, fl"cqucncy and
intensity, as well as purposes, attitudes, education, social pressures, and
heredity. Perception and knowledge were "fusions of sensations," and learning
was the simple formation of associations that were fused together. It was impor-
tant that advertisers understand the laws of learning including repetition, series
length, rhythm, will to learn, and others. Adams's notion of "dynamogenesis"
was the most fundamental law of action declaring that any sensation or idea will
result in some kind of movement, be it automatic, reflexive, habitual, or
random. Adams was one of the original "bottom-liners" in that advertising was
only effective, in his mind, when it led an audience to action.
Strong [1922] was also a "bottom-liner" and suggested in a book on the
psychology of selling life insurance that Sheldon>s [1911] five-step hierarchy
was faulty and had done as much harm as good to the science of selling. It was
Strong's belief that it was both unnecessary and a waste of time to lead a
prospect through the five hierarchical stages, stating, " . . , while the outline is
sometimes true, the ideal is to get action immediately" (p. 155). In a later work,
Strong [1925b] referred to three general theories of selling: Sheldon's AIDAS,
stimulation-response (or appeals-response in advertising), and man as a dynam-
ic being seeking to satisfi wants. According to Strong, the five mental states
(AIDAS) were not the goal the seller should strive for; he believed that this
formulation was likely responsible for many poor advertisements and sales
presentations.
Poffenbcrger [1925] contended that it would be too diffwult to try to analyze
a(.lvcrtising behavioral responses through some hierarchical format stating that
the "processes involved in the human reaction to an advertisement are too
complex and 100 much interwoven with each other to permit [oH analysis into a
chain of distinct and loosely connected operations, each to be examined in
isolation from all the others (p. 566). Rather, it was the association of wants and
needs that provoked responses from advertisements. And it was only where
wants wwre ~~wlk or where consumers faced a large variety of means to satisfy
their wants when " . . . all the devices of the sciences and the arts need to be
I
The IIicrarch~g ~f Effects 261
called upon to make an appeal effective" (p, 566), For example, Poffenberger
wondered how 38 brands of cigarettes and 30 brands of toothpaste could differ
in their power to satis& wants. These situations would call for the skillful use of
advertising and, through association, could effectively generate reactions from
consumers.
Jenkins [1935] maintained, like Strong and others, that the primary aim of
advertising was to influence immediate action or "goodwill actjon" over some
period of time and that, in order to develop effective advertising, the various
stages of action should be understood. All other aims of advertising should be
focused on making the total advertisement an "effective governor of action" (p.
259). In later research, Strong [1938] presented the work of Hawkins, a Sales
Manager for General Motors, who proposed that there were three stages of
attention and three stages of interest that a selling prospect had to be carried
through before desire could be reached. Attention was compulsory, curious, or
spontaneous; interest was attentive, associating, or personal. The main objec-
tive of Strong's later work was to synthesize the three theories of selling
(AIDAS, appeals-response, and the dynamic man theory) mentioned earlier.
His synthesis scheme is presented in Figure 1. Part (a) of Figure 1 conveys
Strong's synthesis of AIDAS and the dynamic man theory; Part (b) synthesizes
the appeals-response theory with the dynamic man theory.
The attention, interest, desire, conviction, action sequence was heavily re-
lied upon by Bedell in his text, How to Write Advertising that Sells [1940]. For
each phase of the hierarchy, Bedell presented five or six "selling
strategems''--proved selling devices--that would create copy that would lead -
ad readers from the attention phase to the buy phase.
In a final advertising work in this early phase, Devoe [1956] referred to two
popular psychological sequences: attention, interest, desire, conviction and
action; and attention, interest, desire, memory, and action. He criticized these
"models" for not differentiating attention to the product versus attention to the
advertisement. He proposed that there is also a difference in the interest
toward the product as well as toward the ad itself.
These early models were descriptive representations based on intuition and
logic. There was virtually no empirical analysis of any of them. Some were more
psychology-bounded than others, but it is evident that these models provided
the foundation for the discussions of the hierarchy models developed over the
next thirty years.
The Modern Development Phase
In contrast to many of their early development predecessors, Lavidge and
Steiner [1961] felt that immediate sales was an incomplete criterion of advertis-
ing effectiveness, even if that criterion were measurable. They felt that adver-
tising was a long-term investment, principally because many of advertising's
Jvallt
1,
2.
(l))
3,
4.
Appeals
Incitcn]ents arousil~g wants of 1.
prospect
Explmmtions and clc~]]li)]]st]-iltio]~s 2.
tlult tl~e comn~odity will adequately
silti!jfy tll~ W}lllts
Induccmcnts to l)UY 1101%" 3.
CUarL\llt~~S tllilt tllc solution will l)C 4.
Response
Wimt
S o l u t i o n -
s. (hmmodity
b. Trade Name
Purchase
Siltistil~tiOll
satisfactory for a reasonal)le length
of tillle
l'rillt['d with pcrl~lission of l~qx'r & Row Plll)lishcrs, inc. FromStrong, lZdwardK.Jr.
(1938) P4s!/c/dogica/ Aspects of Business. NCW York: ?vl&ri~w-I-lill Book Cmnpany, 11~~.,
21(). After I,llcas, D. B. And Benson, C. E. (19{30) Ps@tologq f~)r Advertisers. New York:
Iidq)cr & Bros.
ligum 1, Strong's Synthesis of Thmc Response hlodds
effects were long-term in
consumers do not jump
purchasers; rather, they
purchase.
The seven steps that
threshold were:
The ]Iierarchy of Effects
nature, In their work, Lavidge and Steiner felt
263
t h a t
from a stage of disinterest to that of convinced
go through a series of steps to that threshold of
consumers followed enroute to the purchasing
1. unawareness of product or service existence
2. awareness of product or service existence
3. knowledge of what the product has to offer
4. favorable attitude toward the product-consumers like it
5. a favorable attitude to the point of preferring one brand over others
6. a desire to buy as well as conviction that the purchase would be wise
7. the actual purchase,
Lavidge and Steiner held that these steps were not necessarily equidistant from
each other and that consumers could move "up" several steps simultaneously.
The consumer's psychological or economic commitment would have an impor-
tant bearing on his or her succession to the top (purchase) of the ladder with
more committed purchases taking longer to go through the process. It was also
important to note that the "simple model, " according to these researchers,
assumed that all consumers started from "scratch." In reality, however, they
felt that those consumers who had developed negative attitudes had to be
removed from the negative steps and then moved up the lad&r through the'
process. It was at this stage in the development of the hierarchy of effects work
that the popular terms of "cognition, " "affect, " and "conation" were first used in
reference to the actual hierarchy of advertising responses. Cognition referred
to mental or rational states, afFect to feeling or emotional states, and conation to
the striving or behavioral states.
At the same time of Lavidge and Steiner's work, Coney [1961] developed his
awareness, comprehension, conviction, and action hierarchy in his well-known
work referrecl to as DAGMAR--Defining Advertising Goals for Measured
Advertising Results. Coney reasoned that most advertising objectives were too
vague ancl not easily measured. He suggested that a hierarchy of communica-
tions objectives existed that led to ultimate purchase, These communication
objectives could be more easily measured, compared to direct sales results, to
determine whether the impact of advertising on an audience was effective or
not. The interest in the hierarchy notion was further fueled by the Advertising
Research Foundation's [1961] hierarchy model of exposure, perception, com-
munication (knowledge), communication (attitude), and action.
Wolfe, Brown, and Thompson [1962] published the results of a two-year
Conference Board study that sought to determine how to measure the effects of
2(M Rcviucs of Sdected Areas
advertising. They concluded that it was not reasonable to use sales as a measure
of advertising results unless advertising was the dominant selling force and
other marketing factors remained constant. According to the study, sales result
from advertising via a five-stage process beginning with product awareness,
nurturing product acceptance, establishing product preference, arousing an
intent to buy, and, finally, provoking the sale. The specific influence of adver-
tising at each of these stages was dependent on the industry, the product, the
type of advertising clone, the prevailing market position, and the selling
methods of the company as well as other company factors. The influence of
advertising, according to the study, diminished along each of the successive
stages of the advertising or sales process.
About the same time as the above Conference Board study, Rogers [1962]
developed a response hierarchy model for new product introductions. Referred
to as the adoption model, Rogers proposed that consumers go through a
bier-archy of awareness, interest, evaluation, trial, and adoption. This adoption
process model was expanded upon by Robertson [1971] as awareness,
comprehension, attitude, legitimation, trial, and adoption.
Mendelssohn [1962] was unique in the mass communications branch of social
psychology in his study of the more generic process of communication effects
contending that the Hull learning theory approach as well as the Gestalt
sociological approach were not much help in aiding researchers in understand-
ing the process that people go through in responding to communications. The
former approach was too "gross" and "monolithic," while the latter was too
"multi-dimensional" to be useful for developing theories of communication
effect.
Mendelssohn proposed that communications can affect an individual's actions
directly; that effect is a broad psychological process rather than simply a learn-
ing process; effect is not equally distributed among an audience's recipients;
and that the effect process is cumulative within an individual going from a
"rudimentary psychological brush" with the particular communication to
actually being activated by it. He stressed that effect did not result from simple
learning but was accompanied by emotion. This led Mendelssohn to an early
hypothesis of involvement suggesting " . . . a totality of psychological experi-
ence with a communication in terms of learning, feeling, and preparing oneself
to act as a consequence of exposure. The greater the involvement that the
communication produces, then the more effective it can be considered" (p.
413).
Mendelsohn's response stages were congruent with the cognition, affect, and
conation stages suggested by Lavidge and Steiner. Rudimentary response was
simple recall ability, Emotional response involved recall coupled with some
Locm tmnal reaction, as opposed to indifference, toward the communication.
Active response entailed recall and emotion as well as learning that the object
%1
o
of tllc communication was worth recommending and following up on.
The IIierarch~ of Effects 265
The hierarchy models of Lavidge and Steiner, Wolfe, Brown and Thompson,
and Coney were reviewed by PaIda [1966], palda suggested that these models
were "sketchy views" of the internal psychological process, PaIda cited
Copland's [1963] work that suggested all consumers could not be expected to go
through these hierarchy stages; for example, some may purchase with the mere
awareness of a brand's existence. palda's criticism of the hierarchy was directed
at each step of the process rather than at the hierarchy concept as a whole and
focused on methodologies rather than assumptions. In reviewing previous stud-
ies related to the hierarchy notion, palda concluded that there was little evi-
dence to suggest that changes in awareness had to precede rather than follow
purchase. Further, most studies of cognition showed no relationship to attitude
or behavior change. With regard to effect, some dissonance studies illustrated
that there is no obvious relationship between attitude and behavior. Finally, no
research illustrates that advertising induces intentions to buy,
Palda stated that the real problem in analyzing the hierarchy of effects was
related to the lack of careful experimental design. If there is a hierarchy of
effects operating, Palda reasoned that it would be no more expensive to illus-
trate that relationship with experimental design than to try to illustrate the
direct relationship between advertising and sales.
In somewhat of a contrast to palda's work, O'Brien [1971] sought a more
straightforward approach to test the hierarchy framework. Using a panel of over
600 housewives, O'Brien found support for several traditional hierarchy
hypotheses. Among those supported was one stating that awareness was causal-
ly prior to attitude, intention, and purchase. Further, attitude and intention
were causally prior to purchase. Finally, O'Brien found an indeterminate rela~
tionship between attitude and intention, suggesting this could be due to levels
of involvement.
During the mid-1960s, there were many psychological and sociological
theories that addressed the consistency between attitudes (affect) and beliefs
(cognition). With the exception of dissonance theory and the early work of
Fishbein [1966] a review by Insko and Schopler [1967] illustrated that very
little work looked at the consistency of cognition, affect, and conation. Insko
and Schopler developed a theory of triad consistency--the consistency among
the three dimensions in the response hierarchy. They theorized that the causal
link from behavior change to attitude change is both stronger and more direct
than the causal link from attitude change to behavior change and that most
triadic consistency results from attitudes being adapted to be consistent with
behavior.
The notion of a hierarchy of communications effect was incorporated into
consumer behavior models [Howard and Sheth 1969] in the late sixties. The
Howard-Sheth model outlined attention, brand comprehension, attitude,
intention, and purchase as the response sequence of buying behavior, an S-O-R
paradigm they labeled as "symbolic communication. " Intention was an impor-
266 Reviews of Selected A rCLLY
tant dimension of this theory because of the presumed amount of inertia that
exists between the attitude change resulting from advertising and the actual
purchase situation. It is the integration of a response hierarchy into a more
gcnernl model that creates some confusion as to the ability of theorists to
develop an advertising model of response per se [Preston and Thorson 1984],
Several other hierarchical propositions were presented throughout the 1960s
and early 1970s, Among these was the consumer acceptance hierarchy of Aspin-
wall [1964] suggesting that consumers first accepted products, then preferred
them, and finally insisted on them, Sandage and Fryburger [1967] presented an
exposure, perception, integration, and action model. Schwartz [1969] focused
on retention and attitude change in between the attention and purchase stages
of hierarchy and McGuire [1969; 1978], who proposed the information proces-
sing model, was one of the first to suggest that consumers were information
processors, This model looked at the probabilities of occurrence of each stage in
the hierarchy, McGuire suggested that the probability of a subsequent se-
quence occurring was conditional upon the previous sequence occurring. In
the McGuire model, information has to first be presented, then attended to,
then comprehended. Once comprehended, the recipient must yield to the
message, retain it, and then, behavior occurs.
Longman's [1971] model followed along the lines of the McGuirc
proposition. In Longman's view, audiences were exposed to messages,
attended to them, cleveloped perceptions and comprehension, formed beliefs,
were motivated as a result of those beliefs, and finally acted according to that
motivation and the beliefs. Holbrook [1975] also considered memory or reten-
tion important in his model that suggested consumers attended to messages,
perceived them, submitted them to memory, and then cleveloped attitudes and
intentions around them. Preston [1982] briefly reviews these and other hier-
archy models as they relate to his association model discussed below.
Challenges to and Defenses of the Traditional Hierarchy Models
Among the earliest to suggest that there is an alternate order to the tradition-
al hierarchy were Ray and others [1973]. These researchers suggested that
tl~ere are actually three hierarchy models (three-orders model) that can account
for the response of audiences to advertising messages, The first, the learning
hierarchy, was tt~e thinking model that led audiences to think and perceive,
thcrl feel or develop attitudes, and then behave. This is the traditional
cognition-affect-collation model, The second lmodcl, dissonance-attribution
hierarchy, proposes that consumers respond to advertising in a way that is ~L~
lx~sically the reverse of the learning hierarchy. That is, consumers first behave, ~
then develop attitudes and feelings as a result of that behavior, and then learn
or prvccss information that supports the earlier behavior. This
coi~:~tiol~-:~{~ c~ct-c(}g~~itioi~" model is based on the theories of attribution and
The llicrarchfj q[ Effects 267
dissonance. Finally, as a result of the work of Krugman [1965; 1966] the low
involvement hierarchy maintains that consumers behave, then learn as a result
of that behavior, and then develop attitudes as a result of the behavior and the
learning.
Ray [1974] suggested that "initial processing" could occur at each stage of the
consumer decision-making process and affect the basic attitude structures of
consumers. IIe contended that all three order hierarchies are feasible and can
be correct because consumer responses are likely to be individual- and
situational-specific. The work of Ray and his colleagues has been reviewed
more extensively by Leckenby and Wedding [1982].
Ehrenberg [1974] contended that advertising was not as powerful as its
proponents believed and that repeat buging is the main determinant of sales.
Therefore, advertisil~g must focus more on being a reinforcement mechanism
in the marketplace rather than a persuasive force. Advertisi~lg, Ehrenberg
argued, was not the prime mover of product class demand; in fact, most advc~-
tising was for brands rather than product classes. Furthermore, advertisings
affec~ on sales was neither immediate nor direct,
Ehrenberg suggested the ATR model--awareness, trial, reinforcement. Pre-
vious hierarcl~y models, according to him, did not account for markets where
shares of advertising and sales were "roughly in line
" for each brand. Nor did
these models account for common situations where advertising expenditures
dropped without resultant catastrophic sales declines. Further, these previous
hierarchy models did not explain why four out of five new brands failed in spite
of heavy advertising expendittlres and did not explain why advertising has only
a marginal effect on the total demand for a product group.
The ATR model took into account the fact that most buyers were not totally
brand loyal and that consumers had a "repertoire" of brands from which they
purchased fairly regularly. In other words, repeat buying and brand switching
patterns do not vary materially from one brand or product to another. Rather,
buying is steady and habitual rather than clynamic and erratic. Hence, the role
of advertising, according to Ehrenberg's theory, is generally defensive with the
primary objective of reinforcing previously developed repeat buying habits.
Increasing advertising expenditures in this case would not help, although de-
creasing them would hurt as competitors could make inroads illto this
"repetitive void. "
In their summary of some of the more important traditional hierarchy
models, Anderson and Barry [1979] adcled the sequence of "brand loyalty" to
the decision-making process indicating that advertisers are striving to create
the habitual buying of their brands by consumers as suggested by Ehrenberg.
Rothschild and Gaidis [1981] felt that behavioral learning theory could be a
good model for understandixlg low involvenlent purchases. Their model was
similar to Ehrenbcrg's in that they saw the role of advertising in the hierarchy
as generating awareness of the product or brand which leads to trial purchase
268 IIruiews of Sclccted Areas
behavior, At this point, if the perceived benefit of the product is realized,
satisfaction and reinforcement occur followed by an increased probability of
repeat purchase behavior and, hence, brand loyalty. Thus, the "shaping
procedures" of advertising and other promotional strategies can create be-
havioral learning and make trial purchasing more than just a random behavior.
Vaughn [1980] supported the notion that there is more than one response
hierarchy and suggested that there are actually four possible hierarchy models
that explain advertising's impact on audiences. Vaughn's model is based on low
involvement theory as well as on work in brain specialization and was de-
veioped at Foote, Cone & Belding advertising agency [Vaughn 1980; 1986].
The traditional hicrarchY model was referred to as the "rational" or "thinking"
model. Consumers who were acting in their own self-interest would calculate
consumption decisions. They would process information, develop attitudes,
and then behave accordingly. Vaughn hypothesized that the purchase of cars,
home furnishings, and new products might fall in this hierarchical category,
However, Vaughn suggested three other classifications of consumers. The
first includes responsive or habitual consumers who are lazy and want to shop
with the minimum of effort. They follow the stimulus-response learning pattern
and behave first, learn as a result of that behavior, and finally develop attitudes
based on that Iearning. The purchase of food and household items leads people
to be responsive consumers. The second group constitutes the "feeling" con-
sumer who responds more to aflect or emotion than to information. The hier-
archy of this consumer is affect, cognition, and then behavior. Fashion,
jewelry, and cosmetic products dictate affective behavior. Finally, the imitative
consumer is a reactor following a hierarchical sequence of conation, .a.Rect, and
cognition-- the opposite of the traditional hierarchy. Vaughn [1986] stated that,
while the definitive answer on how advertising works may never be known, the
FCB Grid, as these four theories have come to be known, does help advertisers
develop distinct advertising copy for distinct consumer types. In short, Vaughn
supports the multiple hierarchy thesis of Ray and his colleagues [1973] and
Krugman [1965; 1966].
Given the growing number of detractors from the traditional hierarchy camp,
Sn~ith al~d Swin yard [1982] sought a framework that could reconcile the differ-
ences in conflictiilg findings between the traditional model supporters and their
i"non e:lrllillg model>' colleagues. Called the information response model,
Smith and Swinyard integrated concepts from several research areas and fo-
cused on message acceptance and belief strengths, Their more integrated mod-
C1 claimed flexibility in dealing with the different advertising response se-
quences and illustrated the different effects of advertising and promotion be-
tween conation-trial behaviors and conation-commitment, that is, brand -,
\
lovahy.
The integrated information response model suggests a cogni -
p&
tit~il-tri;ll-:lf~~'ct-collllllitlllent sequence. In general, advertising leads to low
The Ilicrarchtj of Effects 2(W
information acceptance, lower order beliefs (cognition), and lower order affect.
The resultant behavior is trial which then provides direct experience leading to
high information acceptance and higher order beliefs and affects resulting in
commitment or brand loyalty. Advertising generally leads only to lower order
beliefs because of the high levels of source derogation, message rejection,
counterarguing, and so forth done by recipients of advertising messages.
However, when perceived risk and involvement are low, advertising may very
well move consumers directly to purchase.
In a review of the integrated information response model, Finn [1984] ex-
plored the conditions under which advertising might or might not lead to
higher order beliefs. Finn expands on the notion of beliefs and advertising's
influence on them and establishes a set of propositions that, if held valid
through research, will lead to the conclusion that advertising rarely influences
affect directly and supports the proposition of Smith and Swinyard that product
trial is a dominant predictor of attitude.
Strong support for a traditional hierarchical processing sequence in advertis-
ing response is found in Preston's [1982] "association model. " "rhe association
model focuses on a comprehensive advertising process that takes into account
advertising research techniques. Further, the model expands on the three
traditional phases of cognition, aflect, and conation as can be seen in an ex-
panded version of the model [Preston and Thorson 1984] in Figure 2. Previous
hierarchy models failed to portray the important role that consumer informa-
tion processing plays, for example, as in prior states of perception, evaluation,
and stimulation. While prior models may have implied these roles, the associa-
tion model makes them explicit. The association model is a high involvement
hierarchy of effects model; low involvement may threaten the ordering of a
hierarchy model but not the basic notion of the hierarchy concept and its
various steps.
Preston and Thorson [1983; 1984] reviewed the challenges to the hierarchy of
effects model, concluding that, with expanded steps to the original association
model, the notion of a sequential hierarchy could be kept intact, specifically as
that hierarchy related to advertising objectives and effectiveness. They did not
feel that the criticisms of alternate ordering models such as those proposed (and
discussed earlier) by Krugman, Ray, and Vaughn or models of omitted steps
based on the low-involvement work of Krugman and the unconscious proces-
sing work of Zajonc [1968; 1980a] were compelling enough to dismiss the
traditional hierarchy concept as expanded upon in the association model. pres-
ton and Thorson admit that the association model is more of a marketing model
than an advertising model but contend that an advertising model cannot be
solelg an advertising model since advertising is part of marketing. In this
respect, advertising researchers must be content in knowing that this sequen-
tial or behavioral process begins with advertising.
In her review of the hierarchy framework, Moriarty [1983] proposed to
270 Reviews of Sckcted Areas
Dist) il>lltion ~ Ve}]iclc Exposure + Ad Exposure + Ad Awdrrness + Ad Elements Awareness -=-q
Prior
Percept iun
~ ~vf::;i,,n ~ J:;ion
I
L
Action: Search ~Scwch Perception ~ Search Evaluation ~ Search Stimulation
* 1
Action: Trial ,--> Trial Perception ~ Trial Eva[ui\tion + Trial Stim{llution .--l
L * 1
Action: Adoption + Adoption Perception + Adoption Evaluation ~ Adoption Stimulation
%urce: Preston, Ivi~n L. iitld Tllorson, Esther. (1984). "lhe Exp;mded Association Mode]: Keeping
the Hierarchy Concept Alive. " journal oj_A&er-tkit]g lksearch 24 (Feb. /hlarcll): 60.
Figure 2. The Expanded Association Model
replace the process-oriented hierarchy model with one based on the three
domains of advertising: perception, education, and persuasion. She contends
that these three domains are equally important and are interdependent and
simultaneous. Suggesting that there is no theory of how advertising works and
that the linear processes of the hierarchy models do not illustrate how the
sequences are connected, what they have in common, and their patterns of
communication effects, Moriarty's model describes the continua of the three
domains, For example, perception runs the continuum of no awareness of
subliminal or preawareness to recall. Education results in no learning,
association, comprehension, generalization, and discrimination. Finally, per-
1suas on reinforces existing attitudes, forms new ones, or changes old ones.
These domains result in inquiry, trial, considered purchase, or repurchase. By
providing for this continuum of effects from advertising, Moriarty holds that the
debatable concept of process or sequence and the simplistic steps of the hier-
archy model are eliminated.
Finally, Rust and Henderson [1985] contend that the major flaw of the
hierarchy models is that they imply causal relationships between the attitude
structure components rather than an integration of these components. These
models disregard the possible interaction between cognition and ail'ect. Their
work is based on the proposition that affect and cognitive processing can and do
occur separately and that emotion is not the same as a.fleet.
Advertising
The hierarchy of Effects 271
and the Hierarchy Stages
Because the traditional and alternate hierarchy models are so steeped in the
concepts of cognition, affect and conation, it is important to review the clevelop-
ment of the research that investigates these sequences and/or interactions. This
section reviews the development of the literature relating to the impact of
advertising on the cognitive, a.flective, and conative stages of' the hierarchy of
effect models.
Advertising and Cognition
In his discussion of the history of cognition and social cognition, Zajonc
[1980b] explains why the area of cognition has dominated social psychology for
four decades. Among the most important reasons are that social behavior is
vastly complex and necessitates an analysis and understanding of the thought
processes--that is, the way that human beings process information. Too, be-
cause social psychologists have tended to focus on attitudes and they have
always been regarded as cognitive and affective organizations impacting on
behavior, cognition has been a key element of research in social psychology.
Zajonc also contends that World War II, with its totalitarian doctrines finding
solutions in biology and genetics, led important thinkers to conclude that solu-
tions must come through thinking and reasoning rather than through biology
and genetics.
In essence, cognition is the mental processing that occurs when people are
exposed to information. It is quite clear that the information processing skills of
people differ greatly. For example, Petty [1980] contends that cognitive proces-
sing is a function of both the ability and the motivation of people to engage in
that processing. There is, in Petty's terms, a "need for cognition. " In both the
traditional and alternative ordering hierarchy models, the cognitive phase is
integral to the models. While most cognitive and social psychologists believe
that cognition precedes the attitudinal or affect stage [Van Raaij 1984] more
recent researchers such as those supporting the alternate ordering model
disagree. Nevertheless, cognition in advertising is critical.
The impact of advertising on the cognitive component of the hierarchy has
been studied extensively. Among the earliest investigators of advertising and
cognition was Zielske [1959] who studied remembering and forgetting in
advertising. Zielske investigated the number and intensity of advertising
schedules and their impact on how well audiences remembered advertising
content, He contended that advertising will quickly be forgotten by consumers
if they are not continually exposed to it. Rohloff [1966] suggested the use of
quantitative analyses that would shed light on the structural relationship that
existed in measuring recall of an advertisement due to the variance in time and
audience composition. The importance of recall as a measure of advertising
272 Rcvicrcs of Selected Areas
effectiveness has been questioned by many researchers. Haskins [1964] con-
cluded that, while learning and factual recall of advertising do occur, these
measures appear irrelevant to attitude change. Gibson [1983] argued that recall
theory was weak on empirical grounds and suggested that advertisers not use
recall as a measure of copy effectiveness given its weak association with
persuasion. More recently however, Stewart [1986] concluded that recall and
comprehension are important variables of persuasion, and Zinkhan and Gelb
[1986] maintain that Starch recall scores may be good predictors of attitudes,
Woodside and Wilson [1985] demonstrated a strong, positive association be-
tween brand awareness and brand preference.
Krugrnan [1972; 1975] and Naples [1979] believe that advertisers often waste
money by trying to "overhit" target audiences. They argue that advertising is
ineffective unless consumers are in or near their "purchase cycles" and trying to
build up recognition and recall through continuous advertising expenditures
may be inefficient. Krugman [1977; 1986] has argued that recall measures, or
"cognitive effectiveness" of advertising do not take into account the amount of
recognition that is processed by advertising audiences that cannot be captured
by recall measures. Singh and Rothschild [1983] investigated the problems and
distinctions with and between recognition and recall measures in
low-involvement situations and Bettman [1979] suggests that consumer brand
choices may require only recognition when choice is made in the store but
recall when choice is made in the home. Thus a shopping list for groceries made
at home may require recall learning while impulse brand choice in the store
may only necessitate brand recognition. These are both elements of cognitive
processing and their differences have been illustrated through A Study of
printed Advertising Rating Methods (PARM) [Krugman 1985],
Olson, Toy, and Dover [1982] investigated the differences in advertising
effects among cognitive response models, cognitive structure models, and a
combination response or structure mass communication model, A cognitive
7
res onse model states that advertising exposure creates cognitive responses
tha impact on attitude or attitude change. Cognitive structure models are
basically attitude models such as the Fishbein model. Beliefs result from ad
exposure. Beliefs impact on attitudes which impact on intentions and behavior,
The combination model holds that audiences respond cognitively from adver-
tising exposure. Beliefs occur and impact on attitudes, intentions, and
behavior. Their results support the viability of the combination model in ex-
plaining advertising effects and cognitive processing.
hfuch research in the advertising or cognition arena has focused on
audiences' defenses to advertising information, For example, Venkatesan and
IIaaland [1968] illustrated that visual and behavioral distractions for a television
con~mercial can jeopardize the recall of brand names and product types. Cohen
i~l~d I Iouston [ 1972] empirically illustrated that brand loyal consumers for
tootl~pastm tencled to engage in cognitive dissonance towarcl the brands which
The Ilierarch!j of Ej'f(zcts 273
they did not buy. Wright [1973; 1975] concluded that personal characteristics
clearly affect the cognitive responses that consumers have toward advertising.
Cognitive processing is very complex and cognitive models, as developed by
experimental psychologists, do not always provide marketers with unique mod-
els of information processing [Hughes and Guerrero 1971]. Balance models,
reinforcement models, congruity models, dissonance models, attribution
models, and so forth are presented by their supporters as explainers of the
variance in consumer responses to advertising and other social
communications. But it is clear that cognitive processing is only one part of the
complex advertising response puzzle; another is affect.
Advertising and Affect
While not studied to the extent that cognition has been studied, there is,
nevertheless, a plethora of information regarding advertising and the affective
component of the hierarchy of effects model. It is gcnerolly l~cld that affect
consists mostly of attitudes, although Kreshel [1984] in her study of emotion in
advertising suggests that research in this area has tended to confound affect,
emotion, and attitude. On the other hand, Batra [1986] uses the terms feeling,
emotion, and aflect interchangeably.
According to Rokeach [1966-67], "an attitude is a relatively enduring orga-
nization of beliefs about an object or situation predisposing one to respond in
some preferential manner" (p. 530), It follows, then, that attitude change is a
change in predispositions to behavior. Rokeach considered it critical to separate
attitude toward an object from attitude toward a situation and felt that the>
failure to do so by researchers retarded the growth of attitude and
attitude-change theory. The importance of attitude and attitude change relates
to their predictive ability on related behaviors and behavior changes. In this
light, Crespi [1971] held that attitudes must be treated as multidimensional and
should not be generalized as predisposing variables that will have common
effects in a variety of situations. Rather, attitudes should be treated as a com-
bination of beliefs with varying degrees of intensity.
Marketers and advertisers have long sought to show the relationship be-
tween attitudes and sales behavior. Assael and Day [1968] assumed the predic-
tive cognition (recall) and affect (attitudes) on market share performance. They
concluded that affective components of the hierarchy of effects model were
consistently more powerful in explaining market share vari:mce than were
either cognition or conation (usage) components. In answer to the critics of the
hierarchy of effects who contend that sales is the only real measure of advertis-
ing effect, Boyd et al. [1972] concluded that it might be more functional to look
at advertising's ability to impact on attitudes toward salient product characteris-
tics which ultimately will affect sales and profits favorably.
In defining attitudes as mental or neural states of readiness, Smith and
274 kwicu's of Selected A nms
Swinyard [1983] studied the relationship of attitude-behavior consistency in
light of their integrated response model, They showed that there is a low
correlation between attitudes and behavior and pointed to the work of Fazio
and Zanna [1978; 1981], suggesting that the director indirect means of produc-
ing an attitude determined attitude strength ancl, subsequently, consistent
behavior. As an example, advertising produces low-order or indirect attitudes
which are likely to correlate poorly with purchase behavior. On the other hand,
product trial is more likely to lead to stronger attitudes and higher correlations
with consistent purchase behavior. In an experiment, 64 percent of subjects
developed lower-order affect from advertising, while 65 percent of subjects
developed higher-order affect from direct product trial. However, Fazio and
Zanna illustrated that indirect experiences (e. g., aclvertising) can sometimes
]ead to stronger tiective responses than direct experience. Smith and Swill-
yard hold that advertisers should understand those situations when their mes-
sage strategies can lead to higher-order tiect.
Recent research by MacKenzie et al. [1986] illustrates the importance of the
high-order aflect thesis by studying the importance of an audience's attitude
toward an advertisement as well as a brand, They looked at four alternative
models to explain the mediating effect of the ad itself: the affect tran.sfcr
hypothesis, dual mediation hypothesis, reciprocal rneclintion hypothesis, and
independent influences hypothesis. The a.6!ect tral~sfer ]Iypothesis assumes a
direct, one-way causal flow from ad to brand attitl]de. The dual mediation
hypothesis suggests a more indirect flow from ad to brand attitude by way of
source perceptions that influence the cognitive and affective reactions to mes-
sage content. The reciprocal mediation hypothesis, in essence, posits a Heider
theory of balanced cognitive relationships between the receiver of the ad and a
liking or disliking toward it. That is, consumers wilI either like or dislike both
the ~d and brand. Finally, the independent influences hypothesis assumes that
attitudes toward the ad and the brand arc independent. The dual mediation
hypothesis, foc~lsing on source impacts on cognition and attitudes, was con-
sidered superior in explaining the relationship between attitude towards an ad
and towards a brand. The directional impact of cognition on attitudes and
attitudes on conation is critical to understanding the hierarchy of effects
literature, The research highlighted below indicates that we have anything but
definitive conclusions regarding these causal flows,
The Causal Flow of Affect
Hccent research on the role of affect in consumer behavior [Peterson et al.
1986], lcxl to the following conclusions: c
1. &afi'ect l}as long been a step-child to cognition in understanding consumer
behavior;
The llicmrchtj of l?jfects 275
Z, affect should be studied at least as much if not more than cognition
regarding its role in influencing consumption choice;
3. while affect has long been held as being post-cognitive, it may well be
simultaneous or directive to cognition;
4. AfI'ect is difllcult to study because it is a messy construct not easily
defined and very difficult to operationalize,
The causal relationships between cognition and affect and tiect and conation
are at the verY heart of the hierarchy of effects concept. Zajonc and M arkus
[1982] hold that preferences may well involve both cognitive and aRective
dimensions of attitudes and that, in some cases, cognitive components can be
dominant in determining these preferences while at other times affective fac-
tors may dominate this influence 011 preferences. While some [for example>
Mandler 1982; Lazarus 1981; 1982; 1984] argue that cognition must precede
affect, the theoretical argument of Zajonc and Markus suggests the possibility of
tiect without "cognitive participation. " That is, tiect and cognition are inde-
pendent and afl'ect need not always be postcognitive as in the case of acquiring a
taste for chili peppers through habit and surroundings rather than persuasive
information. There is a significant motor component to preferences and emo-
tion is most expressed through this motor component. The importance of this
somatic dimension of af+ect is illustrated in the recent facial response research
of Cacioppo et al. [1986]. The thesis by Zajonc and Markus that a.fIect can be
precognitive has recently been disputed by Tsal [1985] and further defended by
its proposers [Zajonc and Markus 1985]. In viewing the cognitive research of
repeated exposures on attitudes and preferences, Zajonc [1986] comments that,'
while this is the most primitive means of understanding preferences, research-
ers still do not completely understand how exposure effects work. In that
respect, understanding how affect, a much more complex component of
behavior, works will be considerably more dificult for researchers, but neces-
sary indeed.
Emotion in Advertising
Holbrook [1978] argued that attitude research has been too narrowly focused
and more recently [Holbrook 1986] that emotion in consumer behavior is one of
the most needed areas of investigation. Researchers have focused too much on
man as a computer receiving and processing information, evaluating that
information, and then behaving based on that processing. He argues that affect
has been too narrowly conceived as a result of this information processing
approach and that the rich attributes of emotion such as love, hate, fear, anger,
joy, and sorrow have been ignored in the research process. Holbrook proposes
replacing the traditional hierarchy model of cognition, a.fl'ect, and behavior with
"consciousness, " "emotion, " and "value. " In this model, emotiw~ is the "linking
I
pin that holds together the consumption experience" (p. 23). Holbrook points
to the wave of emotion-based work by advertising copywriters as a clue to the
importance of this construct in consumption behavior,
Advertising is laden with emotional messages rather than factual claims. This
was recently illustrated by Holrnan [1986] who investigated the roles of procl-
ucts in the emotional lives of consumers and analyzed advertising from an
emotional perspective. Holman pointed out the importance of differentiating
between consumers' emotional reactions toward advertising and a particular
brand, although she felt that the two should be as congruent as possible. In
positing that visual imagery advertising can be very effective in forming or
changing attitudes, Rossiter and Percy [1980] state,
Wc often lose sight of the fact that the reason why beliefs (cognitive) are
capable of altering attitude (a&ective) is because of their own affective
connotations, It is the affective or emotional feeling that accompanies the
belief that is critical (p. 11).
In an extensive study of emotion and advertising, Kreshel [1984] noted the
tendency to view cognition and emotion as separate rather than interrelated
processes. The three levels of the traditional hierarchy model, according to
Kreshel, are viewecl as three separate levels of response aligned in a sequential
or causal pattern rather than an interfactional one. This makes the hierarchy
model rather simplistic, in spite of its appeal as a planning and decision making
model, A more realistic model, according to Kreshel, would be one that made
emotion the center of the traditional cognition, affect, conation sequence and
changed the levels from sequential to circular as depicted in Figure 3. Thus, as
I-Iolbrook [1986] proposes emotion as the linking pin of consumer behavior,
Kreshel suggests emotion as the linking pin of the advertising response model.
Research in the area of affect, and the hierarchy of effects in general, clearly
supports the Conclusions that affect is a messy construct, difficult to define, and
hard to operationalize [Peterson et al, 1986], In spite of that, Batra [1986]
states,
Such research should bring a more complete understanding of how adver-
tising is processed and a more balanced assessment of the role of cognition
and affect in that processing (p. 85).
The final phase of the traditional hierarchy is conation or the behavioral results
of advertising programs. We turn our attention, now, to the research on cona-
tion ill the hierarchy of effects response process.
Advertising and Conation
Aca(lcmics and practitioners alike have long debated the appropriate meas-
ures to i:sc il~ measuring the effectiveness of advertising programs. Two schools
of tl~o~l~i]( have emergecl in the literature: the hierarchy of effects school of
The Hierarchy of Effects 277
\ Cognition
Source: Kreshe], Peggy J. (1984). "En~otion and Advertising: The `Crand Idea' of the 80's. " JVorkin[
paper No. 17, Advertising Research Center, ~lniversity of I]]ir]ois, No\'enllJer, 23.
Figure 3. The Role of Emotion in the Hierarchy of Effects
thought and the sales school of thought [Anderson and Barry 1979]. Proponwlts
of the hierarchy camp contend that the effectiveness of advertising should be
measured by its impact on some hierarchy of communication responses such as
the ability of the audience to recognize brand names, recall main copy points,
generate positive attitudes, change an image, and so forth. "rhus, advertising is
effective when it generates activity in the cognitive and affective stages of the
hierarchy of effects model. Proponents of the sales camp, on the other hand,
argue that the only real reason for advertising is to impact on sales and profits.
There is little value to the manager who spends millions of dollars on an
advertising program and discovers that there is a ninety percent recall on
his/her brand but sales are flat! As Campbell [1969] states,
Advertising's principal job is to make sales at more profit than would have
been made without advertising, . . . If this is true, then it must also be true
that the proper goals for advertising are those that will be profitable means
for producing sales. Can there be any other sound reason for spending
marketing funds? (pp. 17­18)
Students of advertising realize, however, that a direct measure of sales and
profit results from advertising is highly complex, if possible at all, in most
marketing situations. This is due, simply, to the complexities of uncontrollable
variables such as competitive actions, economic conditions, government
regulations, and such, as well as the contributions of marketing mix variables
other than advertising to the sales and profit picture for any particular brand m-
portfolio of brands in a firm. Additionally, the direction of attitude and sales or
brand preferences and brand choice confound a clear understanding of the
~~ direct sales and profit impact of advertising.
~J Conation measures other than sales and profits have been investigated and
further illustrate the complexities of relating advertising expenditures to var-
ious consumer decision processes. Among these inclilde the problems in
278 Reviews of Selected Areas
measuring advertising expenditures and short term market share [Fothergill
and Ehrenberg 1965; Buzzell 1964; Buzzell et al. 1965], advertising, attitude
$
c anges, and brand choice [Ginter 1974; Cobb and Hoyer 1985; Zinkhan and
"nkhan 1985], differences in responses and media mixes [Leckenby and Kishi
1984; Rust and Leone 1984], and the effects of broadcast bans on the cigarette
industry [Holak and Reddy 1986], One of the more promising behavioral meas-
ures of advertising effect is the analysis of inquiries resulting from advertising
campaigns [Schweiger and Hruschka 1980],
In spite of the conceptual problems in illustrating causation links between
advertising expenditures and sales, much research has been done in this area,
We now turn our attention to that research.
Advertising and the Sales/Profit Link
One of the earliest empirical studies investigating the link between ad ex-
penditures and sales was PaIda's [1964] extensive study of the Lydia Pinkham
Medicine Company. palda sought to determine the cumulative effects of adver-
tising over time using the Koyck distributed lag model, a model sensitive to
effects beyond the period when the stimulus occurs. The use of Pinkham was
ideal in that the firm had no sales force, and provided no credit, discounts, or
point of purchase materials. Additionally, the product studied had no close
substitutes at the time. An investigation of sales and advertising expenditures
between 1907 and 1960 illustrated remarkable stability between the two
variables, palda concluded that the momentum of past advertising can help
sustain sales and that, for Pinkham, advertising was only wise for its long-run
effects.
Bass [1974] reanalyzed the Federal Trade Commission's study of 97 food
companies and the link between profit and the advertising/sales ratio. In criti-
cizing previous methodologies used to illustrate relationships between ad/sales
ratios and firm profitability, Bass concluded that, for some industries, profit and
s!les ratios are independent of each other, and for others, they are highly
related to each other. In fact, there maybe some instances where the profitabil-
ity of a firm causes a higher advertising/sales ratio [Comanor and Wilson 1967;
Schmalensee 1972]. Lambin's [1975] extensive study of 108 brands in 25 mar-
kets in eight European countries also showed mixed results on the causality of
advertising on market power and pefiormance. Lambin concluded that the
economic power of advertising has been overestimated by friends and foes of
advertising, and that an analysis of advertising without a simultaneous analysis
of other marketing mix variables will likely lead to a misunderstanding of
advertising's effects,
In several experimental designs on the Budweiser brand of beer between
1963 and J !368, Ackoff and Emshoff [1975] sought to measure the effect of
advertising on sales as well as to explain the causal effect of advertising on
The ilierarchy of Effects 279
consumer purchases. By varying advertising budgets in different markets and
controlling for sales and sales promotion efforts, the researchers found that no
change in advertising expenditures resulted in no change in sales; a 50 percent
increase in advertising expenditures resulted in a 7 percent increase in sales;
however, a 25 percent decrease in advertising expenditures resulted in a 14
percent increase in sales! Further experimentation illustrated that advertising
reductions would have no negative effects on sales of Budweiser. Over the
period of the research at Anheuser-Busch, a decline in per barrel advertising
expenditures from $1.89 to .80 resulted in sales of 7.5 million barrels to 14.5
million.
In a further investigation of distributed lag models, Rao and Miller [1975]
noted the problems of econometric studies and the association between adver-
tising levels and sales. Since many managers base advertising budgets on sales
levels, the cause-effect relationship is often not clear. Rather than using annual
advertising expenditure and sales data, bimonthly data were analyzed. Results
showed that dollars spent on a variety of media advertising, price-off, and trade
promotions were related to sales of five Lever brands in 15 different districts.
In an investigation of alcohol consumption in Canada, Bourgeois and Barnes
[1979] concluded that advertising expenditures do increase the consumption of
beer but do not increase the consumption of liquor. A positive relationship was
found between the amount of print advertising and per capita consumption of
beer. However, a negative relationship was found between the level of print
advertising and the per capita consumption of spirits. The researchers con-
cluded that uncontrollable factors such as disposable income, population
variables, education, broken homes, and so forth explained more variance in
consumption habits than did advertising expenditures, In a more extensive
study of the impact of advertising on alcohol consumption in the united States,
Wilcox et al. [1986] studied the relationship between 1964 and 1984. They
concluded that, despite the increased advertising expenditures for beer over
the study period, beer consumption actually declined. The same was true for
spirits and would have been so for wine had it not been for the recent surge in
the wine cooler market. Their data illustrated that total alcoholic beverage
advertising and total alcoholic beverage television advertising had no rela-
tionship with total alcoholic beverage consumption, The only important rela-
tionship for consumption of alcoholic beverages was with demographic or eco-
nomic factors.
Among other questions in trying to determine the shape of the advertising
response function, Simon and Arndt [1980] investigated the relationship be-
tween an increase in advertising expenditures and the proportional increase of
units sold. The vast majority of studies pertaining to this question illustrated
that there are diminishing returns for increased advertising expenditures.
Further, they concluded that there is very little evidence illustrating increasing
returns from increased advertising expenditures. In investigate ing the shape of
280 Reuiei.os of Selected Areas
the advertising or sales response function for eight Australian firms, MetwallyI
I [1980] focused on consumer demand for goods from a given firm. Applying the
~oyck distributed lag model, MetwaIly concluded that the more attached con-
sumers are to a particular brand, the lower is the sales elasticity of advertising,
Callahan [1982] more recently investigated a ten-year relationship between
advertising or sales ratios and industry prof3ts, Hypothesizing that heavy adver-
tising expenditures should be associated with high profits and light expendi-
tures with not-as-high profits, Callahan found that these propositions heId true
in 151 out of 166 observations, Further, while no attempt was made to deter-
mine causality, cross-sectional regression analyses indicated a clear relationship
between profits and advertising. On the other hand, Asker and Carman [1982]
imply that profitability of firms that have frequently purchased brands may be
less than optimal in that many of these firms are likely to overadvertise. In a
review of 69 field experiment studies and 60 econometric studies, Asker and
Carman concluded that there are few persuasive studies shedding light on the
relationship between advertising and sales, that this relationship, for reasons
mentioned earlier in this paper, is very difRcult to model, and there is empir-
ical evidence of overadvertising by advertisers. They called for advertisers to
experiment with reduced advertising budgets and focus more on copy quality
than on increased expenditures, In sum, they concluded that, even when
advertising is found to be effective, its significance is most likely rather small,
In a "replication analysis" form of meta-analysis, Assmus et al. [1984] investi-
gated the parameters of 128 econometric models from some 22 studies re-
searching the relationship between advertising and sales. The majority of the
studies were for mature products in the U.S. The authors concluded that more
research must be done outside the U.S. and Europe and for good's other than
packaged goods if we are to ascertain the real impact of advertising on sales,
They also concluded that the quasi-experimental design is too imperfect to
determine the impact of particular factors on explaining the carryover effects of
advertising. They called for a more comprehensive model of advertising effec-
tiveness for theoretical predictions under specific market conditions.
In a study of the German cigarette market, Leeflang and Reuijl [1985] con-
cluded that advertising has a significant influence on primary demand although
that influence tends to diminish over time. using a Koyck transformation, the
authors analyzed annual, bimonthly, and monthly data in the West German
market. As observations are disaggregated over time, the direct influence of
advertising decreases. These findings support the previous empirical work of
Bass and Leone [1983] as well as the conclusion by Clarke [1976] that annual
observations tend to overestimate the cumulative effects of advertising,
In a recent response to the conclusion that overadvertising can impede prof-
itability [Asker and Carman 1982], Tu]l et al. [1986] investigated three separate
sales response functions for advertising: diminishing returns, saturation, and
quadratic response, Because of the illustrative nature of their research, the
The Hierarchy of Effects 281
authors cautioned against drawing hard and fast conclusions about sales re-
sponse to advertising. They did, however, discuss some important implications
for marketing managers and advertising researchers. First, the contribution to
present-period profits and fixed costs changes very little over a rather extended
range of advertising expenditures. Second, all three response function models
indicate that the contribution margin is a partial determinant of an optimal
advertising budget. Thus, firms that emulate their competitors' budgets should
do so only when their contribution margins are similar. Third, finding a signifi-
cant relationship between advertising and profits is difficult in the face of the
"flat maximum principle": the notion that there are only very small changes in
profits over a wide range of advertising expenditures.
In a final study, Batra and Vanhonacker [1986] conducted an empirical analy-
sis in a "quasi-field experiment" containing extensive time-series observations
to determine the cognitive, affective, and conative effects of advertising
repetition. Their results indicated that the effects of advertising on these three
hierarchical variables or stages are neither immediate nor simultaneous.
Rather, they occur in "situationally varying and complex patterns of temporal
precedence" (p. 24). Batra and Vanhonacker concluded that when there is high
prior brand awareness, the traditional cognitive-fiective-conative hierarchy
holds but, over time, a pattern develops where awareness follows rather than
precedes purchase intentions, In addition, the relationships among the hier-
archy variables appear to develop more slowly for low prior brand awareness
than for high prior brand awareness. In short, Batra concludes that there is
clear evidence for "multiple hierarchies" of advertising effect.
one of the more recently studied arguments in the hierarchy literature is
that the level of consumer involvement both with the advertising and the
brand/product advertised is an important mediating variable impacting on the
sequential nature of the three hierarchy phases. It is to that issue that we now
turn.
The Hierarchy of Effects and Involvement
Since Krugman [1965; 1966-67] suggested that much of television learning,
like the nonsensical, is learning without the need for involvement, consumer
behavior and advertising researchers have studied the concept of involvement
:extensively. A fair amount of this research has focused specifically on advertis-
ing impact and the hierarchy of effects. In essence, involvement is considered
to be a mediating variable in the effectiveness of advertising executions and the
sequential nature of the three main phases of the hierarchy of effects. Zaltman
~~and Wallendofi [1983] raised potential problems with the relationship between
"-4 1nvo vement and the hierarchy framework, one is the inappropriateness, from
a lack of research, of concluding that low price products automatically lead to
2821 Reviews of Selected Areas
lowinvolvement purchase decisions. Another isthatthe continuous nature of
involvement makes it difFicult to suggest two sequences of the hierarchy (for
example, the low learning and high learning hierarchies) which are reversals of
each other,
In spite of these problems, advertising and consumer behavior researchers
find much merit in the potential of involvement as an explanatory variable for
the hierarchy of effects and the impact of advertising on receivers. For
example, Krugman [1965] theorized that the level of involvement does not
increase resistance to advertising messages but rather shifts the sequential
nature of them. Thus, under low involvement situations, learning is passive in
that consumers do not link or connect the advertising message directly with
needs and wants. Rather, over time and through repetition, advertising is
persuasive in creating purchasing behavior (conation) followed by cognitive
activity. Thus, consumers learn information randomly in the process of passive
information gathering as opposed to being active information processors and
seekers [Assael 1981].
Involvement theory also stems from the work of Sherif and others [1965] in
social judgment theory. For given persuasive communications, the highly in-
volved recipient increases "latitudes of rejection" or areas of unacceptability
regarding the main communication issues. Therefore, messages such as adver-
tising that deal with important issues for which people are highly involved,
have an increased probability of being rejected compared to those for people
who are less involved, The more uninvolved consumers are willing to consider
more brands but less willing to spend time processing information and evaluat-
ing brands, In short, then, as stated by Assael [1981], ". . . advertising is
perceived with little cognitive activity . . . " (p. 86).
A third theory of involvement is suggested by Petty and others [1983] and is
based on the Elaboration Likelihood Model (ELM) studies of Petty and Caciop-
po [1981]. The ELM posits that consumers are more likely to elaborate or
devote time to cognitive effort when a message issue and/or product involve-
ment is high rather than low, The ELM is based on central and peripheral
routing, In an advertising context, under central routing, a message recipient
will "diligently process" an ad when the issue or the product in the ad is highly
relevant to that recipient. This diligence is likely to lead to attitude change.
With peripheral routing, consumers may change attitudes, not as a result of
committed cognitive activity, but rather because of the positive or negative
cues associated with the issue or product in the ad, such as the entertainment
value of the ad or its having a pleasant celebrity spokesperson. High
involvement, then, leads to central processing along the lines of the traditional
(cognitive-affect-conative) hierarchy model. While the traditional model may ~,a
also be followed with the peripheral processing, the degree of cognitive proces- "-
>
e>
sing is clearly much less than in central processing.
one of the main problems with the involvement construct is getting re-
The IIierarchy O( Effects 283
searchers to agree on its definition and operationalization. Early work by adver-
tising and consumer behavior researchers included that of Lastovicka and
Gardner [1979] who contended that involvement, while a promising construct,
seemed to be a "bag of worms. " They provided empirical evidence that sug-
gested there is a continuum of involvement [also supported by DeBruicker
1979] on which it was possible to measure and place products. Rothschild
[1979] speculated about media frequency, media classes, message content,
price, and the probability of success for advertising strategies in high and low
involvement situations. Tyebjee [1979] refined the concept of involvement in
an advertising context operationalizing it through some form of arousal, for
example, physiological, cognitive, attitudinal, interest, and mood (p. 99). Prod-
uct involvement has been studied in its relationship to brand commitment
[Traylor 1981b] as well as product importance concepts [Bloch and Richins
1983].
Korgaonkar and Moschis [1982] illustrated empirically that product involve-
ment was a mediating variable in the postpurchase evaluation procedure.
Recently, Park and Young [1986] showed that low involvement, cognitive
involve ment, and affective involvement have differential effects on the forma-
tion of brand attitudes. Extensive work on conceptualizing the involvement
construct has been done by Zaichkowsky [1986]. She proposes, as a result of a
review of the involvement literature, three antecedents to the involvement
construct: characteristics of the person receiving the advertising (e. g., value
system, unique experiences), the physical characteristics of the stimulus (e. g.,
media class, ad content, product class variations), and varying situation (e. g.,
whether one is, or is not, in the purchase cycle for the advertised product).'
Further, involvement can be operationalized with respect to involvement with
the advertising itself, the products being advertised, and the purchase decision
itself. A variety of results can follow. Figure 4 illustrates Zaichkowsky's involve-
ment conceptualization, An important implication for advertising managers
(and researchers) is the ability to determine to what extent their products are
perceived as high and low involvement products by target markets. We turn to
that issue now,
Measuring Involvement
Several attempts to scale the involvement concept have been made recently.
Slama and Tashchian [1985] hypothesized a relationship between family life
cycle, sex, work lives, education, income, and product involvement. They
developed a purchasing involvement scale and concluded that demographics do
relate to purchasing involvement in predictable ways. In analyzing data from 37
product categories for over 7500 interviews, Kapferer and Laurent [1985-86]
concluded that involvement is not a single dimension construct. Rather it
should be viewed as a "profile" of a variety of dimensions including perceived
284 Reviews of Selected Areas
I
` Antecedents of Involvement
Derived from the Literature
Possible Results of
Involvement
/
elicitation of
Person Factors
\
counter arguments to ads
-- needs
-- importance \ \
-- interest
-- vatues
Object or Stimulus
-- dtierentiation
of alternatives
-- source of
communication
-- content of
communication
Situational Factors
-- purchaseluse
-- occasion
Factors
Involvement
/
effectiveness of ad
to induce purchase
relative importance
/
of the product class
with advertisements
%
perceived differences
in product attributes
with products
preference of a
with ourchisse
particular brand
decis;ons
E
influence of price
on brand choice
amount of information
search
time spent deliberating
\
alternatives
N type of decision rule
used in choice
INVOLVEMENT = f (Person, Situation, Object)
The level of involvement may be influenced by one or more of these three factors.
Interactions among persons, situation and object factors are likely to occur,
Source: Zaichkowsky, Judith L. (1986). "conceptualizing Involvement. " ]ourmzl ofAduertising 15,
no. 2: 6.
Figure 4. Conceptualizing Involvement
risk, interest, pleasure value, and the ability of a brand to express one's status.
Further, they conclude that the dichotomy of high and low involvement is too
oversimpl~led; these types only accounted for 2/5 percent of the purchase situa-
tions studied, Instead there are many dtierent types of involvement that ex-
hibit very little redundancy among themselves. Their research identified ten
types of involvement profiles ranging from "minimal involvement" to "riskless :
involvement" to "total involvement,"
In a rigorous approach to develop a Personal Involvement Inventory (PII)
Zaichkowsky [1985] sought to devise a scale that would measure involvement
focusing on personal relevance. The approach used met standards for intern - `$P
reliability, reliability over time, content validity, criterion-related validity, anP
construct validity, The resultant PII consists of twenty hi-polar adjectives that
could be used to judge the involvement of respondents with products,
advertising, and purchase decisions, although construct validity of the scale was
supported only for product objects (p. 349).
The Hierarch?i of Effects 285
Zaichkowsky calls for a develop-
ment of scale norms vis-ii-vis empirical studies on involvement in order to
determine its "true validity. " The PII is presented in Figure 5,
Contributions of the Hierarchy of Effects
In a recent discussion of the value of practitioner versus academic orienta-
tions in research, Brinberg and Hirschman [1986] discuss the importance of
"triangulation" in the social and behavioral sciences. Triangulation suggests to
researchers who wish to study a single approach to research that a multiple
perspective relieves the pressure of constraints resulting from a singular focus.
"For the field of marketing to advance . . . multiple perspectives are needed so
that the strengths of one may compensate for the weaknesses of the other" (p.
171). Research in the hierarchy of effects field has demonstrated a multiple
perspective approach. Practitioners have long sought to measure the impact of
advertising on sales in a variety of ways. Academics have also sought to do this
buy have advanced the field by investigating intervening variables that are
more readily measured than sales.
For the vast majority of the hierarchy of effects` "life," the single dominating
a focus has been that audiences must first process information from advertisers,--
then develop (and/or change) attitudes resulting from that information proces-
sing stage, and finally behave (positively or negatively) toward the advertiser's
product as a result of the message and resultant attitude. Only more recently .
through the work of Krugman, Ray and his colleagues, Petty and his colleagues,
and others, has this research focus of cognition-affect-conation been
challenged. Practitioners such as Vaughn and his colleagues at Foote, Cone and
Belding have joined the challenge of the multiple hierarchies notion and sought
to illustrate its validity through a variety of empirical research methodologies.
This triangulated perspective is healthy for the hierarchy concept, and as Zink-
han and Fornell [1986] recently concluded, advertising response theory and
measurement have progressed but measurement problems limit the advance-
ment of advertising theory.
The persistent and pervasive attention given to the hierarchy of effects in
advertising and marketing research and practice attest to its continuing impor-
tance as an academic and practitioner concept. Nevertheless, it seems clear
that questions remain as to the future direction for research into the hierarchy
of effects. Should, for example, research focus on the traditional sequence or
alternate order models? Should involvement be pursued more? Should brain
hemisphere research and other physiological measures of advertising effect be
pursued more in relation to the hierarchy notion? To what extent, and under
what conditions, should consumer information processing be predominate over
attitude change or emotions in advertising? Can advertisers control or even
286 Reviews of Selected Areas
(insert name of object to be judged
important
of no concern
irrelevant
means a lot to mc
Useless
Wluablc
trivial
bcncficikd
matters to mc
unintcrcstcd
significant
vital
boring
unexciting
appcaIing
lnundilnc
essential
undesirable
Wntcd
not needed
--:--: --:--:--:-- :--
--:--: --:--: --:_: --
--:--:--:--:--:--:--
--:-- :_:_ :_:_ .--
--:--: --:--: --:--: --
--:--:--:-- :--:--:--
--:--:-- :--:--:--_ :_
--:--:-- :--:--:-- :_
--:--:-- :--:--:_ :_
--:--:_:-- :--:--:--
--:--: --:--: --:--: --
--:--:-- :--:--:_ :_
--:--: --:--: --_ :_: --
--:--:-- :--:--:-- :--
--:--: --:--:--:-- :--
--:--: --:--:--:-- :--
--:--: --:--:--:-- :--
--:--:--:--:--:--:--
--:--:-- :--:--:-- :--
uninlportant*
of concern to mc
relevant
means nothing to rile*
Uscfhl
wortldcss*
fhndamcntal
not l)cncficial*
docsn't nmtter*
interested
insignikmt*
sup(Xfluous*
interesting
exciting
unappealing*
fascinating
nonessential*
desiral)le
unwwltc Cl*
necclecl
*Indicates item is reverse scored,
Items on the left are scored (1) low involvemeilt to (7) high involvelnent o[l the right.
Totaling the 20 items gives a score from a low of 20 to a high of 140,
Source: Zaichkowsky, Judith Lynne (1985). "Measuring the involvement Construct. " Journal of
Consumer Research 12 (Dec.): 350.
Figure 5. A Scalc for Nlcasuring Involvement
influence the hierarchy? And, finally, is a broader, more-encompassing hier-
archy model necessary to account for all of these issues of emotion,
involvement, and single and multiple hierarchies? There are no simple answers
to these questions. As history conveys, it is clear that all of these investigations
have added to the debate of the hierarchy concept and most assuredly to the
development and advancement of advertising and marketing practice, Both the
rigor of academic research and the relevance of practitioner research [Brinberg
and Hirschman 1986] are evident in hierarchy of effects research. The critical
position of the hierarchy in both of these worlds must remain as a catalyst for
researchers to continue its advancement in the social and behavioral sciences in
general and marketing and advertising in particular. Cn
a
The Hierarchy of Effects 287
NOTES
1. Proponents of the traditional hierarchy framework claim that people must respond to advertis-
ing (or selling) messages cognitively first; affectively, next; and conatively, last. Thus, the
proper sequence of responses is awareness-attitude-behavior.
2. The marketing and advertising literature generally attributes the development of AIDA to
Edward K. Strong, "Theories of Selling," The Journal of Applied Psychology, 1925.
REFERENCES
Asker, David A. and Carman, James M. (1982). "Are You overadvertising?'"~ournal ofAduertising
Research 22 (Aug.lSept.): 57-70.
Ackoff, Russell L. and Emshoff, James R. (1975). "Advertising Research at Anheuser-Busch, Inc.
(1963-68)." Sloan Management Review (Winter): 1-15.
Adams, Henry Foster (1916). Advertising and Its Mental Luws. New York: The Macmillan
Company.
Advertising Research Foundation (1961). Audience Concepts Committee. Toward Better Media
Comparisons, New York.
Anderson, Robert L. and Barry, Thomas E, (1979). Advertising Management: Text and Cases.
Columbus, OH: Merrill Publishing Company.
Aspinwall, Leo V. (1964). "Consumer Acceptance Theory. " In Theory in Marketing, eds. Reavis .
Cox, Wroe Alderson and Stanley J. Shapiro. Homewood, IL: Richard D. Irwin.
Assael, Henry (1981), Consumer Behauior and Marketing Action. Boston: Kent Publishing
Company, 75-98.
and Day, George S. (1968). "Attitudes and Awareness as Predictors of Marketing Share. "
]ournal of Advertising Research 8 (Oct.): 3-10.
Assmus, Gert, Farley, John U. and Lehmann, Donald R. (1984). "HOW Advertising Affects Sales:
Mets-Analysis of Econometric Results. " Journal of Marketing Research 21 (Feb.): 65-74.
Bass, Frank M. (1974). "'Profit and the A/S Ratio. "Journa/ ofAduertising Research 14 (Dec.): 9-19.
and Leone, Robert P. (1983). "Temporal Aggregation, the Data Interval Bias, and Empiri-
cal Estimation of Bi-monthly Relations from Annual Data. " Management Science 29 (Jan. ): 1-11,
Batra, Rajeev (1986). "Affective Advertising: Role, Processes, and Measurement." In The Role of
Affect in Consumer Behavior: Emerging Theories and Applications, eds, Robert A. Peterson,
Wayne D. Hoyer and William R, Wilson. Lexington, Mass.: Lexington Books, 53-85.
and Vanhonacker, Wilfried R. (1986). "The `Hierarchy of Advertising Effects': An Aggre-
gate Field Test of Temporal Precedence. " New York: Columbia Business School, Avis
Rent-a-Car System Working Paper Series in Marketing (March).
Bedell, Clyde (1940). How to Write Advertising That Sells. New York: McGraw-Hill Book
Company, Inc.
288 fleviews of Selected Areas
I
Bethnan, James R. (1979). `"Memory Factors in Consumer Choice: A Review. " Jouma/ of Market-
ing 43, no, 2:37-53.
Bloch, Peter H. and Richins, Marsha L, (1983). "A Theoretical Model for the Study of Product
Importance Perceptions. " journal of Marketing 47 (Summer): 6%81.
Bourgeois, Jacques C. and Barnes, James C. (1979). "Does Advertising Increase Alcohol
Consumption?" Journal of Advertising Research 19 (Aug.): 1%29.
Boyd, Harper W., Jr., Day, Michael L., and Strong, Edward C. (1972). "An Attitudinal
Framework for Advertising Strategy, " Journal of Marketing 36 (April): 27-33.
Brinberg, David and Hirschman, Elizabeth C. (1986), "Multiple Orientations for the Conduct of
Marketing Research: An Analysis of the AcademicfPractitioner Distinction. " Journa/ of Market-
ing 50 (Oct.): 161-73.
Buzzell, Robert D. (1964). "'Predicting Short-Term Changes in Market Share as a Function of
Advertising Strategy. " Journal of Marketing Research 1 (Aug.): 2741.
_ , Kolin, Marshall and Murphy, Malcolm P. (1965). "Television Commercial Test Scores and
Short-Term Changes in Market Shares. " journal of Marketing Research 2 (Aug.): 307-13.
Cacioppo, John T., Losch, Mary E., Tassinary, Louis G. and Petty, Richard E. (1986). "Properties
of Affect and Af3ect-laden Information Processing as Viewed through the Facial Response
System. " In The Role of Affect in Consumer Behauior: Emerging Theories and Applications, eds.
Robert A. Peterson, Wayne D. Hoyer, and William R. Wilson. Lexington, Mass.: Lexington
Books, 87-118.
Callahan, Francis X. (1982). "Advertising and Profits: 1969- 1978." Journal of Advertising Research
22 (April/May): 17-22.
Campbell, Roy H. (1969). Measuring the Sales and Profit Results of Advertising: A Managerial
Approach, Association of National Advertisers.
Clarke, Darral G. (1976). "Econometric Measurement of the Duration of Advertising Effect on
Sales. " Journal Of Marketing Research 13 (Nov.): 345-57.
Cobb, Cathy J, and Hoyer, Wayne D. (1985). "The Influence of Advertising at the Moment of
Brand Choice, " Journal of Advertising 14, no. 4:5-12, 27.
Cohen, Joel B. and Houston, Michael J, (1972). "Cognitive Consequences of Brand Loyalty."
Journal of Marketing Research 9 (Feb.): 97-99.
Coney, Russell H. (1961). Defining Advertising Goals for Measured Advertising Results. New
York: Association of National Advertisers.
Comanor, William S. and Wilson, Thomas A. (1967). "Advertising Market Structure and
Performance. " Reuiew of Economic Statistics 49 (Nov.): 423-40.
Copland, Brian (1963). "An Evaluation of Conceptual Frameworks for Measuring Advertising
Results. " Proceedings, 9th Annual Conference, Advertising Research Foundation.
Crespi, Irving (1971). "What Kinds of Attitude Measures are Predictive of Behavior?" Public
Opinion Quarterly 35 (Fall): 327-34.
DeBruicker, F, Stewart (1979). "An Appraisal of Low-Involvement Consumer Information
Processimz. " In Attitude Research Plays for High Stakes, eds. John C. Maloney and Bernard
Silverman. Chicago, American Marketing Association, 112-30.
Devoe, Merrill (1956). Effectiue Advertising Copy. New York: The Macmillan Company.
cl'!
~. -h
IImtman, Ceol ge R. (1916). Psychology of Salesmanship. New York: A. F. Fenno and Dayton, OH:
Service l'ublishing.
The Hierarchy of Effects 289
Ehrenberg, Andrew S. C. (1974). "Repetitive Advertising and the Consumer,'' Journal of Advertis-
ing Research 14 (April): 2%34.
Fazio, Russell H, and Zanna, Mark P. (1978). `<On the Predictive Validity of Attitudes: The Roles of
Direct Experience and Confidence, " Journal of Personality 46: 22&43.
-- (1981). "Direct Experience and Attitude-Behavior Consistency." In Aduances in Ex-
perimental Social Psychology, ed. Leonard Berkowitz. New York: Academic Press 14, 161­202,
Finn, David W. (1984), "g
The Integrated Information Response Model. "Journa/ of Advertising 13,
no. 1: 24-33.
Fishbein, Martin (1966). `(
The Relationship Between Beliefs, Attitude, and Behavior. " In Cognitiue
Consistency, ed. Shel Feldman. New York: Academic Press, 199-223.
and Ajzen, Icek (1975). Belief, Attitude, intention and Behauior. Reading, MA:
Addison-Wesley.
Fothergill, J. E. and Ehrenberg, Andrew S, C, (1965). "On the Schwerin Analyses of Advertising
Effectiveness. " Journal of Marketing Research 2 (Aug.): 298-306,
Gibson, Lawrence D,' (1983). ""Not Recall', " Journal of Advertising Research 23 (Feb,/March):
39-46,
Ginter, James L. (1974). `"An Experimental Investigation of Attitude Change and Choice of a NewI
Brand. " Journal of Marketing Research 11 (Feb.): 30-40.
Hall, Samuel R, (1915). Writing an Advertisement: Analysis of the Methods and Mental Processes
that Play a Part in the Writing of Successful Advertising. Boston: Houghton Mifflin Co,
Haskins, Jack B. (1964), "Factual Recall as a Measure of Advertising Effectiveness. " Journal of
Advertising Research 4 (March): 2-8.
Holak, Susan L. and Reddy, Srinivas K. (1986). `"Effects of a Television and Radio Advertising Ban:
A Study of the Cigarette Industry, " Journal of Marketing 50 (Oct.): 219-27.
Holbrook, Morris B. (1975). "A Review of Advertising Research. " In Advertising and the Public
Interest, eds. John A, Howard and James Hulbert, Appendix Two, Chicago: Crain.
-- (1978), "Beyond Attitude Structure: Toward the Informational Determinants of Attitudes. "
Journal of Marketing Research 15 (Nov.): 54%56.
-- (1986). "Emotion in the Consumption Experience: Toward a New Model of the Human
Consumer. " In The Role of Affect in Consumer Behauior: Emerging Theories and Applications,
eds, Robert A. Peterson, Wayne D. Hoyer, and William R. Wilson. Lexington, Mass.: Lexington
Books, 17-52.
Holman, Rebecca H. (1986). "Advertising and Emotionality." In The Role of Affect in Consumer
Behavior: Emerging Theories and Applications, eds. Robert A. Peterson, Wayne D. Hoyer, and
William R. Wilson. Lexington, Mass.: Lexington Books, 119-40.
Howard, John A. and Sheth, Jagdish N, (1969). The Theory of Buyer Behavior. New York: John
Wiley & Sons, Inc.
Hughes, G. David and Guerrero, Jose L, (1971). "Testing Cognitive Models Through
Computer-Controlled Experiments." Journal of Marketing Research 8 (Aug.): 291-97.
Insko, Chester A. and Schopler, John (1967). "Triadic Consistency: A Statement of
AfTective-Cognitive-Conative Consistency. " Psychological Review 74 (Sept.): 361-76.
International Correspondence Schools (191 1). The Conducting of Selling,
Jenkins, John G. (1935). Psychology in Business and Industy, New York: John Wiley& Sons, Inc.
29~ Reviews of Selected Areas
Kapferer, Jean-Noel and Laurent, Gilles (1985-86). "Consumer Involvement Profiles: A New
Practical Approach to Consumer Involvement. " Journal of Advertising Research 25 (Dee./Jan.):
48-56,
Kitson, Harry D. (1921). The Mind of the Buyer: A Psychology of Selling. New York: The Macmil-
lan Company.
Korgaonkar, Pradeep K. and Moschis, George P. (1982). "An Experimental Study of Cognitive
Dissonance, Product Involvement, Expectations, Performance and Consumer Judgment of
Product Performance. " ]ourmd of Advertising 11, no, 3: 3&i4.
Kreshel, Peggy J. (1984). `"Emotion and Advertising: The `Grande Idee' of the 80's." Unpublished
Advertising Working Paper, No. 17, University of Illinois at Urbana-Champaign (Nov.).
Krugman, Herbert E. (1965). "The Impact of Television Advertising: Learning Without
Involvement. " Public Opinion Quarterh.j 29, no, 3 (Fall): 349-56.
_ (1966). "Answering Some Unanswered Questions in Measuring Advertising Effectiveness, `"
Proceedings, 12th Annual Meeting, Advertising Research Foundation, New York, 18-23.
-- (1966-67). "The Measurement of Advertising Without Involvement." Public Opinion
Quarterly 30 (Winter): 585-96.
_ (1972). "Why Three Exposures May Be Enough." Journal of Advertising Research 12
(Dec.): 11-14.
-- (1975). "What Makes Advertising Effective?" Haruard Business Reuiew 53 (March/April):
96-103.
-- (1977). "Memory Without Recall, Exposure Without Perception." Journal of Advertising
Research 17 (Aug.): 7-12.
-- (1985). `"Point of View: Measuring Memory-An Industry Dilemma." journal of Aduertis-
ing Research 25 (Aug,/Sept. ): 49-51.
-- (1986). "'Low Recall and High Recognition of Advertising.''Journa/ ofAduertising Research
(Feb,/March): 79-86.
Lambin, Jean-Jacques (1975). "What Is the Real Impact of Advertising?" Haruard Business Reuiew
53 (May-June): 13947.
Lastovicka, John L. and Gardner, David M. (1979). "Components of Involvement. " In Attitude
Research Pk-Jys for High Stakes, eds, John C. Maloney and Bernard Silverman. Chicago, Ameri-
can Marketing Association, 53-73.
Lavidge, Robert C, and Steiner, (h-y A. (1961). "A Model for Predictive Measurements of Adver-
tising Effectiveness." Journal of Marketing 25 (Oct.): 59-62.
Lazarus, Richard S. (1981), "A Cognitivist"s Reply to Zajonc on Emotion and Cognition. " American
Psychologist 36:22243.
-- (1982). "Thoughts on the Relations Between Emotion and Cognition." American psycholo-
gist 37, no. 9 (Sept.): 1019-24.
-- (1984). "On the Primacy of Cognition. " American Psychologist 39, no. 2 (Feb.): 124-29.
Leckenby, John D. and Wedding, Nugent (1982). Advertising Management: Criteria, Analysis and
Decision Making. Columbus, OH: Grid Publishing, Inc.
and Kishi, Shizue (1984). "The Dirichlet Multinominal Distribution as a Magazine Exposure
Model. " Journal of Marketing Research 21 (Feb.): 100-6.
P :*
The Hierarchy of Effects 291
Leeflang, Peter S. H. and Reuijl, Jan C. (1985). "Advertising and Industry Sales: An Empirical
Study of the West German Cigarette Market. " Journal of Marketing 49 (Fall): 92-98.
Longman, Kenneth A. (1971). Advertising. New York: Harcourt Brace, 255.
MacKenzie, Scott B., Lutz, Richard J., and Belch, George E. (1986). `The Role of Attitude Toward
the Ad as a Mediator of Advertising Effectiveness: A Test of Competing Explanations, "Journal of
Marketing Research 23 (May): 1M3.
Mandler, George (1982). `The Structure of Value: Accounting for Taste. " In Cognition and Affect,
eds. Susan T. Fiske and Margaret S. Clarke. New York: Academic Press.
McGuire, William J. (1969). "An Information-Processing Model of Advertising Effectiveness."
Paper presented to Symposium on Behavioral and Management Science in Marketing, Universi-
ty of Chicago.
_ (1978). "An Information-Processing Model of Advertising Effectiveness." In Behavioral and
Management Science in Marketing, eds. Harry L. Davis and Alvin H. Silk. New York: Ronald
Press, 156-80.
Mendelssohn, Harold (1962). "Measuring the Process of Communications Effect. " Public Opinion
Quarterly 26 (Fall): 411-16.
Metwally, M. M. (1980). "Sales Response to Advertising of Eight Australian Products. `" ]ournd of
Advertising Research 20 (Oct.): 59-64.
Moriarty, Sandra Ernst (1983). "Beyond the Hierarchy of Effects: A Conceptual Framework."
Current Issues & Research in Advertising 1983, 45-55.
Naples, Michael J. (1979). Effectiue Frequency: The ReZatwnship Between Frequency and Adver-
tising Effectiveness. New York: Association of National Advertisers, Inc.
O'Brien, Terrence (1971). "Stages of Consumer Decision Making.'" Journal of Marketing Research
8 (Aug.): 282-89.
Olson, Jerry C., Toy, Daniel R., and Dover, Philip A. (1982). `"Do Cognitive Responses Mediate
the Effects of Advertising Content on Cognitive Structure?" Journal of Consumer Research 9
(Dec.): 245-62.
Osborn, Alexander F. (1922). A Short Cow-se in Aduetiising. New York: C. Scribners Sons.
Palda, Kristian S. (1964). The Mea.surermmt of Cumulative Advertising Effects. Englewood Cliffs,
NJ: Prentice-Hall, Inc.
_ (1966). `<The Hypothesis of a Hierarchy of Effects: A Partial Evaluation.''Journal of Market-
ing Research 3 (Feb.): 13-24.
Park, C. Whan and Young, S. Mark (1986). "Consumer Response to Television Commercials: The
Impact of Involvement and Background Music on Brand Attitude Formation. " Journal of
Marketing Research 23 (Feb.): 11-24.
Peterson, Robert A., Hoyer, Wayne D., and Wilson, William R. (1986). "Reflections on the Role of
AfFect in Consumer Behavior." In The Role of Affect in Consumer Behauwr: Emerging Theories
and Applications, eds. Robert A. Peterson, Wayne D. Hoyer, and William R. Wilson.
Lexington, Mass.: Lexington Books, 141-59.
Petty, Richard E. (1980). "The Role of Cognitive Responses in Attitude Change Processes. " In
Cognitiue Responses in Persuasion, eds. Richard E, Petty, Thomas M. Ostrom and Timothy C.
Brock. Hillsdale, NJ: Erlbaum, 13%39.
and Cacioppo, John T. (1981). Attitudes
Approaches. Dubuque, IA: William C. Brown.
and Persuasion: Classical and Contemporary
292 Reviews of Selected Areas
and Schumann, David (1983), "'Central and Peripheral Routes to Advertising Effectiveness:
The Moderating Role of Involvement. " Journal of Consumer Research 10 (Sept.): 13546.
Poffenberger, Albert T, (1925). Psychology in Advertising. Chicago: A. W. Shaw Company.
Preston, Ivan L, (1982). "The Association Model of the Advertising Communication Process. "
Journal of Advertising 11, no. 2:3-15.
and Thorson, Esther (1983). "Challenges to the Use of Hierarchy Models in Predicting
Advertising Effectiveness. " Proceedings of the Annual Convention of American Acao%my of
Aduertbing, 27-33. ~ ~ ~
_ (1984). `"The Expanded Association Model: Keeping the Hierarchy Concept Alive.'' Journa/
of Advertising Research 24 (Feb.lMarch): 59-65.
Printer's Ink (1910). (Dec. 1): 74.
Ramond, Charles (1976). Advertising Research: The State of the Art. New York: Association of
National Advertisers, Inc., 1976.
Ramsay, Robert E. (1921), Effective Direct Advertising: The Principles and Practice of Producing
Direct Advertising for Distribution by Mail or Otherwise. New York: Appleton and Company.
Rae, Ambar C. and Miller, Peter B. (1975). "Advertisin~Sales Response Functions. " Journal of
Advertising Research 15 (April): 7-15.
Ray, Michael L. (1974). `"Consumer Initial Processing: Definitions, Issues, and Applications, " In
Buyer/Consumer Information Processing, eds. G, David Hughes and Michael L. Ray. Chapel
Hill, NC: University of North Carolina Press, 14S.56.
-- , Sawyer, Alan G., Rothschild, Michael L., Heeler, Roger M., Strong, Edward C., and
Reed, Jerome B. (1973). "Marketing Communications and the Hierarchy of Effects. " In New
Models for Mass Communication Research, ed. Peter Clarke. Beverly Hills, CA: Sage
Publishing, 147-76.
Robertson, Thomas S,, (1971). Innovative Behavior and Communication. New York: HoIt,
Rinehart, Winston.
Rogers, Everett iM. (1962). Diffusion of innovation. New York: Free Press.
Rohloff, Albert C. (1966). "Quantitative Analyses of the Effectiveness of TV Commercials, " ]ournaz
of Marketing Research 3 (Aug.): 239-45.
Rokeach, Milton (1966-67). `<Attitude Change and Behavioral Change. " public Opinion Quarterly
30, no. 4 (Winter): 529-50.
Rossiter, John R. and Percy, Larry (1980). "Attitude Change Through Visual Imagery in
Advertising.'' .Journal of Advertising 9, no. 2:10-16.
Rothschild, Michael I+ (1979). "Advertising Strategies for High and Low Involvement Situations. "
In Attitude Research Plays for High Stakes, eds. John C. Maloney and Bernard Silverman.
Chicago, American Marketing Association, 74-93.
and Caidis, William C. (1981). "Behavioral Learning Theory: Its Relevance to Marketing
and Promotions. " Journal of Marketing 45 (Spring): 70-78.
Rust, Roland T. and Robert P. (1984). "'The Mixed-Media Dirichlet Multinominal Distribution: A
Model for Evaluating Television--Magazine Advertising Schedules. " Journal of Marketing Re-
search 21 (Feb.): 89-99.
a
and Henderson, Pamela W. (1985). "Should Emotional Advertising Response Models Re-
flect Brain Physiology?" Proceedings of the American Academy of Advertising, NR204-7.
C3
The Ilierarchy of Effects 293
Sandage, Charles S, and Fryburger, Vernon F. (1967). Advertising Theoy and Practice, 7th ed,
New York: McGraw-Hill.
Schmalensee, Richard (1972), The Economics of Advertising, Amsterdam: Amsterdam-London
Publication.
Schwartz, David A. (1969), "Measuring the Effectiveness of Your Company's Advertising. "Journal
of Marketing 33 (April): 20-25.
Schweiger, Cunter C. and Hruschka, Harald (1980). `"Analysis of Advertising Inquiries, "Journal of
Advertising Research 10 (Oct.): 37-39.
Scott, Walter D. (1903). The Theoy of Advertising: A Simple Exposition of the Principles of
Psychology in Their Relation to Successful Advertising. Boston: Small, Maynard and Company.
-- (1908). The Psychology of Advertising: A Simple Exposition of the Principles of Psychology
in Their Relation to Successful Advertising. Boston: Small, Maynard and Company.
Sheldon, Arthur Frederick (1911). The Art of Selling. Chicago: The Sheldon School.
Sherif, Carolyn W., Sherif, Muzifer, and Nebergall, Roger E. (1965). Attitude and Attitude
Change. Philadelphia: Saunders.
Simon, Julian L. and Arndt, Johan (1980). "The Shape of the Advertising Response Function. "
Journal of Advertising Research 20 (Aug.): 11-28.
Singh, Surendra N, and Rothschild, Michael L. (1983). "Recognition as a Measure of Learning from
Television Commercials. " Journal of Marketing Research 20 (Aug.): 235-48.
Slama, Mark E. and Tashchian, Armen (1985). "Selected Socioeconomic and Demographic Charac-
teristics Associated with Purchasing Involvement. `" Journal of Marketing 49 (Winter): 72-82.
Smith, Robert E. and Swinyard, William R. (1982), "Information Response Models: An Integrated
Approach, " Journal of Marketing 46 (Winter): 81­93.
-- (1983). "Attitude-Behavior Consistency: The Impact of product Trial Versus Advertising."'
Journal of Marketing Research 20 (Aug.): 257-67.
Starch, Daniel (1923). Principles of Advertising. New York: McGraw-Hill Book Company.
Stewart, David W. (1986). "The Moderating Role of Recall, Comprehension, and Brand Dif-
ferentiation on the Persuasiveness of Television Advertising. "Journal of Advertising Research 26
(AprillMay): 4W7,
Strong, Edward K., Jr. (1922), The Psychology of Selling Life Insurance. New York: Harper &
Brothers Publishers,
-- (1925a). The Psychology of Selling and Advertising. New York: McCraw-Hill.
-- (1925b). "Theories of Selling." Journal of Applied Psychology 9:75-86.
-- (1938). Psychological Aspect of Business. New York: McGraw-Hill Book Company, Inc.
Tauber, Edward M. (1982-83), "How Does Advertising Work?'' Journal of Advertising Research 22
(Dee./Jan.): 7.
Traylor, Mark B. (1981a). "Product Involvement and Brand Commitment,'' Journal ofAduertising
Research 21 (Dec.): 51­56,
-- (1981b). "Ego Involvement and Brand Commitment: Not Necessarily the Same." Journal
of Consumer Marketing 1, no. 1:75-79.
Tsal, Yehoshua (1985). "On the Relationship Between Cognitive and Affective Processes: A Cri-
tique of Zajonc and Mw-kus. " Journal of Consumer Research 12 (Dec.): 358-62.
294 Reviews of Sekcted Areas
Tull, Donald S , Wood, Van R., Duhan, Dale, Gillpatrick, Tom, Robertson, Kim R. and Helgeson,
James C, (1986). ""Leveraged' Decision Making in Advertising: The Flat Maximum Principle and
Its Implications." ~ournal of Marketing Research 23 (Feb.): 25-32,
Tyebjee, Tyzoon T, (1979), "Refinement of the Involvement Concept: An Advertising Planning
Point of View. " In Attitude Research PZaysfor High Stakes, eds. John C, Maloney and Bernard
Silverman. Chicago, American Marketing Association, 94-111.
Van Raaij, W. Fred (1984). A~~ectiue and Cognitive Reactions to Advertising. Report No. 84-111,
Marketing Science Institute, Cambridge, Massachusetts.
Vaughn, Richard (1980). `"How Advertising Works: A Planning Model. " Journal of Advertising
Research 20 (Oct.): 27-33.
_ (1986), "How Advertising Works: A Planning Model Revisited." Journal of Advertising
Research 26 (Feb./March): 57-66.
Venkatesan, M. and Haaland, Gordon A, (1968), "Divided Attention and Television Commercials:
An Experimental Study. `" ]ournal of Marketing Research 5 (May): 203-5.
West Coast Life Insurance Company (1920). "The Five Steps of a Sale. "
Wilcox, Gary B., Vacker, Barry and Franke, George R, (1986). "Alcoholic Beverage Advertising
and Consumption in the United States: 1964-1984. " Unpublished Working Paper No. 86-1, The
University cf Texas at Austin,
Wolfe, Harry D., Brown, James K. and Thompson, C, Clarke (1962). Measuring Advertising
Results. New York: National Industrial Conference Board, 7.
Woodside, Arch G. and Wilson, Elizabeth J. (1985). "Effects of Consumer Awareness of Brand
Advertising on Preference.'" Journal of Advertising Research 2S (Aug.lSept.): 41-48.
Wright, Peter (1973). "The Cognitive Processes Mediating Acceptance of Advertising."' Journal of
Marketing Research 10 (Feb.): 53-62.
-- (1975), "Factors ~ecting Cognitive Resistance to Advertising," Jourrud of Consumer Re-
search 2 (June): 1-9.
Zaichkowsky, Judith Lynne (1985). `g
Measuring the Involvement Construct. " Journal ojConsumer
Research 12 (Dec.): 341-52,
(1986). "Conceptualizing Involvement." Journal of Advertising 15, no, 2:4-14,34,
Zajonc, Robert B. (1968). "Attitudinal Effects of Mere Exposure. "Journa/ ofPersonality and Social
Psychology Monograph 9, part 2:1-28.
-- (1980a). "Feeling and Thinking: Preferences Need No Inferences." American Psychologist
35 (Feb.): 151-75.
_ (1980b). "'Cognition and Social Cognition: A Historical Perspective." Reproduced draft
(Nov. 10), for Four Decades of Social Psychology, ed, Leon Festinger. Oxford University Press.
_ (1986). "Basic Mechanisms of Preference Formation. " In The Role of Affect in Consumer
Behauior: Emerging Theories and Applications, eds, Robert A. Peterson, Wayne D. Hoyer and
William R, Wilson, Lexington, Mass.: Lexington Books, 1-16.
and Markus, Hazel (1982). "AKective and Cognitive Factors in Preferences. " Journal of
Consumer Research 9 (Sept.): 12R31.
-- (1985). "'Must
(Dec.): 36$-64,
All AEect Be Mediated by Cognition?" Journal of Consumer Research 12
The Hierarchy of Effects 295
Zaltman, Gerald and Wallendorf, Melanie (1983). Consumer Behavior: Basic Findings and Man-
agement Implications. New York: John Wiley & Sons, chap. 19, "Involvement Theory and
Consumer Behavior" by C. Whan Park and Banwari Mittal, 549-69.
Zielske, Hubert A, (1959). "The Remembering and Forgetting of Advertising. " Journal of Market-
ing 22 (Jan,): 239-43.
Zinkhan, George M. and Zinkhan, F. Christian (1985). "Response Profiles and Choice Behavior: An
Application to Financial Services Advertising. "Journal of Advertising 14, no. 3:3944,51,66.
Zinkhan, George M. and Gelb, Betsy D. (1986). "What Starch Scores Predict.'' Journal ofAdvertis-
ing Research 26 (Aug.lSept.): 45-50.
Zinkhan, George M. and Fornell, C1aes (1986). "A Test of the Learning Hierarchy in High-Low
Involvement Situations. " Paper presented at the University of California at Irvine (Nov.).