You are on page 1of 16

Are Product Attribute Beliefs the Only Mediator of Advertising Effects on Brand Attitude?

Author(s): Andrew A. Mitchell and Jerry C. Olson


Source: Journal of Marketing Research, Vol. 18, No. 3 (Aug., 1981), pp. 318-332
Published by: American Marketing Association
Stable URL: http://www.jstor.org/stable/3150973 .
Accessed: 06/11/2014 07:06

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Marketing Association is collaborating with JSTOR to digitize, preserve and extend access to
Journal of Marketing Research.

http://www.jstor.org

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
I ?

ANDREW A. MITCHELLand JERRYC. OLSON*

Fishbein's attitude theory posits that beliefs are the only mediators of attitude
formation and change. The validity of this proposition for consumers' beliefs
about product attributes and brand attitudes was exmained in the context
of an advertising effects study. To manipulate product attribute beliefs and
to create settings in which other mediation processes might occur, the authors
exposed subjects to simple advertisements that contained either a verbal claim
or visual information. Level of repetition also was varied. As expected, product
attribute beliefs mediated attitude formation. However, another variable,
termed attitude toward the advertisement, also mediated brand attitudes and
purchase intentions. The authors discuss alternative explanations for the results
and offer suggestions for future research.

Are Product Attribute Beliefs the Only


Mediator of Advertising Effects on Brand
Attitude?
I

Attitude, defined here as an individual's internal marketers must develop a clearer understanding of
evaluation of an object such as a branded product, has the causal determinants of attitude formation and
been an important concept in marketing research for change (Olson and Mitchell 1975; also see Lutz 1977;
the past 20 years. There are at least two major reasons Lutz and Bettman 1977). These causal influences are
for this long-term interest. First, attitudes often are of critical importance because they mediate the effects
considered relatively stable and enduring predisposi- of marketing decision variables such as advertising
tions to behave. Consequently, they should be useful or price on consumer attitudes. Thus, a better under-
predictors of consumers' behavior toward a product standing of the causal dynamics of attitude formation
or service. Second, social psychology has provided would not only aid marketing researchers in measuring
several theoretical models of the attitude construct the attitudinal impact of marketing variables, but also
(e.g., Fishbein 1963; McGuire 1968; Rosenberg 1956; help managers develop more effective marketing strat-
Triandis 1971; Wyer 1974). These conceptual frame- egy.
works, especially Fishbein's, have stimulated much
of the attitude research in marketing. Most of this ATTITUDE THEORY
work, however, has been descriptive and pragmatic Fishbein (1963, 1967; Fishbein and Ajzen 1975)
in orientation; theoretical issues have been relatively presented perhaps the clearest theoretical exposition
neglected. of the causal basis of attitudes. According to Fishbein
To increase the usefulness of the attitude construct, and Ajzen (1975, p. 222), "A person's attitude is a
function of his salient beliefs at a given point in time."
Beliefs are the subjective associations between any
two discriminable concepts. Salient beliefs are those
*Andrew A. Mitchell is Associate Professor of Marketing, Carne-
gie-Mellon University. Jerry C. Olson is Associate Professor of
activated from memory and "considered" by the
Marketing, The Pennsylvania State University. The research was person in a given situation (Fishbein and Ajzen 1975;
funded by a Research Initiation grant to both authors from The Olson, Kanwar, and Muderrisoglu 1979). As these
Pennsylvania State University and an NSF grant to the first author. ideas are operationalized in the typical marketing
The authors thank Meryl Gardner and S. P. Raj for their assistance
in data analysis and three reviewers for their constructive comments. research study, the attitude concept of interest is a
brand and the related concepts are product attributes.
318

Journal of Marketing Research


Vol. XVIII (August 1981), 318-32

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
PRODUCT
ATTRIBUTE AND BRANDATTITUDE
BELIEFS 319

That is, marketing researchers have been mainly 6). Beliefs can be modified by changing the strength
concerned with consumers' beliefs about attributes of a salient belief (be), changing the evaluation of
of a brand. a belief (e,), creating a new salient belief, or making
Fishbein's now-familiar attitude model specifies the a salient belief nonsalient (Lutz 1975, 1977). Fishbein's
relationship between the set of salient beliefs about algebraic model (equation 1) then can be used to
a concept (often termed cognitive structure) and an estimate the subsequent effect on overall attitude. Our
overall evaluation of, or attitude toward, the concept. key concern is with the proposition that beliefs cause
n attitudes. Stated in its strongest sense, the presumption
( is that beliefs mediate all effects on attitude formation
(1) b,e, = A,
i=l or attitude change.
where: Developing an empirical test of this causal relation-
ship is straightforward, at least logically. Assume that
b, = the strengthof the associationbetween the attitude a persuasive message has a significant effect on
concept, o, and the ithsalient concept, attitudes. Assume further that we can identify and
e, = the evaluationof the ithsalient concept, measure all salient beliefs. If Fishbein's theory that
Ao = the overall evaluationof, or attitudetoward, con- beliefs mediate effects on attitude is correct, we
cept o, and should find that statistically removing the effect of
n = the numberof salient beliefs. the message on beliefs (e.g., by analysis of covariance
procedures) also removes the significant message
Fishbein clearly intended the attitude model in equa- effect on attitude. To the extent that the message
tion 1 to describe only the predicted relationship still has significant effects on attitude, other causal
between measures of the theoretical constructs (e.g., processes must have occurred (or else measurement
see Fishbein and Ajzen 1975, p. 222-3). That is, the errors, etc., are present). In reality, problems arise
algebraic model provides a means of estimating the because empirical data cannot always be interpreted
belief-attitude relationship that was formed as a result as rigorous tests of the theory (cf. Carnegie-Mellon
of causal mechanisms. The model itself does not University Marketing Seminar 1978; Dickson and
"state" the causal proposition nor does it specify the Miniard 1978; Lutz 1978a, b). Perhaps the most diffi-
causal mechanisms. cult problem is demonstrating that all of the salient
Fishbein also proposed that the attitude-belief rela- beliefs, and only the salient beliefs, were measured.
tionship specified in equation 1 holds for attitudes Another problem is unambiguously demonstrating the
toward a specific behavior such as buying a product, temporal ordering of the effects. Perhaps because of
Aact. However, the set of salient beliefs may not be these difficulties, very few studies in either psychology
the same as for Ao (Fishbein and Ajzen 1975). In or marketing have addressed the validity of the basic
turn, attitude, especially Aact, is presumed to have proposition that beliefs are the sole determinant of
a causal influence on behavioral intentions (BI).' In attitudes. That issue motivated the research reported
sum, Fishbein proposed a causal flow among three here.
cognitive variables-beliefs, evaluations or attitudes, PAST A TTITUDE RESEARCH IN MARKETING
and intentions (see Lutz 1977). According to this view,
a marketing stimulus such as an advertisement affects Much of the marketing research based on the Fish-
consumers' beliefs first. Then the influenced salient bein model has focused on demonstrating statistically
beliefs mediate the marketing variable's effect on significant correlations between a "direct" measure
attitude, and attitude in turn mediates subsequent of brand attitude (e.g., A,) and the cognitive structure
effects on behavioral intention. index of attitude (Ebiei) based on product attribute
The basic theoretical proposition of Fishbein's atti- beliefs (see Wilkie and Pessemier 1973). Most of the
tude theory is that beliefs cause attitude. Because studies obtained statistically significant correlations.
attitude is determined by a set of salient beliefs, Moreover, when the model constructs were measured
changes in attitude must be mediated by changes in correctly (Cohen, Fishbein, and Ahtola 1972), the
those beliefs. Therefore, to change a person's attitude resulting correlations were reasonably high, ranging
toward a concept, one must modify the salient beliefs from .40 to .70 (e.g., Lutz 1977). Considerable research
about that concept (Fishbein and Ajzen 1975, Ch. effort has been directed at determining whether various
modifications of the basic model are empirically "bet-
ter" than the Fishbein formulation, even though the
'Actually, Fishbein'sextended attitudemodel assumes that be- appropriate criteria for selecting one model over an-
havioralintentions are a function of both A^,c and a social norm other are somewhat ambiguous. In general, this re-
construct (Fishbein and Ajzen 1975). Recent research findings, search has emphasized measurement issues and pre-
however, raise the question of whether the social norm element dictive validity (Lutz and Bettman 1977).
is really a separateconstruct(Miniardand Cohen 1979). Because
of these operationaland conceptual problems, the social norm Recently, other interests related to Fishbein's atti-
constructis not includedin our research. tude model have begun to emerge. Several researchers

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
320 JOURNALOF MARKETINGRESEARCH,AUGUST 1981

have suggested that this model provides a theoretically research suggested that repeated exposure enhances
integrated set of measures of the cognitive effects attitudes, there has been considerable controversy
of marketing variables, particularly advertising (Lutz about the necessary conditions for these effects (e.g.,
1975; Mazis and Adkinson 1976; Olson and Mitchell Matlin 1970; Stang 1975) and their theoretical basis
1975). In applying this approach, beliefs about attri- (e.g., Berlyne 1970; Harrison 1968; Sawyer 1977).
butes of the advertised brand, brand attitudes, and However, substantial evidence supports their exis-
purchase intentions are measured to indicate the mul- tence (Harrison 1977). Recently, Wilson (1979) dem-
tiple effects of a particular communication message onstrated that repetition affects attitude even when
on cognitive structure variables (e.g., Olson and Dover subjects are unaware of being exposed to the stimulus.
1978b). The measures of brand attribute beliefs seem On the basis of such evidence, Zajonc (1980) suggested
particularly useful in providing diagnostic information that attitudes may precede beliefs in certain situations;
about the effectiveness of a message strategy. that is, the causal flow may be reversed. Briefly stated,
A second emerging interest of a few researchers the notion is that attitude formation apparently can
is the theoretical basis for attitude (cf. Calder 1975, occur without belief formation under certain condi-
1978; Holbrook 1978; Lutz 1975; Lutz and Swasy 1977; tions of repeated exposure.
Mitchell, in press; Olson and Dover 1978a; Olson and The other alternative proposition investigated is
Mitchell 1975). Although massive correlational evi- derived from the classical conditioning approach to
dence supports the static relationship between beliefs attitude formation (e.g., Staats and Staats 1967). Ac-
and attitudes (cf. Fishbein and Ajzen 1975; Wilkie cording to this perspective, attitudes may be formed
and Pessemier 1973), experimental evidence support- by repeatedly pairing a neutral (unconditioned) stimu-
ing the causal, mediating effect of beliefs in attitude lus with a positively or negatively evaluated (condi-
formation and change is meager (Lutz and Bettman tioned) stimulus. For instance, an unknown brand
1977). Lutz (1975), for instance, demonstrated that name (UCS) could be paired in an advertisement with
manipulations of product attribute belief strength were a picture (CS) that evokes positive feelings. The
mirrored by corresponding shifts in Ao. Olson and association might cause the positive feelings evoked
Dover (1976) created changes in the strength of product by the picture to become conditioned to the brand
attribute beliefs that were consistent with changes in name.
brand attitudes. However, few other studies have Both alternative propositions identify conditions
examined the causal influence of product attribute under which attitudes may be formed or changed
beliefs on brand attitudes. In fact, relatively few without forming or changing beliefs. Our study was
marketing researchers have even included beliefs, designed to create the conditions necessary to test
attitudes, and intentions as multiple dependent vari- whether brand attribute beliefs or the causal factors
ables in the same study. suggested by the alternative propositions mediate
attitude formation. We manipulated two variables: (1)
RESEA R CH PURPOSE
repetition-the number of times a particular adver-
The purpose of our study was to examine the validity tisement was repeated, and (2) advertising content-
of the basic theoretical proposition of the Fishbein whether the information presented in each ad was
attitude theory in a marketing research context. Spe- a verbal claim about a product attribute or merely
cifically, we were interested in whether beliefs about a pairing of a brand name with one of three visual
product attributes are the only mediator of brand stimuli. By repeatedly exposing subjects to different
attitude. The experiment involved exposing subjects types of information, we created conditions in which
to different advertisements intended to create different attitudes could have been formed through mechanisms
brand attitudes. Because the brands were fictitious, other than belief formation. That is, we experimentally
consumers had no prior brand beliefs or attitudes. created opportunities for a disconfirmation of the
Thus, the research concerned only belief and attitude proposition that beliefs are the sole mediator of atti-
formation, not change. If Fishbein's theory is correct, tudes. For instance, if advertising content creates
the advertisements should also create differences in brand attitudes without parallel effects on product
brand attribute beliefs. Moreover, the variation in beliefs, such results would constitute a "strong"
these beliefs should account for (i.e., mediate) all the disconfirmation of the beliefs-cause-attitudes proposi-
experimentally produced variation in brand attitudes. tion.
To enhance the inferential logic of the study (Platt In another article based on this research (Mitchell
1964), we also examined two alternatives to the be- and Olson 1977), we presented a theoretical discussion
liefs-cause-attitude model. Both imply that an attitude of repetition effects and the results of a preliminary
toward a concept may be formed (or changed) without data analysis.2 The analysis reported here produced
the corresponding formation of salient beliefs about
the concept (or changes in those beliefs). One alterna-
tive proposition suggests that sheer repetition of a In that analysis only the cell means were used as the dependent
stimulus may cause changes in an individual's attitude variable. Because we used all the observations in each cell as
toward that stimulus. Since Zajonc's (1968) original the dependent variable, our analysis is more complete.

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
PRODUCT
ATTRIBUTE
BELIEFS
AND BRANDATTITUDE 321

the same results for repetition as reported previously; Table 1


that is, there were no effects of repetition on any EXPERIMENTAL
DESIGNSHOWINGNUMBER
OF
of the cognitive variables. Thus, the data cannot be REPETITIONS
used to address the question of whether sheer repeti-
tion might "directly" affect attitudes but not beliefs, Brand name/advertising content
because repetition had no effect on brand attitudes
Brand I, Brand R,
or beliefs. Therefore, in this article we are concerned Experi- explicit Brand J, Brand L, picture of
only with the effects of the visual-verbal information mental claim of picture of picture of abstract
manipulation of advertising content. group softness kitten sunset painting
A (n = 21) 4 8 6 2
METHOD B (n = 19) 2 6 8 4
C (n = 17) 6 2 4 8
D(n= 14) 8 4 2 6
Subjects
Subjects were 71 junior and senior undergraduate
students of both sexes who were recruited from an Product
introductory marketing course. Subjects volunteered
to attend one of four 45-minute experimental sessions For several reasons, facial tissue was selected as
in return for a $2.00 payment. As there is no reason the product to be advertised. First, the salient product
to suspect that subjects' choices of an experimental attributes of the product category were well known
session were biased, this procedure is assumed to to the subjects and thus could be determined relatively
have produced a random assignment of subjects to easily. Second, subjects' knowledge structures for
groups. facial tissues seemed rather simple, involving a rea-
sonably small number of salient product attributes.
Unstructured interviews with several secretaries and
Experimental Design students indicated six salient attributes for the facial
To avoid the demand characteristics that might be tissue product class: softness, price, absorbency, tear-
created by requiring subjects to rate the same stimulus ing ease, scent, and color. Finally, it appeared likely
repeatedly, we used an experimental design in which that the softness attribute could be communicated in
reactions were measured only once, after all exposures a pictorial, nonverbal manner as well as through an
to the four different advertisements. The resulting explicit verbal claim.
design was a 4 x 4 Latin square with four groups
of subjects and repeated measures over both treatment Brand Advertisements
factors (Table 1). Advertising content was manipulated For each of the four "brands" of facial tissue,
by four different advertisements, each for a different advertisements were designed to look like mockup,
"brand" of facial tissue. Three advertisements were partially completed print advertisements. To eliminate
nonverbal "image ads," each containing a different the influence of prior brand learning on the measured
picture, and the fourth contained a verbal product effects of repetition and advertising content, we used
claim but no picture. Repetition was manipulated at unfamiliar brand names. The letters I, J, L, and R
four levels-2, 4, 6, and 8 exposures. As illustrated were selected as the "brand names" on the basis
in Table 1, subjects in the four groups were exposed of pilot research indicating that individuals made
to all four advertisements, each advertisement under relatively few associations with these letters. Each
a different level of repetition. In total, each subject advertisement contained a headline and about the same
received 20 exposures divided among four different amount of simulated copy in the form of drawn lines
advertisements. representing rows of words arranged in paragraph form
This design does not allow an ideal examination below the headline. The headlines were set in the
of the belief and attitude formation processes as they same type face and size (e.g., "Brand J Facial Tis-
occur for each individual subject, over time, with sues"). The artwork, simulated copy, and reproduction
successive repetitions. However, it does eliminate the of the finished advertisements on 35-mm slides were
potential reactivity of within-subject designs that re- of professional quality.
quire repeated responses to the same measures. The Of the four advertisements, only one contained an
design provides complete information about the main explicit product claim but no picture. Each of the
effects of group, advertising content, and repetition three nonverbal "image" advertisements contained
level, but only partial information on the interaction only a brand name headline and a half-page color
between advertising content and repetition. Finally, photograph but no product claim. The verbal claim
this design requires the assumption that interactions advertisement and one of the nonverbal advertise-
involving the group factor are negligible. There was ments were intentionally designed to communicate that
no a priori reason to expect treatment x group the brand had the specific attribute of softness. That
interactions because subjects were presumed to have nonverbal advertisement contained a picture of a fluffy
been assigned randomly to groups. kitten, assumed to be a positively evaluated stimulus

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
322 JOURNALOF MARKETINGRESEARCH,AUGUST 1981

and to connote "softness," photographed against a subjects signed an informed consent form. As a
dark background. Below the picture was the headline, rationale for the experiment, subjects were told that
"Brand J Facial Tissues," and the simulated copy. the basic purpose of the study was to investigate
The verbal claim ad contained only the headline, whether different advertisements affected their ability
"Brand I Facial Tissues are Soft," below a half-page to remember brand names. Subjects were told to
blank space where a picture might have been. The concentrate on remembering the brand names and the
other two nonverbal advertisements contained pictures advertisement for each brand. This was done to ensure
considered to be essentially irrelevant to the product. that subjects could associate each "brand name" with
The advertisement for Brand L included a picture the appropriate advertising content, because all post-
of a spectacular sunset over an ocean, which was exposure measures of cognitive structure were brand
assumed to be a positively evaluated stimulus. The specific.
advertisement for Brand R contained a presumably Then subjects were shown 20 advertisements (4
neutral picture of an abstract painting. brand advertisements with either 2, 4, 6 or 8 repeti-
tions), timed so each remained on the screen for 10
Experimental Procedures seconds. The ordering of the advertisements was
After being seated at tables in a large room, subjects randomly determined for each group, except for the
in each group were told that they would see partially stipulation that the same ad not appear consecutively.
completed advertisements for four different brands After the 20 exposures, subjects responded to the
of facial tissues. Then, to meet university regulations, dependent variable measures contained in a question-

Table 2
MAIN EFFECTSOF ADVERTISINGCONTENTON BELIEFSTRENGTH,ATTITUDES,AND INTENTIONS

Brand name/advertising content


marginal means
F-ratio' for Brand I, Brand R,
advertising explicit Brand J, Brand L, picture of
content claim of picture of picture of abstract
Dependent variables main effect softness kitten sunset painting
Belief levels (main e,)
Not at all soft (-2.82) 27.30' 2.49 2.32 2.59 [3.96]
Slightly soft (-0.74) 3.82 4.83 4.51 4.25 4.46
Fairly soft (1.11) 16.58' 5.42 5.31 4.99 [4.45]
Very soft (2.86) 25.83' (4.80) 5.45 4.89 [3.59]
Tears easierb (-1.88) 4.30d 3.99 4.08 3.68 3.46
Tears about as easilyb (0.07) 2.80 5.18 5.06 4.92 4.73
Does not tear as easily (1.58) 5.42d 3.34 3.30 3.68 (3.90)
More absorbentb (2.21) 3.98d 3.87 4.17 3.90 (3.48)
About as absorbentb (0.47) 4.99d 5.21 5.17 4.96 (4.70)
Less absorbentb (-1.56) 1.95 3.49 3.25 3.39 3.68

More economicalb (1.83) 9.35' 3.82 3.28 3.31 (4.18)


About as economicalb (0.56) 2.34 4.79 4.76 4.45 4.82
Less economicalb (-1.25) 4.86d 3.70 4.04 4.25 (3.51)
Comes in more attractive colorsb (0.90) 30.92c [3.04] 4.51 5.11 4.79
Comes in the same colorsb (0.35) 6.46c (5.04) 4.93 4.37 4.34

Attitudes and intentions


Attitude toward brand (A,) 27.43c [3.03] 3.65 3.68 [2.82]
Attitude toward act of purchasing (Aa,,) 14.54c 3.36 3.61 3.61 [2.95]
Behavioral intention to purchase (BI) 15.62c 4.08 4.66 4.81 [3.20]
"d.f. = 3 and 204.
bTo complete description read "than (or as) most other brands."
< .001.
p < .01.
Notes: Interpretation of the Newman-Keuls pairwise comparisons of advertising content means is: (a) means enclosed by [ ] are
significantly different (p < .01) from each of the other three means, (b) means enclosed by ( ) are significantly different
means
(p < .05) from underlined means, (c) underlined means are not significantly different (p > .05) from nonunderlined
or other underlined means.

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
PRODUCT
ATTRIBUTE AND BRANDATTITUDE
BELIEFS 323

naire booklet, then they were debriefed, paid, and were 0.88 and 0.85, respectively, indicating acceptable
dismissed. levels of internal consistency. Behavioral (purchase)
intention toward each brand (BI) was measured on
Dependent Variables a single seven-point bipolar scale anchored by the
The major dependent variables were the various phrases "not at all likely to buy" and "very likely
cognitive elements defined by Fishbein and Ajzen to buy."
(1975) that are commonly measured in consumer Consumer evaluations of each advertisement and
research (e.g., Gardner, Mitchell, and Staelin 1977; the two pictorial symbols were also measured. Near
Lutz 1977; Mazis and Adkinson 1976; Olson and Dover the beginning of the post-exposure questionnaire and
1978b). These included belief strength for the salient before the cognitive structure measures, subjects rated
brand attributes (b,), evaluations of each attribute each advertisement on seven five-point bipolar scales.
belief (e,), attitude toward each brand Ao), attitude A factor analysis of these measures yielded two factors
toward the act of purchasing and using each brand with eigenvalues greater than 1, which together ac-
Aact), and behavioral intention to purchase each brand counted for 68% of the variance. A varimax rotation
(BI). of these factors identified four variables (good-bad,
Because of the importance of precisely measuring like-dislike, irritating-not irritating, interesting-unin-
subjects' attribute beliefs about facial tissues, Ahtola's teresting) that loaded highly on the first, clearly
(1975) vector model modification of Fishbein's basic evaluative, factor. The mean of the four evaluative
attitude model was adapted as the belief measurement measures was interpreted as attitude toward the ad
procedure. In terms of attitude formation, Ahtola's (Aad, coefficient alpha = .87). Later in the question-
vector model is theoretically identical to Fishbein's. naire, after the belief and attitude measures, subjects
However, the Ahtola approach seems to capture more rated the concepts of kittens and sunsets on four
subtle aspects of consumers' belief structures (Ahtola five-point scales (good-bad, like-dislike, pleasant-
1975; Olson and Dover 1978b). To operationalize unpleasant, and nice-awful; coefficient alpha = .85).
Ahtola's model one must know not only the salient The mean of these evaluative measures was interpreted
attributes for a product, but also the levels, categories, as indicating attitude toward the pictorial symbols
or amounts of each attribute dimension that consumers (Apic) portrayed in the nonverbal advertisements.
use to discriminate objects in terms of that attribute.
In the Ahtola model beliefs are measured for each RESULTS
discriminable attribute level, and the set of beliefs Demand Characteristics
about the various levels of an attribute constitutes
Immediately after completing the questionnaire,
a belief vector. The discriminable belief levels for
each salient attribute (listed in Table 2) were derived subjects wrote a brief paragraph describing their
from subjects' comments in the pilot interviews and perceptions of the purpose of the experiment. These
comments indicated that virtually all of the subjects
the authors' intuition.
accepted the explanation given to them. No subjects
Following Ahtola, we measured belief strength for mentioned that the experiment might pertain to the
each attribute level, although our measurement proce- effects of conditioning or of different types of adver-
dure differed from the constant sum method suggested
tising content on the formation of beliefs, attitudes,
by Ahtola (1975). For each of the four brands, re- or intentions. Thus, although demand characteristic
spondents rated each of the 15 attribute levels as to influences cannot be completely ruled out by this rough
how likely that level was to be associated with the
check, at least the basic purposes of the experiment
brand (1-7 scale, very unlikely-very likely). The were not apparent.
evaluation (ei) of each attribute level was measured
on a bipolar scale labeled good-bad (+3 to -3). Brand Name Learning
Because e, values for particular attribute levels are To ensure that the measures of beliefs, attitudes,
generally assumed to be the same for all brands within and intentions for each brand were interpretable, we
a product category, these measures were not brand first had to determine that all subjects had accurately
specific. learned the correct association between the advertising
Attitude toward the brand (A ) was measured by content and the four brand names. Therefore, the
the mean of four five-point evaluative scales (good- first question in the post-experimental questionnaire
bad, dislike very much-like very much, pleasant- asked subjects to match each brand name with the
unpleasant, poor quality-high quality). The mean of correct written description of its advertisement. Four
three five-point evaluative scales (good-bad, foolish- subjects were unable to answer this question correctly
wise, beneficial-harmful) was used as a measure of and were dropped from the analysis along with two
attitude toward the act of purchasing and using the other subjects who failed to complete substantial parts
brand (A.,t). Multiple scales were used to enhance of the questionnaire. The resultant sample sizes were
the reliability of the important attitude measures. 21, 19, 17, and 14 for groups A, B, C, and D,
Coefficient alpha scores for the Ao and Aat measures respectively.

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
324 JOURNALOF MARKETINGRESEARCH,AUGUST 1981

Analyses softness attribute (not at all soft, fairly soft, and very
An ANOVA for a 4 x 4 Latin square design with soft, all p < 0.001). Pairwise tests indicated that for
all three of these attribute levels the beliefs created
repeated measures on both the advertising content
and repetition factors was used to analyze each of by the abstract painting advertisement for Brand R
were significantly different from the beliefs created
the cognitive variables. Following Winer (1971) and
Kirk (1968), we used an estimate of the mean square by the other three advertisements (Newman-Keuls
due to subjects within groups to test the between-sub- tests, p < .01). In addition, the advertisement pairing
the J brand name with the kitten picture created a
jects group factor. Advertising content, repetition, and
their interaction are within-subject factors which were stronger belief that the tissues were very soft (the
most positively evaluated attribute level) than did the
tested against the residual error after the sum of
explicit verbal claim advertisement (5.45 vs. 4.80;
squares due to subjects within groups was removed. Newman-Keuls test, p < .01), perhaps because the
Because the cell sizes were unequal, the advertising
verbal claim stated that Brand I tissues were soft,
content and repetition factors are not orthogonal.
not very soft.
However, the group factor is orthogonal to these two. The advertising content factor also had strong main
As Perreault and Darden (1975) point out, nonorthog-
effects on both belief levels for the color attribute
onality creates problems in testing the significance
of the nonorthogonal effects, because a portion of (p < 0.001). For Brand I in the explicit verbal claim
the total sum of squares will be common to the advertisement, the strength of the positively evaluated
belief, comes in more attractive colors than other
nonorthogonal factors. Perreault and Darden suggest brands, was significantly lower than that for the three
a procedure for hypothesis testing that sequentially
adds the common variance to each nonorthogonal visually advertised brands (Newman-Keuls test, p <
factor. However, this procedure may produce a signif- .01). For the essentially neutral comes-in-the-same-
color attribute level, stronger beliefs were created for
icant effect for a particular factor when the common
the verbal-claim brand than for the brands paired with
variance is included and a nonsignificant effect when
the sunset and abstract paintings.
it is not included. A more conservative procedure
is to test the significance of each nonorthogonal factor Finally, scattered, somewhat weaker effects of
advertising content were found for six of the remaining
using the sum of squares unique only to that factor nine brand attribute beliefs (p < .01). These advertis-
(Timm and Carlson 1975). A preliminary analysis of
the mean scores of each cell had indicated that only ing content effects were due primarily to differences
between the advertisement containing the abstract
the advertising content factor was significant (Mitchell
and Olson 1977). Therefore, only the unique sum of painting and one or more of the other brand advertise-
ments. Thus, Brand R paired with the abstract painting
squares due to advertising content was used to test was perceived to be less absorbent, to not tear as
that effect, and the sum of squares common to both
easily, and to be more economical than the other
nonorthogonal effects was included with the sum of brands. In summary, the advertising content
squares unique to repetition to test that effect. In
manipulation had substantial multiple effects on prod-
essence, this procedure "stacked the deck" against uct attribute beliefs.
finding an advertising content effect and for finding
a repetition effect.
Effects on Attribute Evaluations
Fishbein's attitude theory proposes that attitudes
Effects on Belief Strength are a function of belief strength and evaluation. Thus,
Using these procedures, we ran ANOVAs on the whether or not the advertising content manipulation
belief strength scores for each of the salient beliefs had any effects on the 15 attribute evaluation (ei)
about facial tissue (i.e., the 15 attribute levels for scores is of interest. Our design, however, is not
the five belief vectors, as shown in Table 2). Neither capable of detecting specific effects of advertising
the group nor repetition factors had significant main content (or repetition) on attribute evaluations. Be-
effects on any of these beliefs (p < .05). However, cause advertising content was a within-subject
significant main effects due to advertising content were manipulation and the ei's were stated in terms of the
obtained for 11 (73%) of the 15 belief strength scores generic product category (i.e., they were not brand
(p < .01). Only one (7%) of the 15 possible interaction specific), the ei's were measured only once for each
effects between repetition and advertising content was subject, after multiple exposures to the varied contents
significant (p < .05), an event possibly due to chance. of the four advertisements. Therefore, the best that
Thus, only the main effects of advertising content can be done is to examine the ei scores as a function
on belief strength are of interest. of the four groups to determine whether different
As can be seen in Table 2, the advertising content confoundings of the advertising content and repetition
effects were most pronounced for the beliefs about factors (or a nonrandom assignment of subjects to
softness. Significant effects of advertising content experimental groups) influenced the e,'s (see Table
were obtained for three of the four levels of the 1). One-way ANOVAs for each of the 15 ei scores

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
BELIEFSAND BRAND ATTITUDE
PRODUCTATTRIBUTE 325

indicated a significant effect of groups for only one Alternative Mediators of Attitude Formation
belief, fairly soft, which is close to the chance level. Fishbein proposed that attitudes toward either the
On the basis of this evidence, and because the adver- act or the object are a function only of the strength
tising content of the four test advertisements does of the respective salient beliefs and the evaluations
not seem relevant to. attribute evaluations, we have of those beliefs. Alternatively, classical conditioning
assumed that any observed mediation effect of beliefs
on attitudes is probably due to the impact of advertising principles suggest that if a brand is repeatedly paired
with a positively evaluated stimulus a "direct transfer"
content on belief strengths, not belief evaluations. of that evaluation to the brand might occur (Olson
and Mitchell 1975). In terms our study, an individual's
Effects on Attitudes and Intentions
attitude toward an advertised brand might be affected
The ANOVA procedure used for the belief strength
by that person's evaluation of the advertisement as
measures was also used to examine the effects of a whole (Aad), or of a visual stimulus in the ad (Ap,).
advertising content on the more global cognitive vari- Presumably, such attitudinal impact would be
ables, Ao, Aa,, and BI. Again, neither the main effects independent of the semantic meanings that might be
due to groups or repetition nor any of the interactions
acquired from an advertisement, as represented here
were statistically significant (all p > .05). However,
by the product attribute beliefs. The relative validity
as is consistent with the belief strength results, signifi- of these two alternative explanations for the mediation
cant main effects due to advertising content were of attitude formation was examined by using regression
obtained for the Ao, Aact, and BI variables (p < .01).
analysis and analysis of covariance.
The marginal means and main effect F-ratios for these
Regression analysis. First, AOand A act were treated
effects are reported in Table 2. Specifically, the two as separate criterion variables in four multiple regres-
brands associated with the positively evaluated visual sion models. The independent variables included the
stimuli (Brands J kitten and L sunset) had significantly evaluation index for the softness belief vector (. b,e,),3
more positive brand attitudes (3.65 and 3.68) than did
Aad, and Api, as well as dummy variables for three
the brand associated with the neutral abstract painting of the four brands when A ,, was included in the
(2.82) or the brand in the explicit verbal claim adver- regression. The results of these alternative prediction
tisement (3.03; Newman-Keuls test, p < .01). The Aact
models, summarized in Table 3, indicate that either
and BI scores followed the same pattern, except that
Aad or Aic, but not both, contributes significantly
the two brands associated with the kitten and sunset
pictures were not significantly different from the
verbal-claim brand. However, all three of the latter
brands were significantly more positive than the brand 3Thepredictedattitudetowardthe brandwas computedby using
associated with the presumably neutral abstract paint- the Ahtola/Fishbein formulafor all the attributevectors (X:b,ei)
and for only the softness vector (bi ei). Correlationswere then
ing (Newman-Keuls tests, p < .01). computedbetween the respondents'ratedbrandattitude(Ao) and
The preceding results show that the manipulation these two variables.The softness vector score alone predictedA,
of advertising content had significant effects on the better(r = .42)thandid the evaluationscorebasedon allthe attribute
strength of beliefs about several product attributes, vectors (r = .28). One possible reason for these results is that
attitudes toward both the brand and the act of buying becausesoftness was the only attributeemphasizedin the advertise-
the brand, and purchase intentions. Now let us con- ments, it representedthe most stable and salient component of
cognitive structure for these brands. Olson and Dover (1978a)
sider the mediation processes by which the attitude suggestedthat the predictivevalidityof a Fishbein-typemodel may
and intention effects may have occurred. depend on the stabilityof the salient beliefs in cognitive structure.

Table 3
ALTERNATIVE
MODELSFOR PREDICTINGBRAND ATTITUDES
(beta coefficients and t-statistics)

Criterion Predictor variables


variable Model bie, A A pic R2
Ao 1 0.40 (7.33)b -.16
2 0.31 (6.10)" 0.31 (3.37)b .35
3 0.22 (5.37)" 0.65 (15.74) - .56
4 0.20 (4.71)b 0.61 (11.73)" 0.11 (1.35) .57
A^t 1 0.39 (7.12)b - .16
2 0.31 (5.65)b 0.14 (1.40) .23
3 0.28 (5.45)" 0.41 (7.99) .31
4 0.24 (4.47)" 0.42 (6.54)b -0.001 (0.01) .34
aEvaluationbased on softness belief vector.
bp < 0.005.

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
326 JOURNALOF MARKETINGRESEARCH,AUGUST 1981

Table 4
EFFECTOF COVARIATESON THE F-RATIOFOR THE MAIN EFFECTOF ADVERTISINGCONTENT

Covariate(s)
Dependent No
variable covariate Ib, eia ?bie,a & Aad Aac, Aact & .bie,I Aac, & Aad
A. 27.43b 16.03b 0.13 -
Aact 14.54b 3.18' 0.86
BI 15.62- 3.61b 2.52 2.47
aEvaluation based on softness belief vector.
bp < 0.001.
cp < 0.05.

to the prediction of Ao beyond that provided by the Mediators of Behavioral Intentions


evaluation (5 b,ei) associated with the softness beliefs For completeness, the BI scores were analyzed
(R = .74 or .59 vs. r = .40, respectively). Because the similarly (see Table 4). As specified by the Fishbein
correlation between Aad and A pi was only 0.34, extended model (cf. Lutz 1977), Aact is the critical
multicollinearity probably did not account for this mediator of BI. Including Aact as a covariate reduced
result. Much the same pattern of results was obtained the magnitude of the advertising content main effect
for the prediction of A at, except that only Aad (not on BI considerably, but not quite to nonsignificant
A ic) contributed significantly to the prediction of Aact
levels. When either Xbie, or Aad was included as a
in addition to lbiei. Empirically, Aad is a stronger second covariate, the advertising content effect on
predictor of brand attitudes than Apic. Conceptually, BI became marginally significant (F = 2.52 and 2.47,
Aad seems to be the more general construct because
both p = .06). These results suggest that Aad had its
it may incorporate the evaluation associated with the
major mediating effect on attitudes, either Ao or Aac,,
picture symbol. and had only a weak mediating influence on BI. Thus,
Covariance analysis. Next, using analysis of co-
as predicted by the Fishbein theory, Aactwas the major
variance, we examined the mediating roles of product mediator of BI.
attribute beliefs and Aad in explaining the obtained
differences in Ao and Aact (e.g., Osterhouse and Brock DISCUSSION
1970). If product attribute beliefs are the only mediator
of brand attitudes, including the presumed mediator We tested the basic theoretical proposition of Fish-
(represented by the evaluative belief index, lbie,) as bein's attitude theory in a marketing research context.
a covariate should reduce the main effect of advertising Specifically, we examined whether product attribute
content on Ao to a nonsignificant level. However, beliefs were the only mediator of advertising effects
if a significant Ao effect remains after statistical on brand attitudes. The beliefs-cause-attitudes propo-
removal of the presumed mediating effect of brand sition was contrasted with two alternative models of
attribute beliefs, these beliefs may not be the only attitude formation, sheer repetition and classical
mediator of attitude formation. The ANCOV results conditioning, neither of which treats beliefs as a causal
are summarized in Table 4 in terms of the changing mediator. Thus the study was designed to provide
F-ratios for the main effects of advertising content conditions in which either beliefs or the alternative
given different mediators/covariates. The results causal processes could have mediated the effects of
indicate that including the evaluation associated with advertising on attitude formation.
the softness belief vector (bi,ei) as a covariate did The experiment involved two manipulations: (1) the
not completely remove the effects of advertising number of times a subject saw a particular advertise-
content on either Ao or A at. However, using both ment (repetition) and (2) the type of visual or verbal
.biei and Aad as covariates eliminated any reliable information contained in the advertisement (advertis-
effects on attitude due to advertising content. When ing content). Repetition had no reliable effects on
these analyses were repeated using the combined attitude formation or any other cognitive variable.
evaluations associated with all the attribute belief Therefore, we could not test the proposition that
vectors (5zbiei) as a covariate, the results were multiple exposures to a stimulus can directly affect
identical. attitude without influencing beliefs.4 In contrast, the
In summary, the results of both the regression and
ANCOV analyses suggest that brand attitudes are not
solely a function of the attribute beliefs that are formed 4However, the lack of repetition effects on both beliefs and
attitudes is consistent with, although only weak evidence for, the
about the brand, but may also be influenced by beliefs-cause-attitude proposition. The Fishbein proposition would
consumers' general liking for the ad itself or the visual have been disconfirmed only if repetition had affected either beliefs
stimulus presented in the advertisement. or attitudes, but not the other.

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
PRODUCTATTRIBUTE
BELIEFSAND BRANDATTITUDE 327

advertising content factor produced significant effects a highly valenced visual stimulus (conditioned stimu-
on three cognitive variables-product attribute beliefs, lus). In this interpretation, the conditioning process
attitudes, and purchase intentions. due to contiguous presentation caused the evaluation
Given these main effects of advertising content, associated with the advertisement in general (Aad),
the critical question was whether product attribute or with a prominent part of the advertisement such
beliefs were the only mediator of the variation in as a picture (Api), to become associated with the
brand attitude produced by the advertisements, or brand name (see Staats and Staats 1967). Presumably,
whether another process such as classical conditioning this "direct" influence on attitude is independent of
was operating. The obtained mediation relationships the message's effect on the formation of or change
are summarized in Figure 1. They indicate that beliefs in product attribute beliefs. If this classical condition-
about product attributes were a major mediator of ing interpretation is accepted, our results are partially
the advertising content effect on attitudes, but not inconsistent with the information processing perspec-
behavioral intentions, as predicted by the causal flow tive of Fishbein.
of the Fishbein extended model. These results are Alternatively, one might argue that A ad is a surrogate
consistent with those obtained by Lutz (1977). indicator of unmeasured salient beliefs about product
However, the mediation analyses also suggest that attributes not included in the experimental question-
product attribute beliefs were not the sole mediator naire. Hence, processing the advertised information,
of attitudes. The Aadmeasure also seemed to mediate impoverished as it was, may have created beliefs about
attitude formation. In fact, together, the Aad scores unmeasured product attributes that somehow are re-
and the evaluative belief indices (X bie,) totally mediat- flected by the Aad measure. The idea is that if we
ed the effects of advertising content on brand attitudes. had measured all the salient attribute beliefs for each
These results are similar to the mediation effects of subject, the Aadmediation effect would not have been
product attribute beliefs and advertising evaluations obtained. However, as it is not obvious how the
reported by Holbrook (1978). four-item Aad score could represent beliefs about
Do the results obtained confirm or disconfirm Fish- unmeasured product attributes, this explanation seems
bein's basic proposition that beliefs cause attitude? somewhat strained. In any case, this explanation
They seem to do both. Although product attribute cannot be addressed by our data. Because we used
beliefs did mediate message effects on attitude as a structured questionnaire to measure b, and e,, all
specified by Fishbein's attitude theory, such beliefs subjects rated a fixed set of presumably salient attri-
do not appear to be the only mediator. Another butes. To examine this issue we must identify and
potential mediator, represented by the Ad measure, measure the set of beliefs that are salient for each
may be operating. To understand better the meaning subject. Lutz and Swazy (1977) proposed such a
of these results, we must examine the Aad concept procedure, and Olson and Muderrisoglu (1979) de-
more closely. scribed free-elicitation procedures by which it could
be accomplished. To date, however, no marketing-
Attitude Toward the Advertisement
oriented study has investigated how subjects' idio-
A straightforward interpretation of the Aadmeasure syncratic attribute beliefs are influenced by advertising
is that it accurately reflects subjects' evaluations of messages.
the overall advertising stimulus. Accordingly, Aad Along similar lines, the advertising content
should be treated as a construct that is conceptually manipulation could have created beliefs about con-
distinct from brand attribute beliefs and brand attitude. cepts other than product attributes (Calder 1978). For
Thus the mediation effect of Aad can be interpreted example, a consumer could have a belief that X Facial
as capturing the classical conditioning effect of pairing Tissue is manufactured by a particular company, or
an unknown brand name (unconditioned stimulus) with a person may associate a visual image with Brand
Z. Because in a general sense beliefs are the subjective
associations between cognitive representations, con-
Figure 1 sumers presumably can have beliefs about any concept
OBSERVEDMEDIATORSOF ADVERTISINGCONTENTON that they have represented cognitively. However,
COGNITIVEVARIABLES beliefs about the nonattribute concepts that may be
associated with a brand are not measured by the typical
marketing-oriented multiattribute questionnaire. The
point is that we know relatively little about the types
of cognitive representations consumers may associate
with a brand. Possibly at least some of these salient
beliefs do not concern traditional product attributes.
If semantic memory is conceptualized as a network
of associations between cognitive representations (cf.
Anderson 1976), a set of product beliefs is a semantic

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
328 JOURNALOF MARKETINGRESEARCH,AUGUST 1981

memory structure. A semantic memory concept pro- reactions to advertisements and to differentiate those
vides the basis for another explanation of the Aad effects from advertising impacts on brand attribute
mediation effect. Assume that a visual image of each beliefs, attitudes, and intentions.
advertisement is associated with the brand name How can these ideas be explored in future research?
representation in a consumer's memory. Although If Aad is a direct mediator of attitude formation, the
marketing researchers usually would not treat the Aadscore might be expected to have a greater mediation
"connection" between a brand name and the remem- effect on brand attitudes for advertisements that
bered advertisement as a belief, such an association arouse more intense evaluative or affective reactions,
is conceptually equivalent to a belief. Thus, if the whether positive or negative. Such responses might
consumer is asked to evaluate the brand, the related be due to vivid visual content or strong symbols used
visual image might be activated from memory along in the advertisement. Also, the verbal (written or
with other beliefs about the brand, perhaps including spoken) content of the advertisement might evoke
product attribute beliefs. If so, the evaluation of the intense affective feelings, perhaps by stimulating visu-
visual image, indicated by Aad in our study, should al imagery (cf. Kosslyn 1975; Pavio 1977). Thus, we
influence the overall brand attitude, Ao. By this might "test" the Aadmediation hypothesis by examin-
interpretation the Aad concept is absorbed into the ing the magnitude of its effect for advertisements that
semantic memory structure of beliefs that are asso- differ greatly in evaluative appeal but contain about
ciated with the attitude concept. Thus, Aad is not a the same verbal content.
separate construct, bttt rather is a surrogate measure Another way in which the evaluation associated with
of biei; that is, Aad is the evaluation of the advertise- an advertisement may become conditioned to a brand
ment "attribute" of the brand. If this explanation occurs when an individual pays general attention to
is accepted, our results are totally consistent with an advertisement, but does not actively process the
Fishbein's proposition that beliefs cause attitude, brand-relevant information in the advertisement
because Aadis considered to be another belief mediator (Krugman 1965; Mitchell, Russo, and Gardner 1980)-
of Ao. for instance, when a consumer attends to an advertise-
In sum, the interpretation of our results hinges on ment only for its entertainment value. Under such
the conceptual meaning of the Aad construct. At conditions, a consumer may achieve only limited
minimum, Fishbein's theoretical proposition is partial- comprehension of the brand-related information in the
ly supported. However, other processes may be message. However, when considering the advertised
operating that are not captured by the belief structure brand at some later time (say, during the purchase
model, at least as typically operationalized by market- decision process), the individual might recall charac-
ing researchers in terms of product attribute beliefs. teristics of the advertisement from memory. Then the
Although empirical data will probably be valuable in Aad that is also retrieved might influence the formation
reaching an accepted interpretation, the decision about of a brand attitude (Mitchell, in press).
how to treat Aad ultimately may hinge on more prag- Future research could examine the relative media-
matic considerations. In advertising effects research, tion effects of beliefs (Eb,ie) and advertising evalua-
for example, it may be useful to consider Aad as a tions (e.g., Aad) for advertisements of widely differing
separate construct that can provide separate diagnostic semantic and affective content, in conditions involving
information about an advertisement's attitudinal different purchase deliberation goals, and for varied
impact on consumers. In contrast, if the researcher time intervals after exposure. Such research would
is interested in memory structure, it may be more help to clarify the construct of Aad by determining
useful to consider A adas the evaluation of an image-like the specific conditions under which Aad seems to
cognitive representation that has become associated mediate attitude formation and change.
with the brand name representation. Fishbein's atti-
tude theory can incorporate the latter approach, but Visual Versus Verbal Information in Advertising
not the former (see Zajonc 1980). Relatively few studies of advertising effects have
Future research. Our research design is not capable examined the influence of visual versus verbal in-
of distinguishing between these alternative explana- formation content. Exceptions include Wright's (1979)
tions for the Aad effect. In fact, designing an empirical study of the efficacy of different visual content for
test of these causal explanations will be difficult. inducing certain desired behaviors, Rossiter and
Researchers might begin by reviewing past research Percy's (1978, in press) examinations of the attitudinal
on how consumers' reactions to the advertisement impact of visual information, and Holbrook's (1978)
per se are related to other responses (e.g., March research on the effects of more and less affective
and Swinbourne 1974; Schlinger 1979; Silk and Vavra verbal advertising copy on brand attitudes. In our
1974; Wells, Leavitt, and McConville 1971). As our study we examined four ads, only one of which
results and those of Holbrook (1978) indicate, such contained verbal information about a product charac-
evaluations of the advertisement seem to have an teristic. The other three advertisements contained only
independent impact on brand-related responses. More visual information (except for the brand and product
research is needed, however, to explicate consumers' names). Despite the relative lack of specific informa-

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
BELIEFSAND BRAND ATTITUDE
PRODUCTATTRIBUTE 329

tion, the belief strength measures indicate that subjects attribute is likely to have a certain level of another
formed rather different perceptions of the four brands attribute. For instance, if consumers believe a brand
(see Table 2). of facial tissue is very soft, they may also consider
These results are of interest for several reasons. it likely the brand will tear easily or be more absorbent.
At a descriptive level they suggest that subjects, Such inferences could be based on previously learned
apparently by some inferential process, develop beliefs associations about the generic product category which
about brand attributes based on minimal brand-specific enable a person to form numerous beliefs based on
information. In addition, they show that the measures only one piece of information (cf. Olson 1978). Second,
used in our study are sensitive enough to detect these memory structures for generic product categories may
effects. The results also suggest that subjects convert- provide default values which enable a person to make
ed visual information that was not directly related inferences about specific attributes when information
to the product into meaningful semantic information. is missing (e.g., Cantor and Mischel 1977; Rosch 1978).
This process is indicated by the obtained differences For instance, individuals may have generic knowledge
between brands in the strength of beliefs about brand structures for different categories of facial tissue
attributes, particularly the softness beliefs. For in- which provide typical values for the absorbency and
stance, the picture of a fluffy kitten produced the tearing ease of brands in that category. Basically, both
strongest beliefs that the facial tissue was very soft. of these explanations derive from semantic memory
A possible explanation for the softness belief theory. That is, people have structures of knowledge
strength effects is that seeing the advertisement con- or memory schemata which may be activated and used
taining the verbal claim about softness primed subjects as a basis for interpreting new information and for
to think about softness during exposure to the other making inferences (cf. Olson 1978, 1980).
three visual advertisements. This alternative hypothe- A third explanation for the observed "inferences"
sis was examined by taking written cognitive responses was suggested by Mitchell (1979). The overall brand
(thought protocols) from two additional groups of attitude may influence specific ratings of belief
student subjects (n = 18 and 16) immediately after strength through a "halo effect" process. The process
exposure to each of the four advertisements. We found seems especially likely to occur when subjects are
that when the verbal claim preceded the kitten adver- forced to rate beliefs about nonsalient attributes. This
tisement, about as many subjects mentioned softness explanation cannot be ruled out for some of our belief
in their protocols for the kitten advertisement (72.2%) strength results. Our procedures for establishing the
as when the kitten advertisement preceded the verbal saliency of the attribute dimensions were admittedly
claim (75.0%). Very few mentions of softness were intuitive, and all subjects had to rate every attribute.
made in response to the advertisements portraying Thus, some subjects probably rated some nonsalient
the sunset and the abstract painting. Thus, these data beliefs. The ideal situation would be for each subject
provide additional support for the idea that the kitten to rate only those beliefs that are salient for him or
picture independently connoted the semantic concept her. Although difficult, a procedure of eliciting the
of softness. salient beliefs of each individual is not impossible
For the other product attribute beliefs, the connec- (cf. Lutz and Swazy 1977; Olson and Muderrisoglu
tions between the pictorial stimuli and the inferentially 1979). In fact, such an approach is probably necessary
derived beliefs about facial tissue are somewhat less in future research if we are to develop clearer answers
obvious. For instance, subjects seem to have inter- to these theoretical questions about the basis for
preted the picture of the sunset as meaning that the inferential belief formation.
advertised brand comes in more attractive colors (than Finally, like most studies of attitude formation, our
other brands). Subjects also seemed to make in- design cannot rigorously establish when the measured
ferences about other characteristics of the four brands attribute beliefs were formed. Possibly some of the
(e.g., absorbency) even though no relevant information brand beliefs were formed during completion of the
was provided. Such inferences were especially evident questionnaire, rather than during or immediately after
for the brand paired with the abstract painting, which exposure to the advertisements. Cognitive response
was perceived as more negative on nearly every protocols taken during exposure could be used in future
attribute. Inferences are clearly important in attitude research to provide evidence of the extent of belief
formation (cf. Fishbein and Ajzen 1975), and more and attitude formation at exposure.
research on consumer inference processes is needed
(Lutz 1976; Olson 1978). To stimulate that research, Limitations
we offer the following comments on inferential belief Certain characteristics of our research were included
formation. to enhance internal validity, but obviously limit exter-
nal validity-that is, the generalizability of the results
Inferential Belief Formation to natural advertising settings. Certainly the test ad-
Inferences may be formed in at least three ways. vertisements contained much less information than
First, as suggested by Wyer (1974), individuals may actual print advertisements. Unraveling the relative
believe a product having a particular level of one effects of verbal and visual content in real, complex

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
330 JOURNALOF MARKETINGRESEARCH,AUGUST 1981

advertisements will be difficult because most contain advertising and marketing practice, they warrant
substantial amounts of both verbal and visual content further research attention.
designed to be complimentary. Another limitation of
our research is the artifical exposure environment.
For instance, the advertisements were not embedded REFERENCES
in editorial material, exposure was forced and timed,
Ahtola, Olli T. (1975), "The Vector Model of Preferences.
and the repetitions occurred within a short time period. An Alternativeto the Fishbein Model," Journal of Mar-
The last feature, in fact, may account for the nonsig-
keting Research, 12 (February), 52-9.
nificant effect of repetition. Multiple exposures to Anderson, John R. (1976), Language, Memory and Thought.
an advertisement over more natural intervals (e.g., Hillsdale, New Jersey: LawrenceErlbaum.
weeks) might produce measurable changes in beliefs, Berlyne, D. B. (1970), "Novelty, Complexityand Hedonic
attitudes, and intentions. Also, effects measured Value," Perception and Psychophysics, 8, 279-86.
in our study were taken almost immediately after Calder, Bobby J. (1975), "The Cognitive Foundations of
exposure. Whether similar mediation effects would Attitudes: Some Implications for Multi-Attribute
be obtained after post-exposure delays of days or Models," in Advances in Consumer Research, Vol. 2,
weeks is not known. Finally, by using fictitious "brand MaryJ. Schlinger,ed. Chicago:Associationfor Consumer
Research, 241-7.
names," we avoided the effects of previously acquired
brand knowledge. In natural settings, however, peo- (1978), "CognitiveResponse, Imagery,and Scripts:
What Is the Cognitive Basis of Attitude?" in Advances
ples' stored knowledge will interact with the advertised in Consumer Research, Vol. 5, H. Keith Hunt, ed. Ann
information. In sum, moving to more natural research Arbor, Michigan: Association for Consumer Research,
settings and advertising stimuli will complicate the 630-4.
research design problems and will require rigorous Cantor,Nancy and WalterMischel(1977), "Traitsas Proto-
controls and measurements. types: Effects on RecognitionMemory," Journalof Per-
sonality and Social Psychology, 33 (January), 38-48.
SUMMARY Carnegie-Mellon University Marketing Seminar (1978),
"AttitudeChangeor AttitudeFormation?An Unanswered
Our results indicate that product attribute beliefs Question," Journal of Consumer Research, 4 (March),
had a major mediating effect on brand attitudes, 271-6.
whereas attitudes substantially mediated behavioral Cohen, Joel B., MartinFishbein,and Olli T. Ahtola (1972),
intentions. These findings are consistent with Fish- "The Nature and Uses of Expectancy-ValueModels in
bein's attitude theory. However, contrary to Fish- Consumer Attitude Research," Journal of Marketing Re-
bein's major assumption as it is operationally inter- search, 9 (November),456-60.
preted in most marketing research, the product attri- Dickson, Peter R. and Paul W. Miniard(1978), "A Further
Examinationof Two LaboratoryTests of the Extended
bute belief index, lbiei, was not the sole mediator Fishbein Attitude Model," Journal of Consumer Research,
of attitude formation. Rather, a measure of attitude 4 (March),261-6.
toward the advertisement, Aad, partially mediated Dover, Philip A. and Jerry C. Olson (1977), "Dynamic
advertising effects on brand attitudes. Because these Changes in the Expectancy-ValueAttitude Model as a
results can be accounted for within the beliefs-cause- Functionof MultipleExposuresto ProductInformation,"
attitudes theoretical framework, however, our data in Contemporary Marketing Thought, Barnett A. Green-
cannot be interpreted as strongly disconfirming the berg and Danny A. Bellenger, eds. Chicago: American
basic proposition of Fishbein's attitude model. Addi- MarketingAssociation, 455-60.
tional research is needed to develop a better under- Fishbein, Martin(1963), "An Investigationof the Relation-
standing of the Aad construct, its measures, and its ships Between Beliefs About an Object and the Attitude
Towardthe Object," Human Relations, 16, 233-40.
causal dynamics, as well as to identify the specific
conditions under which it may have separate mediation (1967), "A Considerationof Beliefs and Their Role
in Attitude Measurement," in Readings in Attitude Theory
effects on attitudes and intentions relative to belief and Measurement,MartinFishbein, ed. New York: John
measures. Wiley & Sons, Inc., 257-66.
Our results also indicate that individuals can develop and Icek Ajzen (1975), Belief, Attitude, Intention
different perceptions of brands based only on visual and Behavior: An Introduction to Theory and Research.
information that provides no explicit brand informa- Reading,Massachusetts:Addison-Wesley.
tion. That is, consumers seem able to convert visual Gardner,Meryl, Andrew A. Mitchell, and Richard Staelin
information into semantic knowledge or beliefs about (1977), "The Effects of Attacks and Inoculations in a
attributes of the advertised brand. The process seems PublicPolicy Context:A CognitiveStructureApproach,"
in Contemporary Marketing Thought, Barnett A. Green-
to occur even for visual stimuli that are seemingly
irrelevant to the product. These phenomena may be berg and Danny A. Bellenger, eds. Chicago: American
MarketingAssociation, 292-7.
considered in terms of people's memory structures Harrison,A. A. (1968),"ResponseCompetition,Frequency,
for the generic product categories and the inference ExploratoryBehaviorand Liking,"Journalof Personality
processes that are guided by these cognitive structures. and Social Psychology, 10 (August), 363-8.
As these theoretical issues also have implications for (1977), "More Exposure," in Advances in Experi-

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
BELIEFSAND BRAND ATTITUDE
PRODUCTATTRIBUTE 331

mental Social Psychology, Vol. 10, Leonard Berkowitz, sure to Advertising," in Information Processing Research
ed. New York: Academic Press, 39-83. in Advertising, Richard J. Harris, ed. Hillsdale, New
Holbrook, Morris B. (1978), "Beyond Attitude Structure: Jersey: Lawrence Erlbaum.
Toward the Informational Determinants of Attitude," and Jerry C. Olson (1977), "Cognitive Effects of
Journal of Marketing Research, 15 (November), 545-56. Advertising Repetition," in Advances in Consumer Re-
Kirk, R. E. (1968), Experimental Design: Procedures for search, Vol. 4, William D. Perreault, Jr., ed. Atlanta:
the Behavioral Sciences. Belmont, California: Brooks/ Association for Consumer Research, 213-110.
Cole. , J. Edward Russo, and Meryl Gardner (1980),
Kosslyn, Stephen M. (1975), "Information Representation "Strategy Induced Low Involvement Processing of Ad-
in Visual Images," Cognitive Psychology, 7 (July), 341-70. vertising Messages," working paper, Graduate School of
Krugman, Herbert E. (1965), "The Impact of Television Industrial Administration, Carnegie-Mellon University.
Advertising: Learning Without Involvement," Public Olson, Jerry C. (1978), "Inferential Belief Formation in
Opinion Quarterly, 29 (Fall), 349-56. the Cue Utilization Process," in Advances in Consumer
Lutz, Richard J. (1975), "Changing Brand Attitudes Through Research, Vol. 5, H. Keith Hunt, ed. Ann Arbor, Michi-
Modification of Cognitive Structure," Journal of Con- gan: Association for Consumer Research, 706-13.
sumer Research, 1 (September), 49-59. (1980), "Encoding Processes: Levels of Processing
(1976), "First-Order and Second-Order Cognitive and Existing Knowledge Structures," in Advances in
Effects in Attitude Change," in Communicating with Consumer Research, Vol. 7, Jerry C. Olson, ed. Ann
Consumers, Michael L. Ray and Scott Ward, eds. Beverly Arbor, Michigan: Association for Consumer Research,
Hills, California: Sage Publications, 101-12. 154-60.
(1977), "An Experimental Investigation of Causal and Philip A. Dover (1976), "Effects of Expectation
Relations Among Cognitions, Affect, and Behavioral In- Creation and Disconfirmation on Belief Elements of
tention," Journal of Consumer Research, 3 (March), Cognitive Structure," in Advances in Consumer Research,
197-208. Vol. 3, Beverly B. Anderson, ed. Cincinnati: Association
(1978a), "Rejoinder," Journal of Consumer Behav- for Consumer Research, 168-75.
ior, 4 (March), 266-71. and (1978a), "Attitude Maturation: Changes
(1978b), "Rejoinder," Journal of Consumer Re- in Related Belief Structures Over Time," in Advances
search, 4 (March), 276-8. in Consumer Research, Vol. 5, H. Keith Hunt, ed. Ann
and James R. Bettman (1977), "Multiattribute Models Arbor, Michigan: Association for Consumer Research,
in Marketing: A Bicentennial Review," in Consumer and 333-42.
Industrial Buying Behavior, Arch G. Woodside, Jagdish and (1978b), "Cognitive Effects of Deceptive
N. Sheth, and Peter D. Bennett, eds. New York: North- Advertising," Journal of Marketing Research, 15 (Fe-
Holland, 137-50. bruary), 20-38.
and John L. Swazy (1977), "Integrating Cognitive , Rajesh Kanwar, and Aydin Muderrisoglu (1979),
Structure and Cognitive Response Approaches to Moni- "Clarifying the Confusion Regarding Salience, Impor-
toring Communications Effects," in Advances in Con- tance, and Determinance Concepts in Multi-Attribute
sumer Research, Vol. 4, William D. Perreault, Jr., ed. Attitude Research," in AMA Educators' Conference Pro-
Atlanta: Association for Consumer Research, 363-71. ceedings, Neil Beckwith, et al., eds. Chicago: American
March, Robert M. and Donald W. Swinbourne (1974), "What Marketing Association, 286-90.
is 'Interest' in TV Commercials?," Journal of Advertising and Andrew A. Mitchell (1975), "The Process of
Research, 14 (August), 17-22. Attitude Acquisition: The Value of a Developmental
Matlin, M. W. (1970), "Response Competition as a Mediating Approach to Consumer Attitude Research," in Advances
Factor in the Frequency-Affect Relationship," Journal in Consumer Research, Vol. 2, Mary J. Schlinger, ed.
of Personality and Social Psychology, 16 (November), Chicago: Association for Consumer Research, 240-64.
536-52. and Aydin Muderrisoglu (1979), "The Stability of
Mazis, Michael B. and Janice E. Adkinson (1976), "An Responses Obtained by Free Elicitation: Implications for
Experimental Evaluation of a Proposed Corrective Adver- Measuring Attribute Salience and Memory Structure,"
tising Remedy," Journal of Marketing Research, 13 (May), in Advances in Consumer Research, Vol. 6, William L.
178-83. Wilkie, ed. Ann Arbor, Michigan: Association for Con-
McGuire, William J. (1968), "The Nature of Attitudes and sumer Research, 269-75.
Attitude Change," in Handbook of Social Psychology, Osterhouse, R. A. and Timothy C. Brock (1970), "Distrac-
2nd ed., Vol. 3, Gardner Lindzey and Elliot Aronson, tion Increases Yielding to Propaganda by Inhibiting Coun-
eds. Reading, Massachusetts: Addison-Wesley, 136-314. terarguing," Journal of Personality and Social Psychology,
Miniard, Paul W. and Joel B. Cohen (1979), "Isolating 15 (August), 344-58.
Attitudinal and Normative Influences in Behavioral Inten- Pavio, Allan (1977), Imagery and Verbal Processes. New
tions Models," Journal of Marketing Research, 16 (Febru- York: Holt, Rinehart and Winston.
ary), 102-10. Perreault, William D. and William R. Darden (1975), "Un-
Mitchell, Andrew A. (1979), "Predicting Choice with Prefer- equal Cell Sizes in Marketing Experiments: Use of General
ence Functions Estimated from Perceptions of Brands Linear Hypotheses," Journal of Marketing Research, 7
Within a Product Category," in Analytic Approaches to (August), 333-42.
Product and Marketing Planning, Allan Shocker, ed. Platt, John R. (1964), "Strong Inference," Science, 146
Cambridge, Massachusetts: Marketing Science Institute, (October), 347-53.
368-87. Rosch, Eleanor (1978), "Principles of Categorization," in
(in press), "Cognitive Processes Initiated by Expo- Cognition and Categorization, Eleanor Rosch and Barbara

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions
332 JOURNALOF MARKETINGRESEARCH,AUGUST 1981

B. Lloyd, eds. Hillsdale, New Jersey: Lawrence Elbaum, Learning and Affect," Journal of Personality and Social
27-48. Psychology, 31 (January), 7-12.
Rosenberg, Milton J. (1956), "Cognitive Structure and At- Timm, N. H. and J. E. Carlson (1975), Analysis of Variance
titudinal Affect," Journal of Abnormal and Social Psy- Through Full Rank Models, Multivariate Behavioral Re-
chology, 53, 367-72. search Monograph No. 75-1, Texas Christian University,
Rossiter, John R. and Larry Percy (1978), "Visual Imaging Forth Worth, Texas.
Ability as a Mediator of Advertising Response," in Ad- Triandis, Harry C. (1971), Attitude and Attitude Change.
vances in Consumer Research, Vol. 5, H. Keith Hunt, New York: John Wiley & Sons, Inc.
ed. Ann Arbor, Michigan: Association for Consumer Wells, William D., Clark Leavitt, and Maureen McConville
Research, 621-28. (1971), "A Reaction Profile for TV Commercials," Journal
and - (in press), "Visual Communication in Ad- of Advertising Research, 11 (December), 11-17.
vertising," in Information Processing Research in Adver- Wilkie, William L. and Edgar A. Pessemier (1973), "Issues
tising, Richard J. Harris, ed. Hillsdale, New Jersey: in Marketing's Use of Multi-Attribute Attitude Models,"
Lawrence Erlbaum. Journal of Marketing Research, 10 (November), 428-41.
Sawyer, Alan G. (1977), "Repetition and Affect: Recent Wilson, William R. (1979), "Feeling More Than We Can
Empirical and Theoretical Developments," in Consumer Know: Exposure Effects Without Learning," Journal of
and Industrial Buying Behavior, Arch G. Woodside, Personality and Social Psychology, 37 (June), 811-21.
Jagdish N. Sheth, and Peter D. Bennett, eds. New York: Winer, B. J. (1971), Statistical Principles in Experimental
North Holland, 229-42. Design, 2nd ed. New York: McGraw-Hill Book Company.
Schlinger, Mary Jane (1979), "A Profile of Responses to Wright, Peter L. (1979), "Concrete Action Plans in TV
Commercials," Journal of Advertising Research, 19 Messages to Increase Readings of Drug Warnings," Jour-
(April), 37-46. nal of Consumer Research, 6 (December), 256-69.
Silk, Alvin J. and Terry G. Vavra (1974), "The Influences Wyer, Robert S. (1974), Cognitive Organization and Change:
of Advertising's Affective Qualities on Consumer Re- An Information Processing Approach. Hillsdale, New
sponse," in Buyer/ Consumer Information Processing, G. Jersey: Lawrence Erlbaum.
David Hughes and Michael L. Ray, eds. Chapel Hill: Zajonc, Robert B. (1968), "Attitudinal Effects of Mere
University of North Carolina Press. Exposure," Journal of Personality and Social Psychology
Staats, Arthur W. and Carolyn Staats (1967), "Attitudes Monograph Supplement, 9 (February), 1-27.
Established by Classical Conditioning," in Readings in (1980), "Feeling and Thinking: Preferences Need No
Attitude Theory and Measurement, Martin Fishbein, ed. Inferences," American Psychologist, 35 (February), 151-
New York: John Wiley & Sons, Inc., 337-81. 75.
Stang, David J. (1975), "Effects of 'Mere Exposure' on

AG-CHEM MARKETING CONFERENCE


FEES: Early Late
October 25-27, 1981 AMA Members $105 $120
Stouffer's National Center Hotel Non-Members $120 $135
Arlington, Virginia
How-to conference for professional agricultural researcher covering all
aspects of marketing research including theory and application. General
sessions on technique and discussion sessions on problem solving. Sponsored
by the Agricultural-ChemicalMarketingResearch Section of the Industrial
MarketingDivision of AMA.
EDUCATION DEPARTMENT
American MarketingAssociation
250 S. Wacker Drive
Chicago, Illinois 60606

This content downloaded from 200.89.68.193 on Thu, 6 Nov 2014 07:06:56 AM


All use subject to JSTOR Terms and Conditions

You might also like