The Effects of Music in Advertising on Choice Behavior: A Classical Conditioning Approach

Author(s): Gerald J. Gorn

Source: Journal of Marketing, Vol. 46, No. 1 (Winter, 1982), pp. 94-101
Published by: American Marketing Association
Stable URL: http://www.jstor.org/stable/1251163 .
Accessed: 24/09/2014 06:20

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.

http://www.jstor.org

This content downloaded from 175.159.14.205 on Wed, 24 Sep 2014 06:20:52 AM

All use subject to JSTOR Terms and Conditions
 Gerald J. Gorn

The Effects of
Music In Do features like humor, sex, color, and music in
a commercial merely increase our attention to
Advertising
On product information in a message, or can they
directly influence our attitudes? The results of an
experiment using a classical conditioning ap-
Choice proach suggest that hearing liked or disliked mu-
sic while being exposed to a product can directly
Behavior: A affect product preferences. A second experiment
differentiated communication situations where a
classical conditioning approach or an information
Classical processing approach might be appropriate in ex-
plaining product preference.
Conditioning
Approach

OMMERCIALS typically contain both product conditioning suggests that positive attitudes towards
specific information and background features an advertised product (conditioned stimulus) might
such as pleasant music, attractive colors, and humor. develop through its association in a commercial with
This paper examines the impact of the background other stimuli that are reacted to positively (uncondi-
features on product preferences. One experiment was tioned stimuli). Attractive colors, pleasant music, and
conducted to determine whether background features humor are examples of potential unconditioned stim-
of a commercial affected product preferences when uli in a commercial. Potentially, classical condition-
only minimal product information was presented. A ing could, in fact, explain the effect of many variables
second experiment examined the relative importance in communication-attitude change situations. For ex-
of background features and product information in ample, the communicator effect may to some extent
different situations. be due to the association of the attitude object with
The impact of product information in a commer- positive affect attached to the communicator.
cial on beliefs and attitudes would typically be inter- In marketing, classical conditioning is often men-
preted within an information processing framework. tioned and generally accepted as a process relevant to
It is suggested here that a classical conditioning advertising (e.g., Engel, Blackwell, and Kollat 1978;
framework could account for the potential impact of Schiffman and Kanuk 1978). There is, however, little
background features on product attitudes. Classical empirical research on whether preferences for objects
can actually be classically conditioned. In psychol-
ogy, where classical conditioning has been investi-
GeraldJ. Gornis a professor
intheFaculty of CommerceandBusiness gated more extensively, there is little evidence that
Administration,
Universityof BritishColumbia. Theauthorwouldlike attitudes can be classically conditioned (Brewer 1974,
to thankthe Institut
D'administrationdes Entreprises, Aix-
University Fishbein and Ajzen 1975). The limited popularity of
Marseille,
France,forits helpfulassistanceinformulating
andpretest-
classical conditioning may be due to several difficul-
ingthe project.
ties associated with typical conditioning experiments.

Journal of Marketing
94 / Journalof Marketing,Winter1982 Vol. 46 (Winter 1982), 94-101.

This content downloaded from 175.159.14.205 on Wed, 24 Sep 2014 06:20:52 AM

All use subject to JSTOR Terms and Conditions
 Difficulties With Classical tempt to think well of oneself (Greenwald 1980). A
rationalanalysis of behavior in a communicationsit-
Conditioning uation might be biased in favor of informationand,
Demand Characteristics therefore, of an information processing model. For
Fishbeinand Ajzen (1975) suggest that attitudechange example, it may be more logical to say that you like
in conditioningexperimentsmay be a function of the the productmore after seeing a commercial, because
demand characteristicsof the experimentalsituation. you now believe it has X characteristic,ratherthan
because you liked the music in the commercial. Thus
They believe that the subject must be consciously
aware of the presence of the unconditionedstimulus evidence supportingclassical conditioningis unlikely
when the conditioned stimulus is present in order for to emerge in self-reports.
conditioning to take place. This awareness may lead
subjects to believe that the experimenterwants them Arousal
to respond positively to a conditioned stimulus if it
is paired with a positive unconditionedstimulus. But In testing for classical conditioning, product infor-
awareness of the conditioned stimulus/uncondi- mation in the commercialmust be kept minimal, oth-
tioned stimulus contingency should not automatically erwise the unconditioned stimuli in the commercial
mean that attitudechange is more likely the result of might merely be arousing interest in product infor-
demand characteristics than conditioning. In fact, mation. Background features such as sex, music,
there is evidence of attitudeconditioning even where color, and humor have typically been treated as
demandtype responses were minimizedthroughelab- arousal stimuli or stimuli that reinforce the informa-
orate cover stories later verified as believed by sub- tion in the commercialratherthanunconditionedstim-
jects (Zanna, Kiesler, and Pilkonis 1970). uli (see Engel, Blackwell, and Kollat 1978, for a dis-
cussionof this material).In the printmedium,Kroeber-
Lack of Awareness Riel (1979) found that ads that were arousing pro-
The consumer may not always be aware that the un- duced better recall of the information in the ad. A
conditioned stimuli in a commercial may affect his/ greater knowledge of the product informationcould
her product attitudes. Even more generally, the con- presumably create more positive product attitudes.
sumer may not be aware of the real forces impacting However, by minimizing product information in a
on both attitudes and behavior. For example, many commercial one can investigate the potential ability
in-store features may influence consumers although of unconditionedstimuli to change consumerattitudes
they may not be easily evoked in post-purchaseex- directly.
planationsof buying (amount of shelf space devoted
to a product, color of the package, and so on). Mere Exposure Versus Classical Conditioning
Nisbett and Wilson (1977) reviewed empiricalevi-
dence casting doubt on people's ability to introspect Mere exposure effects can be confounded with clas-
sical conditioningeffects. While people may develop
accuratelyregardingtheir cognitive processes.' They
suggest that people often speculateabout the potential favorableattitudestowardsproductsadvertisedin the
causes of their own behaviorand select stimuli which context of unconditionedstimuli, these attitudeshifts
they think are probable reasons for their behavior. may be simply a function of mere exposure to the
They may conclude that what they think should be the advertised product. Zajonc (1968) has emphasized
cause of their behavior is in fact the cause of their that exposure leads to liking. Krugman (1965) sug-
behavior (Nisbett and Wilson 1977). gested that in low involvement situations,people may
In a typical communicationsituation, people may purchase a product simply because they have been
not realize or accept the impactof unconditionedstim- exposed to it before through a commercial. It is in-
uli on their responses, since unconditioned stimuli teresting to note that with the Krugmanmodel, ex-
should not logically be related to behavior. Possible posure leads to behavior, with attitude not concep-
classical conditioning effects might, therefore, be tualized as an intervening variable. With classical
underestimatedand underreportedin self-reports. In conditioning, however, the person is not choosing an
analyzing their behavior, people may attemptto con- object simply because it's been seen before. His/her
vince both themselves and the researcherthat they are affect is also involved, so that if the product(condi-
strictlyrational. Cognitive bias can result from an at- tioned stimulus) is paired with a negative uncondi-
tioned stimulusit would be avoided despite exposure.
The following experimentinvestigatingthe impact
'The Nisbett and Wilson (1977) article is controversial; see both of the backgroundfeatures of a commercialon prod-
Ericson and Simon (1980) and Weitz and Wright (1979) for criti- uct preferences allowed for a test of mere exposure
cisms. versus classical conditioning effects.

Effectsof Musicin Advertising/ 95

This content downloaded from 175.159.14.205 on Wed, 24 Sep 2014 06:20:52 AM

All use subject to JSTOR Terms and Conditions
 Experiment One supportedthis notion, as 22 out of 41 picked the beige
pen, 19 the light blue pen.
Basic Structure and Hypothesis
In the present study, subjects were exposed to a neu- Design and Procedure
tral product (conditioned stimulus) in the context of The experimentwas carriedout duringclass time. The
a backgroundfeature (unconditionedstimulus), often experimenter explained that an advertising agency
found in commercials. To minimize the difficulties was trying to select music (unconditionedstimulus)
mentioned earlier in employing a conditioning para- to use in a commercial for a pen (conditioned stim-
digm, a number of steps were taken. To lessen de- ulus) producedby one of its clients. They would hear
mandcharacteristics,an unobtrusive,behavioralmea- some music that was being considered while they
sure of productpreferencewas used insteadof a paper watched the slide of the pen which the agency was
and pencil measure. Informationon the conditioned planningto advertise. The pen was inexpensive look-
stimuluswas kept as close to zero as possible to dem- ing and cost forty-nine cents. Very little information
onstratethat the unconditionedstimuluscould directly regardingthe pen's attributeswas visible in the slide.
affect product preference even where product infor- The following conditions were structured:1) con-
mation was minimal. A negative unconditionedstim- dition LI-Liked music, light blue pen; 2) condition
ulus condition was structuredto test for mere expo- L2-Liked music, beige pen; 3) conditionDl-Disliked
sure versus classical conditioning. Consistent with a music, light blue pen; 4) condition D--Disliked mu-
conditioning interpretation,it was hypothesized that sic, beige pen.3 While subjects had been randomly
subjects would prefer an unexposed versus exposed assigned to the two sections of the course, a counter-
product if the exposed product were paired with a balancing procedure was used to control furtherfor
negative unconditioned stimulus. In contrast, where any differences that might have existed between the
the conditioned stimulus was paired with a positive two sections. To counterbalancethe music and the
unconditioned stimulus, subjects would choose the color of the pen within each class, one-half of the first
class was randomlyassigned to the liked music-light
exposed ratherthan unexposed product.
blue pen condition (LI), while the other half was as-
Sample signed to the disliked music-beige pen condition
244 undergraduatesrandomly assigned during regis- (D2). In the second class, one-half of the students
were randomly assigned to the liked music-beige
tration to two sections of a first year management
course at McGill University served as subjects. pen condition (L2), with the other half assigned to the
disliked music-light blue pen condition (DI). The
Pilot two conditions in each class were run sequentially
with one-half of the class taking a break while the
In a pilot, a group of 10 subjects evaluated 10 dif-
other half participatedin the project.
ferent pieces of music on a scale rangingfrom dislike
The subjects in each condition heard the music
very much (1) to like very much (5). The results of
the pilot led to the choice of a one-minute extract of while they watched the slide. They then evaluatedthe
music from the movie "Grease" as the positive un- music on a scale ranging from dislike very much (1)
conditioned stimulus (x=4.3). One minute of classi- to like very much (5). They were told that they would
cal Indian music (x= 1.5) served as the negative un- receive either a light blue or beige pen for their help,
conditioned stimulus.2 donated by the company that manufacturesthe pen.
Two different colors of a pen were used as con- The experimenterheld up each pen briefly and com-
ditioned stimuli, light blue and beige, neutralenough mentedthat if they wanted a light blue one, to go over
so that associating them with liked or disliked music to the left side of the room to pick one up and drop
could change color preferences.A pilot on 23 subjects off their question sheet in a box provided. If they
had revealed that 80% felt neutralaboutlight blue and wanted a beige one, they were told to go to the right
side of the room to pick up the pen and drop off their
beige pens, whereas, for example, yellow pens were
generally disliked, with black pens generally liked. question sheet. Using two different locations for the
A second sample (N=41) of subjects were asked di- boxes (150 pens per box) of beige and blue pens, with
rectly whetherthey preferredbeige or light blue pens. question sheet drop-off boxes next to them, allowed
It was expected that if subjects were neutralabout the
two colors, then half would choose each. The results
3No condition containing both positive and negative music was
structured, since it was felt the demand characteristics would be too
great under such circumstances; subjects would be more likely to
2Although musical preferences are affected by prior learning, music think that we expected them to pick the color of the pen associated
was treated as the unconditioned stimulus since it was not being con- with the positive music if both positive and negative music were used
ditioned in the present experiment. in the same condition.

96 / Journalof Marketing,
Winter1982

This content downloaded from 175.159.14.205 on Wed, 24 Sep 2014 06:20:52 AM

All use subject to JSTOR Terms and Conditions
 for a relatively unobtrusiverecordingof choices and
linking of these choices to the evaluation of the mu- TABLE1
sic.4 As a result of the music evaluation ratings, ap- Liked Versus Disliked Music and Pen Choices
proximately10 subjects were eliminatedfrom each of Pen Choice
the four conditions. They were eliminated if in the "Advertised" "Non-advertised"
liked music conditions, they did not evaluate the mu- pen pen
sic from "Grease" as either like very much (5) or like Liked
somewhat (4), or in the disliked music conditions, if music 74 20 94
they did not evaluate the Indianmusic as either dislike Disliked
very much (1) or dislike somewhat (2). For the re- music 30 71 101
maining subjects, it was predicted that they would 104 91 195
pick the color of the pen they saw when they heard x2 = 47.01 (p < .001)
the liked music and the alternativecolor when they <2 = .24
heard the disliked music.
Once out of class, subjects were handed the fol-
lowing free response question: "Why did you pick the
color of the pen you picked rather than the other Only five said the music had an influence on their
color? If you have any reasons please list them be- choice and none mentionedthat they were simply fol-
low." There was room for thiee answers. lowing a friend.When time permitted,subjects(n=54)
Finally, a discussion about the natureof the proj- who gave color as a reason for choice were pressed
ect was held with 10 students at the end of class one furtherand asked if the music had any influence on
and 10 studentsat the end of class two. Not one men- them. They respondedno, making comments such as
tioned that our real purpose was to influence their "I have always liked light blue," or with beliefs such
particularcolor preferences. When told that this was as "beige is stylish" (eight people).
in fact the purpose, some wonderedhow we were able
to detect their preferences, since they did not put the Discussion
color of the pen they picked or their names down on The majoraim of the presentstudy was to take a prod-
the question sheet. They also did not see anyone re- uct relevant to the subject sample (studentspurchase
cording their choices when they picked the pens from and use pens) and to advertise it in a favorable or
the boxes. unfavorable context simply by associating it with
liked or disliked music. The results supportedthe no-
Results tion that the simple association between a product
Comparingthe effect of liked versus disliked music (conditioned stimulus) and another stimulus such as
(Li and L2 collapsed versus DI and D2 collapsed), music (unconditioned stimulus) can affect product
there was a very clearcut impact of the music in the preferencesas measuredby productchoice.
expected direction. As can be seen in Table 1, 74 out While classical conditioning can parsimoniously
of 94 subjects (79%) picked the color of the pen as- explain the results, the data could also be interpreted
sociated with the liked music, while only 30 out of within an informationprocessing framework. While
101 subjects (30%) picked the color of the pen as- only minimal productinformationwas presented, the
sociated with the disliked music.5 music might have stimulatedproduct-relevant thoughts.
When asked for possible reasons for their choice, It might have suggested potentialattributesof the pen
126 out of 205 subjects (62%) indicatedthat they did or appropriatecontexts in which it could be used. For
have a particularreason. Of these 126 people, 114 example, liked music with an upbeat sound might
(91%) mentioned color preference as their reason. stimulatethe development of beliefs that a particular
color of a pen is a fun color or that it is appropriate
for an active lifestyle. These beliefs linking the pen
4A behavioral rather than verbal measure of preference was used, to the music might then influence product choice.
in part, because it could be recorded unobtrusively. In addition, be-
havioral measures of attitudes are generally preferred to verbal mea- However, there were very few belief type comments
sures, although less prevalent in the literature (Carlsmith, Ellsworth, that were elicited when subjects did in fact give rea-
and Aronson 1976). sons for theirchoice. Taken togetherwith the fact that
'The color of pen did not seem to matter. 36 out of 43 people (84%)
picked the light blue pen when it was associated with the liked music only minimal product information was presented,
(L,); 38 out of 51 (74%) picked the beige rather than the light blue classical conditioning would seem to be at least as
pen when the beige pen was associated with the liked music (L2). plausible an explanationof the results as information
Similarly, 17 out of 52 (33%) chose the light blue pen when it was
associated with disliked music (D,) while 13 out of 49 (26%) picked processing.
the beige pen when it was associated with the disliked music (D2). When a reason was given for choice, it was pri-
These differences were not statistically significant.
marilythatthe pen of a particularcolor was preferred.

Effectsof Musicin Advertising/ 97

This content downloaded from 175.159.14.205 on Wed, 24 Sep 2014 06:20:52 AM

All use subject to JSTOR Terms and Conditions
Very few subjects pointed out that the music had an the primacy of affect and presents a viewpoint con-
impact on their choice. Zajonc (1980) notes that even sistent with a perspective that people may not always
when questioned by others, people may have diffi- be actively processing information and evaluating sit-
culty verbalizing reasons for feelings or describing uations.
them. Instead, they may utilize more rational and Kassarjian (1978, 1981) and Olshavsky and Gran-
more easily verbalized cognitions (e.g., product spe- bois (1979) have suggested that there are situations
cific information, or general statements of attitude where information search and evaluation are not rel-
like "beige is my favorite color") when justifying a evant to purchase behavior. Probably there are also
product preference or purchase to themselves and es- many communication situations where the individual
pecially to others. This would be particularly true if is not really searching for or evaluating information.
people are not aware of the forces that influence their A second experiment was conducted to examine the
behavior, which may be true in some cases (Nisbett importance of product information relative to the
and Wilson 1977). background features of a commercial, in two different
The verbal reports were obtained with a free re- communication situations.
sponse format. This format was chosen and the ques-
tions presented in a matter of fact way, to elicit salient
reasons for choice without encouraging the subject to Experiment Two
engage in any justification of choice to both him/
herself or to us. A more formal measure of beliefs Rationale and Hypothesis
regarding the colors of the pens chosen and not cho- We are often exposed to a commercial when we are
sen could be incorporated in future research. How- not thinking of even potentially buying the advertised
ever, even if beliefs consistent with choice are ob- product. We are not in a decision making mode and
tained, the causal relationship between beliefs and are not likely to be seeking information from any
choice might remain unclear. Are the beliefs influ- source. In such situations, the commercial's impact
encing choice, or to justify behavior, is the choice may often be more related to stimuli that can arouse
influencing the formation of beliefs? emotion, such as music and colors, than to product
Had time permitted in the present study, post hoc information. A recent study suggests that in fact, very
verbal reports of message evoked thoughts might have little product information is contained in most Amer-
shed light on the extent and nature of cognitive activ- ican and Canadian commercials (Pollay, Zaich-
ity during exposure. Message evoked thoughts related kowsky, and Fryer 1980).
to the product would have favored an informational An information processing explanation of post-ex-
rather than conditioning interpretation. It is interesting posure attitudes and behavior might be more appro-
to note that in the post project discussion with 20 sub- priate when the person at the time of exposure to the
jects (10 per class), most said that they were thinking commercial is thinking about making a purchase in
of very little when listening to the music. The six the relevant product category. For example, a buyer
people that did recall something specific, recalled may see a commercial for car X while thinking of
pleasant or unpleasant situations from their past. buying a new car. In such situations the buyer might
be regarded as being in a decision making framework
Exposure and, therefore, perhaps very interested in the infor-
The results of the present study do not support an mation that commercial might contain about car X.
interpretation that mere exposure leads to liking. The causal impact of such information on cognitive
Those who heard disliked music avoided the color of structure, and in turn, its effects on attitudes and be-
the pen they were exposed to. Zajonc (1968) has havioral intentions have been demonstrated in care-
pointed out that in many situations, exposure is con- fully controlled research (e.g., Lutz 1977).
founded with a number of other variables (in the case Product information should be most important
of this experiment, liked or disliked music) that may where the decision making context is salient at the
be more crucial in attitude formation. time of exposure to a commercial. However, the un-
While there were no exposure effects in this study, conditioned stimulus or background features could be
feelings play as important a role in exposure theory more important than the product information where
as in a classical conditioning theory of communication exposure is a nondecision making context. The pres-
effects. Zajonc (1980) notes that people tend to de- ent study was conducted to test this hypothesis.
velop a positive feeling toward objects they have seen
before even if they cannot recall prior exposure. Sample
While it may be true that there are reasons for these The sample consisted of 122 students randomly as-
feelings, these reasons may come later and perhaps signed during registration to two sections of an upper
only when questioned by others. Zajonc (1980) stresses level, undergraduate management course. Both classes

98 / Journalof Marketing,Winter1982

This content downloaded from 175.159.14.205 on Wed, 24 Sep 2014 06:20:52 AM

All use subject to JSTOR Terms and Conditions
 were taught in the middle part of the day and by the A significant difference in choices between the
same instructor.The experimentwas conductedat the decision making context condition and the nondeci-
beginning of the term so that different class experi- sion making context condition was hypothesized. It
ences were less likely to be a significant factor. was predicted that subjects in the former condition
would pick the light blue pen (the one advertisedwith
Design and Procedure information),and subjects in the latter condition, the
As a cover story, subjects in the decision makingcon- beige pen (the one advertisedwith music).
dition (n=59) were told that an advertising agency
would like their help in determining whether or not Results
to purchase advertising time in an upcoming major In the decision making condition, 42 out of the 59
networkTV program.They were told that one of the (71%) subjects chose the light blue pen, advertised
advertising agency's specializations was producing with information,with the remaining17 (29%) choos-
commercialsfor companies that marketedpens. Two ing the beige pen, advertisedwith music. There was
pens, each manufacturedby a different company, a strong crossover in the nondecision making condi-
were slated to be advertisedon the program.For help- tion with 40 out of 63 (63%) choosing the beige pen
ing the agency evaluate the programthey would later and the remaining 23 (37%) choosing the light blue
be given a threepen packet of the companyA or com- pen. As expected, there was a significant difference
pany B pen. Three pens were given ratherthan just in choices between the two conditions (decision vs.
one because it was thought that informationrelevant nondecision making conditions: x2(1) = 14.72, p <
to a choice would be more influential if the decision .001).
were made somewhat more important.
Subjects were then told that they would be given
Discussion
an idea of the kinds of commercials planned for the The results of this second experimentsuggest that an
program. They were shown a slide of company A's individual in a decision making mode when exposed
pen and told that the company was planning a musi- to a commercialis affected by the informationit con-
cally-based ad to sell its pen. They then listened to tains. In contrast, that informationhas less impact in
one minute of "liked" music (the "Grease" music nondecisionmaking situations. In such situations, the
used in the first experiment) while they watched the classicalconditioningof the advertisedproductthrough
slide of a beige pen. Afterwardsthey were shown a its pairingwith the unconditionedstimuli in the com-
slide of a light blue pen and were told that company mercial appearsto account for subsequentchoice be-
B was planning to emphasize in its commercial that havior.
the pen wrote very smoothly, that it never smudged, Possibilities for future research include varying
and could last a long time even with regularuse.6 the importanceof the decision to be made and/or the
After seeing the pen slides and viewing the pro- time lag between exposure and actual choice. One
gram (entitled "Drugs and Teenagers"), subjectscir- could hypothesize that the importanceof information
cled yes or no on their answer sheets as to whether to the process of selection is a function of the impor-
the ad agency should purchase time slots in the pro- tance of the decision. This should be particularlytrue
gram. They were instructedthat at the end of the class if the individual is in a decision making mode when
(one hour and 10 minutes later), they should drop the exposed to that information. In addition, as the time
answer sheet in the box provided, and pick up their lag between exposure and actual choice is increased,
pens. The time delay between exposure and choice the relative impact of informationshould be reduced,
was included to try to mirror the real life situation particularlywhere the individual is not in a decision
where people usually do not make their actual pur- making mode when exposed to the information.In a
chases until some time after exposure to a commer- nondecision making mode, one might not be inter-
cial. At the end of the class, the experimenterpointed ested in productinformation.Therefore, even if such
to two boxes in different locations where subjects informationis acquired and rememberedfor a short
could pick up the packets of beige or light blue pens. while, with time it should be easily forgotten.Length-
Subjects in the nondecision making condition (n=63) ening the time lag between exposureand choice could
followed exactly the same procedurebut were not told also help determine the longevity of classically con-
until the end of the class that they would get a three- ditioned preferences.
pen packet of the beige or light blue pens. Future research might explore the tentative find-
ings of this experimentin a context where real com-
mercials are utilized. However, real commercials
6The results of experiment one suggested that beige and light blue
were both neutral colors for a pen. Therefore, it did not seem nec- usually reflect the use of a numberof techniquesand
essary to incorporate color of pen as a variable in the second exper- may not always be appropriatefor theory testing. For
iment.
example, it might be very difficult to isolate an in-

Effects
of MusicinAdvertising
/ 99

This content downloaded from 175.159.14.205 on Wed, 24 Sep 2014 06:20:52 AM

All use subject to JSTOR Terms and Conditions
formation based commercial totally devoid of such cording of the activity in the two hemispheres during
unconditioned stimuli as attractive colors or music. exposure to musically based versus product informa-
tion based commercials would yield more activity in
the right than left hemisphere with the former type of
Implications and Speculations commercial, and the reverse with the latter. Further-
Measures of the effectiveness of commercials typi- more, activity in the right hemisphere should also be
cally stress recall of the basic selling points of the higher in nondecision making contexts where people
commercial. The impact of music and other back- should not be seeking or analyzing product informa-
ground features is usually neglected (Honomichl tion from any source, with activity in the left hemi-
1981). A recent study by the Foote, Cone, and Beld- sphere higher in decision making situations.
ing ad agency found, however, that recall of good The situation in which many commercials are
"feeling" commercials, created through the skillful viewed may be characterized as nondecision making.
use of music and visuals, was equal to that obtained In watching TV we make decisions about, and are
with good "thinking" commercials, created through interested in, programs, not commercials. The com-
the rational presentation of product attributes (Hon- mercials we do see are often for brands in product
omichl 1981). This paper further suggests that the categories we may not be interested in for one reason
background features of commercials can influence or another. We may have recently purchased in that
product choice. It is argued that the positive emotions category or we may not be interested in the whole
they generate become associated with the advertised product category in the first place. Even where pur-
product through classical conditioning. chase in a product category is contemplated we may
The actual choices people made in the nondecision not be attentive to product information in a relevant
making conditions in these experiments would sup- commercial because the product may be an unimpor-
port a conditioning interpretation. To further test this tant one.
interpretation physiological recording could be used This suggests to the advertiser that an audience
in future research. The right hemisphere of the brain may be largely comprised of uninvolved potential
is the nonverbal, musical, and intuitive side; the left consumers rather than cognitively active problem
hemisphere is the analytical and verbal side (see Kas- solvers. Reaching them through emotionally arousing
sarjian 1981 for a summary of consumer research in background features may make the difference be-
this area). The results of this study suggest that re- tween their choosing and not choosing a brand.

REFERENCES
Brewer, W. F. (1974), "There is No Convincing Evidence for (1981), "Consumer Psychology," Working Paper
Operant and Classical Conditioning in Humans," in W. B. Series, Center for Marketing Studies, U.C.L.A., Paper No.
Weiner and D. S. Palermo, eds., Cognition and Symbolic 104.
Processes, Hillsdale, NJ: Erlbaum. Kroeber-Riel, Werner (1979), "Activation Research: Psycho-
Carlsmith, J. J., P. C. Ellsworth, and E. Aronson (1976), biological Approaches in Consumer Research," Journal of
Methods of Research in Social Psychology, Reading, MA: Consumer Research, 5 (March), 240-250.
Addison-Wesley. Krugman, H. E. (1965), "The Impact of Television Adver-
Engel, J. F., R. D. Blackwell, and D. J. Kollat (1978), Con- tising: Learning Without Involvement," Public Opinion
sumer Behavior, Hinsdale, IL: The Dryden Press. Quarterly, 29 (Autumn), 349-56.
Ericson, K. A., and H. A. Simon (1980), "Verbal Reports Lutz, R. (1977), "An Experimental Investigation of Causal
as Data," Psychological Review, 87(3), 215-251. Relations Among Cognitions Affected and Behavioral In-
Fishbein, M. and J. Ajzen (1975), Belief, Attitude, Intention, tention," Journal of Consumer Research, 3 (March),
and Behavior: An Introduction to Theory and Research, 197-208.
Reading, MA: Addison-Wesley. Nisbett, R. E. and J. D. Wilson (1977), "Telling More Than
Greenwald, A. G. (1980), "The Totalitarian Ego: Fabrication We Know: Verbal Reports on Mental Processes," Psycho-
and Revision of Personal History," American Psycholo- logical Review, 84 (3), 231-259.
gist, 35(7), 603-618. Olshavsky, R. W. and D. H. Granbois (1979), "Consumer
Honomichl, J. (1981), "FCB: Day-After Recall Cheats Emo- Decision Making, Fact or Fiction?" Journal of Consumer
tion," Advertising Age, 52(no. 20), 2. Research, 6 (September), 93-100.
Kassarjian, H. (1978), "Presidential Address, 1977: Anthro- Pollay, R. W., J. Zaichkowsky, and C. Fryer (1980), "Reg-
pomorphism and Parsimony," in Advances in Consumer ulation Hasn't Changed TV Ads Much!" Journalism
Research, H. Keith Hunt, ed., 5, Ann Arbor, MI: Asso- Quarterly, 57 (3), 438-446.
ciation for Consumer Research, 12-14. Schiffman, L. G. and L. L. Kanuk (1978), Consumer Behav-

100 / Journalof Marketing,Winter1982

This content downloaded from 175.159.14.205 on Wed, 24 Sep 2014 06:20:52 AM

All use subject to JSTOR Terms and Conditions
 ior, Englewood Cliffs, NJ: Prentice-Hall. (1980), "Feeling and Thinking: Preferences Need
Weitz, B. and P. Wright (1979), "Retrospective Self-Insight to Inferences," American Psychologist, 35, 151-175.
on Factors Considered in Product Evaluations," Journal of Zanna, M. P., C. A. Kiesler, and P. A. Pilkonis (1970),
Consumer Research, 6 (December), 280-294. "Positive and Negative Attitudinal Affect Established by
Zajonc, R. B. (1968), "Attitudinal Effects of Mere Expo- Classical Conditioning," Journal of Personality and Social
sure," Journal of Personality and Social Psychology Mon- Psychology, 14 (no. 4), 321-326.
ograph Supplement, 9, Part 2, 1-27.

mr- %k

wiFACE
with
erer
Ferber

Readings In The Analysis Of Readings In Survey Research

Survey Data Robert Ferber, editor
Robert Ferber, editor 604 pp. 1978
249 pp. 1980 $1 0/member $13/nonmember
$16/member $24/nonmember
A collection of readings which form
Key pieces of the published literature an extension of the special issue on
concerning applications of multi- survey research of the August 1977
variate and related techniques to sur- issue of Journal of Marketing Re-
vey data, and new, innovative ap- search. The articles focus on three
proaches to the analysis of survey aspects of survey research: sampling,
data are brought together in this book questionnaire preparation and data
of previously published articles. collection. An extensive bibliography
Emphasis is given to recent material is included.
although some of the classics in the
field have also been included. Biblio-
graphies follow each piece, to stimu- TO ORDER call or write Order
late the researcher to go further in Department, American Market-
ing Association, 250 S. Wacker
examining the various techniques. AMERICAN
MARKETING Drive, Chicago, IL 60606, (312)
A$OCIATION
648-0536.
L J

Effects
of MusicinAdvertising
/ 101

This content downloaded from 175.159.14.205 on Wed, 24 Sep 2014 06:20:52 AM

All use subject to JSTOR Terms and Conditions