Professional Documents
Culture Documents
.
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.
Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to Journal of
Advertising.
http://www.jstor.org
Children's
Television
Advertising
Brand
A
and
Choice:
Laboratory
ALAN RESNIK
Experiment
I.
BRUCE L. STERN
Journalof Marketing.
IBruce L. Stern is currently an Assistant Professor of Marketing at
Portland State University. His D.B.A. was received at Arizona State
University. Dr. Stern's articles have appeared in the Journal of
ABSTRACT
This experimentattemptsto investigatethe vulnerabilityof children
to television advertising.Subjects between six and eight years of age
were exposed to commercialstimuliundervaryingconditions. Groups
were randomlyassigned to view segments of "Jeannie," a network
commercialof a
children'sshow, eitherwith an experimenter-produced
previously-unknownbrand or a control ad inserted into the program
context. In addition,the effects of traditionalmethodologiesemployedin
relatedliteratureto one which minimizesreactanceeffects were studied.
Children'sbrandchoices werefoundto be influencedsignificantlyby the
communicationof the commercialmessagefor the previously-unknown
brand.
11
JOURNAL OF ADVERTISING
such advertisingis beneficialin thatit creates awareness of products that potentially increases one's
satisfaction.This issue has largelyremainedan emotional one with industrysupporterson one side, and
organizationslike Action for Children'sTelevision
(ACT) on the other side. The FTC is now wedged in
the center of the controversy, and is underpressure
to make public policy decisions despite the dearth
and lack of clarity of available experimentaldata.
The vast majorityof studies attemptingto measure the effect of television advertisingon children
have employed survey or observationaltechniques.
The survey methodgenerallyinvestigatedsuch questions as: How much television does your child
watch? How often do you yield to these requests?
How do you feel television advertising influences
your child?
A second approach to assessing the impact of
television commercialson childrenis throughdirect
observation. This technique generally employs
motherstrainedto unobtrusivelyobserve one of their
children watching television commercials. Generally, observation periods are determined by the
child's normalviewing habits and usually last for a
minimumof six and a maximumof ten hours over a
seven to ten day period. Participatingmothers are
often paid for their efforts.
Laboratory experimentation is the third
technique used for assessing the effect of television
commercials on children. Within this setting, children are often asked to view a video-tapedsegment
of a television programwith commercial(s)inserted
by the experimenter.The child is then subjectedto a
post-viewinginterviewand/ora productchoice situation.
Each research technique has its own strengths
and weaknesses. The survey and observational
techniques lack scientific documentation and are
subject to experimenterbias, while the laboratory
method has been criticizedfor the artificialityof the
viewing situation.
PREVIOUS RESEARCH
It is surprisingthat more research has not been
publishedin this highly controversialarea. Existing
12
Viewing Behavior
Ward, Levinson and Wackmaninvestigated the
areas of attention to, consequences of and alternatives to commercialviewing (12). Using the trained
mother observational approach, the researchers
found that approximatelyone-half of the 65 children
were paying full attention, while one-quarterwere
payingpartialattentionto a programbefore the onset
of a commercial.The child usually did not changehis
viewing pattern, but the degree of attentiondid vary
with age. Their study also supportedWard'sfindings
that children of all ages were more interested in
13
JOURNAL OF ADVERTISING
VULNERABILITY STUDY
The specific purpose of this experimentis to determinewhethermere exposure to a non-informative
product advertisement on television is sufficient to
create a desire for that product. To this end, the
investigators will first obtain a baseline measure of
the vulnerabilityof childrento television advertising,
and second, will establish a check on the reactive
effects of the traditionallaboratorymethods used to
research this area. The null hypotheses to be tested
are as follows:
I. There is no significantrelationshipbetween a
child's selection of a particular brand and
whetheror not it was advertisedon television.
2. There is no significantdifferencebetween the
research findings obtained by utilizing quasinatural and traditionalartificialexperimental
methods.
Sample
The subjectsfor the study were chosen from children in the Marquette,Michiganpublic school system. It was impossible to obtain names and phone
numbersof childrenfrom school principalsbecause
of a policy established by the superintendent.This
problem is pervasive for studies using children as
subjects. In lieu of an idealrandomselection process,
five hundredletters were sent home with students in
14
Results
The data collected in this study are dichotomous,
that is, a S chooses the advertised or unadvertised
brand of potato chip. Hence 2 x 2 contingency
analyses were used to test both hypotheses. Because
some of the cell sizes containedless than five observations, the Fisher Exact test was used in lieu of
chi-square analysis.
The contingencytableassociatedwith hypothesis
one, concerning the relationshipbetween seeing an
advertisementand choosing the brandpresented, is
shown in Table 1.
sion.
Table I
AN ANALYSIS OF BRAND CHOICES
IN THE QUASI-NATURAL AND
QUASI-NATURAL CONTROL CONDITIONS
Conditions
Quasi-Natural
Quasi-Natural Control
Number of
advertised
brand choices
Number of
non-advertised
brand choices
8
1
6
13
The contingencytableassociatedwithhypothesis
two concerning the viability of the two divergent
methodologies is presented in Table 2.
15
JOURNAL OF ADVERTISING
of consumerbehavior, or in most areas for that matter, is the eclectic approachto its exploration.
The test of the second hypothesis demonstrated
that the reactance effects associated with the traditional condition did little to produce statisticallydifferent results compared to the quasi-naturalcondition. However, when the Ss were told to watch the
television, a higher proportionof them (73 percent)
chose the advertisedbrand, comparedto 57 percent
of the quasi-naturalSs. Although the Fisher Exact
score was not statistically significant, there was a
difference in brand choice between the two groups
which would have to be attributed to the special
instructionsgiven to the traditionalSs and/orchance.
The lack of statistical significancecould well be due
to the smallcell sizes, and therefore,deserves further
investigation. If successive samples yielded similar
results, the findingsshouldreflecta moderatecaution
to futureresearchersin the area. The implicationsare
that placing childrenin an unnatural,forced (versus
optional)viewing situationmightproducesubtle contamination toward the advertised products which
might overstate the effects of normalviewing situations on brand choice behavior.
Table 2
AN ANALYSIS OF BRAND CHOICES
IN THE QUASI-NATURAL
AND TRADITIONAL CONDITIONS
Conditions
Quasi-Natural
Traditional
Number of
advertised
brand choices
Number of
non-advertised
brand choices
8
11
6
4
Discussion
The small sample size and limited scope somewhat limits inferences that can be drawn from the
study. However, the experiment has dramatically
demonstrated a generalized cause and effect relationship that is rarely present in the previous research. As reflectedin Table 1, 57 percent(8 of 14)of
the Ss in the quasi-naturalconditionchose the advertised brandcomparedto 7 percent (1 of 14)of the Ss
in the quasi-naturalcontrol condition. Although the
57 percentproportionis not impressivewhen viewed
in isolation, it comparesquite favorablyto the overwhelming non-preference for the same advertised
brand offered to the control group Ss. Apparently,
the non-advertisedbrand,by virtueof its brandname
or package, was inherentlymore appealingto those
who were not exposed to the experimentalmanipulation. Because of the statistically significantresults
and the tightnessof the experimentaldesign, it can be
said (keepinglimitationsin mind)that a child is more
likely to choose a brandthat he has seen advertised
on television over another previously unknown
brandthat he has not seen advertised.
The importantquestion remainingis why the advertised product is chosen. A myriad of potential
explanationsexist. Perhapstelevision lends credibility to advertisedproducts. Equally likely is that the
advertisedbrandwas fresh in the child's mind and it
was chosen due to unconscious familarity.Possibly
the child really believed the model's use of superlatives in describingthe product. The outcome may be
a result of a conditioned response guided by the
informationreceived (6,15). The list of possible explanationscould be vastly expandedwith little effort.
Unfortunately, although previous research gives
some clues concerningthe reasons for the choice of
the advertisedproduct,it would be difficultto justify
one explanationbased solely on their findings.Even
theoretical explorations of the area do not provide
in-depth or predicative foundations for evaluating
the process involved. One majorproblemin this area
CONCLUSIONS
The issue of children's vulnerabilityto advertising messages is central to determiningthe aggregate
effect of advertising. When one thinks of a child
watchingan averageof 25,000commercialsannually,
it is understandablethat there is great concern over
their potential effects. When one further considers
that the majorityof a child's television viewing is not
done duringchildren's programming(i.e., Saturday
morning) times but during prime adult hours, the
concern increases. Does constant exposure to similar messages increase or reduce their credibility?
Does a child, because of the sheer numberof commercialmessages, develop a type of cynicism toward
them as some researchresults suggest?The answers
to these and a myriadof other questions relate to the
malleabilityof a child's mind-the degree to which it
can be manipulatedby the advertiser.
This potential for manipulationhas made children's television advertising a highly emotional
topic. The scenario of the innocent child exposed to
hundredsof drugads, as an example, and developing
the philosophythat "drugscan cure all of Man'sills"
is an outgrowth of such emotion with no empirical
support. However, it is an undeniablefact that television plays a majorrole in a child's developmentand
such a scenario cannot be dismissed as utter nonsense. Of course childrenare not passive receivers of
all commercial messages, but we know very little
16
9. Rossiter,J. andT. S. Robertson."Children'sTV Commercials:Test-ingand Defenses," Journal of Communication, Vol. 24 (Autumn, 1974),
pp. 137-44.
10. Blatt, J., L. Spencer and S. Ward. "A Cognitive Developmental
Study of Children'sReactions to Television Advertising," Working
Paper:MarketingScience Institute, Cambridge,Mass., 1971.
11. Robertson,T. and J. R. Rossiter. "Childrenand CommercialPersuasion: An AttributionTheory Analysis," Journal of ConsumerResearch, Vol. 1 (June, 1974),pp. 13-20.
12. Ward,S., D. Levinsonand D. B. Wackman."Children'sAttention
to Television Advertising," WorkingPaper: MarketingScience Institute, Cambridge,Mass., 1971.
13. McNeal,J. U. "An ExploratoryStudyof the ConsumerBehaviorof
Children," in James U. McNeal, Dimensions of Consumer Behavior.
1969.
New York:Appleton-Century-Crofts,
14. Goldberg,J. and G. Gorn. "Children'sReactionsto Television Advertising: An Experimental Approach," Journal of Consumer Research,
REFERENCES
1. Breen, M. P. and J. T. Powell. "The RelationshipBetween Attractiveness and Credibilityof Television Commercialsas Perceived by
Children," Central States Speech Journal, Vol. 24 (Summer, 1973), pp.
97-101.
2. Frideres,J. S. "AdvertisingBuyingPatternsand Children,"Journal
of Advertising Research, Vol. 13 (Feb., 1973), pp. 34-36.
17