TIte Spai,sh Jeuroal of Psvcholog~ Cupynght 2 0 ( 1 1 1 w TIre Spanislr Joarnal of Psychology

2 0 0 1 vol 4 No 1 4 ~ f4 1 1 38-74 1 6
Consumer Preferences and Brand Equity Measurement of Spanish
NationM Daily Newspapers: A Conjoint Analysis Approach
Jess Varela Malloti. Antonio Rial Boubcta, and Teresa Braa Tobio
University of Santiago de Compostela
Braud is a product attributc thai. for many typcs of goods or services, makcs a major
COrItObOtIOrI lo consuTner prefcrenccs. Con 0 1 1 1 1 analys 5 5 a tisti1 techniquc for thc
asscssrnent of brand values for a taven consu ma or group of consumers, l u ihis paper,
an appl Canon of ceujoin anaiysis u, 1 1 w estimaion of Stand values o tSe Spanish daiiy
ne~vspaper rnarket is reponed. Fnur ncw spapcr altributes Web c( ,flsidred: braud ( i.e..
new spaper Llame), pric ( 0 .60 , lOS. or 1 .50 euros). S,,oday sopplccnent ( yesnu). ami
/ci/v pci/oid ( ves/no). A tolal of 51 0 regular readcts of tSe nationa) prcss. sratified by
ago aud sex. w ere asked to rank 1 6 profiles rcpresenting an orihogonal fractior, of ihe
possihle auribute Level combi natons. Braud w as by tau ihe mos importan altrihute.
w hei~cas ptice liad uegligi SIc e ( tea. More generally. ( Se results con ni tOe uri?ity of
con;nirit anaLysis br assessing Stand equiy u tOe ncw spaper markel and for estimatirw
a
tSe relative iroportance of tIte varjeus atrihutes 1 .0 dllerent suhgroups of consumcrs.
CO, ivtifli&r belio to,: hin, cl ecu t~; coooiit oii.tilv&,s
La marca es el atributo de un producto o servicio que puede ejercer gran influencia en
las preferencias de los consumidores. E l anlisis conjunto es una tcnica muy til para
conocer el valor que una marca tiene para un grupo de consumidores. E n este trabajo
se presenta una aplicacin del anlisis conjunto en la estimacin del valor de la marca
en el mercado de la prensa nacional espaola. Se han considerado cuatro atributos de
un peridico: la marca ( el nombre del peridico), el precio ( 0 .60 , 1 .0 5 1 .50 euros), el
Suplemento Dominical ( con o sin suplemento), y el Cuadernillo Diario ( con o sin
cuadernillo). La muestra, compuesta por un total de 51 0 lectores habituales de prensa
nacional, fue estratificada por edad y sexo. Los sujetos tuvieron que ordenar 1 6 estmulos
que representaban una fraccin ortogonal de todas las combinaciones de atributos y
niveles posibles. Los resultados muestran que la marca es el atributo ms importante,
mientras que el precio tiene un efecto mnimo. Adems, stos confirman la utilidad que
tiene el anlisis conjunto para conocer el valor de la marca en el caso de un peridico,
asi como para estimar la importancia relativa de distintos atributos en diversos grupos
de consumidores.
Palabras clave: comportamiento del consumidor, valor de la marca, anlisis conjunto
fis w ork w as carried nut w ih financial support frorn tOe Furopean Union ( FE DE R 1 F0 9l-0 l58) and tOe collaboration of Sondaxe
( Crupo Voz).
Addrcss correspondence tu: Prof. Dr. Jess Varela Mallou. Opto. Metodologa de las Ciencias del Comportaniettu. Facultad de
Psicologa. Universidad de Santiago de Compostela. [570 6 Santiago. A Corua ( Spain). E -mail: rntsuso@ usces Http: w w w .usc.es/psicom
48
CONJOINT BRAND E QIJITY 49
In both the academie and busines.s werlds, it 5
newadays widcly recegnized that a prestigious braed
iniage is critical lo the success of most types of product
er service. Brand equity -defined by Farquhar (1989) as
the added value supplied by the braud (o the product-
is often a companys key asset, exceeding the value of lis
financial asse(s. Hew cisc can Wc explain ihe fact that
Phulip Monis bought ihe Kraft braud for more than 6 times
ts estimated financial value, er (he fact thai the mashead
of a national newspaper rnay cest up te $5 0 million?
Streng brand equity allews dic company to negetiate from
a pesition of srength with distribuers, te take up dic lead
pesition in a given predLct caiegory, to meve witb relative
case into ciher product categories, er to opt fer ce-
branding in association with other prestigleus brands. In
elber words, dic product is no longer king, and ihe
success of companies ends up depending en marketing.
Such success is, cf ceurse, not easily achieved. Numercus
recent siudies have indicated thai braud equity develeps
gradually over time, and requires careful management II
it is te beceme a valuable asset (Aaker, 1991, 1996;
Kapferer, 992; Kapferer & Theenig, 991; Low &
Fullerton, 1994: Park & Srinivasan, 1994; Shocker,
Srivastava & Ruekcrt, 994).
In ibis context, accurate and ebjcctive assessment of
brand value lo consumers (and rhus, (he value of the brand
te the company, j.c., brand equity) becomes a key gea fer
managcrs. One mcthed for the assessmcnt of braud value
is conjoint analysis. This rucihod is based en Lancastcrs
(1 966) approach to consumer theery, within which the
censumer is considered an active dccisien-maker.
Specifically, ihe censumer is seen as a preducer of
satisfactions, and goods (products or services) are
considered attribute seis. Accerding lo this approach, any
preduet oc servce can be broken down nCc a basic
funetional utility (the basic service) and a set el secondary
utilities ef varjeus typcs (aesthctic, social, cultural, safcty,
gL I arantee, prestige) thai supplemcnt thc basic service
(Leuden & Della Billa, 198% Wilkie & Pessenier, 1973).
Ihe sum el dic partial utilities (partworth) adds up to thc
total utility of the preduct, as perceived by a given
censumer ci group of consumers. Frem the theoretical
framewerks of Rosenberg (195 6) and Fishbein (1967), thc
tota] ulility of a product is a key dc(errninant of the
probabiliy of its purchase. Consideratien ef the individual
partial utilities wi II thus aid understandi ng the relative
impertance of dic individual atiributes. For example. if
ene of the attributes under sudy is brand, wc can estimate
the extent te which this atti-ihute gevcrns purchase choice,
Ihus moving inte dic ficd el decision models, and more
specifically, thai of cempensatory decisien-rnaking, as
extensivcly censidered in ihe informatien integration tlcory
proposed Uy Andersen (1974).
In ibis pape, we apply cenjeint analysis t.e the estimation
of brand values in the Spanish daily newspaper market.
Conjoint Analysis
The letal utility U of a given product can be considered
te be the sum of the partial utilities of the various altributes
of thai produc:
U = Zu (xi),
=1
( 1 )
where u~ is the partial utility funetion ter altribute 1, and x
is the perceived level of attribute i in that product. Given
ebserved values of U (e.g., as inferred frem preferences
exprcssed Uy the consumer) aud known values of x~ , conjeint
analysis uses rcgression procedures te estimate nr Fuji delalis
of diese procedures can be seen in Orcen and Wind (1975 ),
Varela and Braa (1996), or Huber (2000). In ihe present
study, we used the program Conjoint (Nerusis, 1994).
For each attribute, Conjeint additionally calculates
reladve impuriance (w
1), which provides a nieasure of dic
iniportance atached lo that attributc by the censumer or
group of consumers under study:
w.=n
t[Maxtupinu,;]
x 100,
(2)
whcre u ~ is the value of at level j.
L i
The principal aim of tUis study was te apply conjojut
analysis te the assessment of consumer preferences as
regards major Spanish national daily newspapers, with
particular reference lo brand (i.e., the naine or masthead of
the newspaper). Te cur knowledge. this is the firs application
of this approach w the assessmen of brand value in Ibis
produc catcgory. Additionally, estimation of thc partial
utilities of each attribute considered can be expected te
facilitarc producr design, whcreas evaluarion of partial
utilities in diffcrent censumer groupings (sex, age, etc.) rnay
facilitate identification of markct segmentation.
Method
Participants
Thc sample conprised 5 10 habitual readers of majer
natienal daily newspapers, al of whom werc residcnt in
Madrid and aged betwecn 18 and 74 years. A randem
sampling precedure was used, with sratification by age
group and scx. Rere were 323 men (18-24 years: 20: 25
34: 75 : 35 -44: 83: 45 -5 4: 62: 5 5 -64: 33; ever 65 years: 5 0)
aid 187 womcn (18-24 years: 28; 25 -34: 5 6; 35 -44: 46:
45 -5 4: 24; 5 5 -64: 20; over 65 years: 13). Thc sample was
alse balanced accerdingLo participants prcferred newspapcr
(ncwspaper A, B, C, or D), taking into acceun thc data
properlioned by thc Estudie General de Medies (General
Media Study) corresponding te 1998. Sample size was
5 0
VARE LA. RIAL. AND BRAA
sufftcient <o ensure a95 % ceufidence level aud a maximum
estimated error of 3%.
Stimulus
Evidently, consumcrs evaluaticus of newspapers are
influenced Uy a large number of variables (altributes).
Howcver, eur experience of dic newspapcr market in Spain
suggested censideration of feur variables only, namely. braud
(i.c., name), price, Sunday supplement (yes/no), aid daily
central pulleul (yes/ne). Te naintain confidentiality, we refer
te thc four ncwspapers ([ he feur nen-sperts dailies with
largest readership in Madrid) as A, B, (2. and D.
Given dic large number of types that would resul from
consideration of al attbute x level combinations (4 x 3 x
2 x 2), consumer preferences wcre directly determined only
for an orthogonal fraction of the complete factorial desigu
( Kirk, 1982), namely, dic 16 types detailed in TaiMe 1.
Procedure
Paticipants wcrc randornly selected for interview in thcir
bornes Uy dic random reutes procedure. with progressive
fulfillment of <he age, sex, and prcferred-ucwspapcr
stratification quetas (see Participants). Once thc participant
had agreed lo take pan, he er she was presented with sixteen
cards (14 x 8 cm), each represcnting ene of Ihe 6
ncwspaper types shown iii Table 1, and was asked te rank
them physically according te preference. The raukiwg task
typically lasted about 15 minutes.
Dura unalvsis
Thc dala were anayzed using [ he prograin Cenjoin
(Norusis, 1994). The observed prefercuces of each participant
(1 fur the most preferred product, 16 for Pe casi preferred
pro(luct) were entered as Sequence data. The altribule price
was modeled as Linear LessJ (i.c., (he pardal utility funetion
was assumed a priori lo be decrcasinglinear). Tle tmainiug
altributes were mWelcd as Discrete (i.c., as categorical vadables
wiuh no a priori assumptions abon which would be preferred).
Results
Tlc resuls of applicarien of conjeint analysis t.e <he data
for tle 5 10 participants are shown in Figure 1.
Ube model acconated fer 82% of <he total variance
(R = 0.911, p < .0001), indicating that u provides a fairly
accurate representatien uf participan preltrences.
The mos important dctcrninant uf prelerence was clearly
Urand (relative impurlance, 74.2%): Sunday suppleucnt
(8.4%) and daily pullout (6.6%) were much lcss impertan,
aud price had negligible effect (0.9%).
Ilie a priori assumplioi tha participans would prefer
low prices te high prices vas cenfirmed. Ncvcrlheless, <he
mean increase in utility per 75 -peseta drop in mice was a
negligible: 0.0213 utility unils.
Note Chal tUis relatienship allows us te estimale the ancun
of meney a reader weuld be prepared te pay Coy a given
improvement in a given attribute. For example. the difference
TaiMe 1
Attribute leveLs frr cadi of the 16 cards presented fo parlic;pwitr
Price
( u pesetas)
Wc oid y
Supplernent
25 0 W
175 W
175 Wo
25 0 W
175 Wo
25 0 Wo
lOO No
25 0 No
loo No
175 W
lot) W
lOO Wo
loo W
100 No
lOO XV
1<11) XV
F3rand
Stimulus 1
Stimulus 2
Stirnulus 3
Stimulus 4
Stinulus 5
Simulus 6
Stimulus 7
Stimutus 8
Stirnulus 9
Stimulus 10
Sti nulus II
Sri mulus 12
Slmulus 13
Stimulus 4
Stimulus 5
Stimulus 16
U
3
( 2
A
A
D
D
A
A
D
A
( 2
D
A
1 3
( 2
Daily
Pulleul
Wo
W
w
Wo
No
W
No
W
Wo
w
W
X V
No
w
No
W
Not. W = with; No without.
5 1
(2CNJOINT BRAND EQUITY
Relative
Importatice
Ut [y
+ +
1 7 4 . 1 7
+ +
Al tr i b u te
BRAND
. 8 220
1 . 1 24 1
1 . 6 7 5 6
1 . 9 7 7 8
BRAND
A
1 3
( 2
O
SUPPLEMENT
1
I
PtJLLOUT
1
PRI(2E
WEEKLY SUPPLEMENT
Nith s u p p e m e n t
Ni th o u t s u p p e m e n t
DAILY PULLOUT
Wi th p u l l o u t
Wi th o u t p u l l o u t
PRI(2E
2 30 pesetas
1 7 5 p e s e ta s
l OO p e s e ta s
( 2 ON STANT
Si g n i f i c a n c e . 0 0 0 0
Si g n i f i c a n c e = . 0 0 0 0
cunjoint analysis.
tu partial utility beween brand (2 arid braud D was 3.65 units
(s e e Ta b l e 2) . In p r i c e te m i s , th i s i s c q u i v a l e n t te a d i f f e r e n c e
o f 7 2 p e s e ta s (s e e Ta b l e 3 ) . Wi th i n th e p r i c e r a n g e
c o n s i d e r e d , th i ~ . . s u g g c s ts th a t p a r ti c i p a n [s wo u l d b u y b r a n d
(2 even il it cos 25 0 pesetas and braud D cos 100 pesetas.
Th c to ta l p r e d i c l e d u ti l i ty e l th e i d e a l n e ws p a p e r (b r a u d
(2, with weekend s u p p l c m e n t, d a i l y c e n tr a l p u l l o u t. a n d p r i c e
1 0 0 p e s e ta s ) wa s 1 0 . 8 2 (1 . 6 7 + 0.45 +0.16 + 0 . 0 6 + 8 . 4 6 ,
wh e r e 8 . 4 6 i s th e c o n s l a n t) . Th e c f f e c t o f i n c r e a s i n g p r i c e
t~ rom 100 to 25 0 p e s e ta s wo u l d b e m i n i m a l (o v e r a i l u ti l i ty
1 0 . 7 8 ) . Th c to ta l p r e d i c te d u ti l i ty o f b r a n d (2 wi th th e we r s t
p o s s i b l c a ttr i b u te s e t (n o wc c ke n d s u p p l e m c n t, n o d a i l y
p u l l e u l , p r i c e 25 0 p e s e ta s ) wa s 9 . 5 5 . TUi s wa s s Ci Il h i g h e r
Ih a n th c to ta l p r e d i c te d u ti l i ti e s f o r Ih e Ie a s t p r e f e r r e d b r a n d s
(b r a n r i s A a u d D) wi th th e b c s t p o s s i b l e a t< r i b u < c s e l (s e e r e s
8 . 3 1 a n d 7 . 1 6 , r c s p e c ti v c l y) . Th i s a g a i n i l l u s tr a te s Ch e c r i ti c a l
i m p o r ta n c e o f b r a n d .
1< should be stresscd that <hese q u a n ti f i c a ti e n s o f b r a n d
value are relative, in thc sense that <he value of a braud is
q u a n ti f i e d r e kti v e < e th a t o f th c o th e r c o m p e ti n g b r a n d s . Th u s ,
Ta b l e 2
DTerences ja tite
Ncwspape rs
I-arlial Uilhy of Brand between tit Four
Table 3
Diflerences in tlw Partial Utility of Branbetween tite Parir
Newspape rs, Ex;~ ressed in Ter,ns ofrice
Br a n d
A 0 . 8 220
A 1 3 ( 2 D
< 1
1 3 1 . 1 24 1 1 . 9 4 6 1 O
( 2 . 6 7 5 6 2. 4 9 7 6 0.5515 0
U 1 ) 4 7 7 8 1 . 1 5 5 8 3 . 1 0 1 9 3 6 5 3 4 <1
Br a u d
A 0 . 8 220
A B (2 [ 3
0
13 1.1241 91.3 0
( 2 1 . 6 7 5 6 1 1 7 . 2 25 . 8 0
3 [9 7 7 8 5 4 . 2 4 5 . 6 1 7 1 . 5 2 0
.4 52 <)
. 4 5 20
1 8 . 3 5 II
++
6 . 6 1 II
++
.86
13
. 1 6 28
. 1 6 28
. 0 2) 3
. 0 4 26
. 0 6 3 9
. 0 21 3
8 . 4 6 27
Pearsons R = . 9 1 1
Ke n d a l l s r = . 7 8 3
Figure 1. G e n e r a l rcsuLts uf Ihe
52
VARELA. RIAL. ANO BRAA
Ta b l e 4
Rlative hnportancc (iv,.) Q/ each Atiribute br
Participattt Subg roups
Braud
Delmcd
Price
bv Se
., Age, md Nnvspapr Readh ~-Freq~e;tcv
Su p p l c m c u t (2. Scctiou
Se x
To ta l
Me n
Wo m e n
74 . 7
7 7 . 6 7
6 9 . 4 9
tI. 8 6
0 . 7 7
0 . 9 4
8 . 3 5
7 . 6 8
1 8 . 2
6 . 6
3 . 8 7
. 3 6
Ag e
8 2 4 Yc a r s
25 34 Years
35 44 Years
45 5 4 Ye a r s
5 5 64 Years
=65 Years
43.82
7 3.69
77.47
75 ,43
75 .7 1
65 .8 3
3.5 9
2.88
3.61 1
3.34
3.61
3.29
26.29
4 .71
4.73
9.5 9
6.5 5
29.83
26,29
8.72
4.2<)
.64
4 . 4
.05
Re a d i n g F r e q u c u c y
E very Da y
5 6 Da ys /We e k
3 4 Days/Ncek
2 Da ys /We e k
Ouy Suudays
Ta b l e 5
odia! Urilities ofcadi Leve! olead, Attribute u Participant Subgroups I)efi;ed lw 5kv. Ag, and Nwspapr Rcading-Prcquetcy
Braud A lraud 3 Braud ( 2 Braud D With Without Ni th X v i th o u t 7 5 25 0 1 1 ) 0
Su p p l e we u t Su p p l e m e n t Pu f l e u s Pu e u t Pesetas lesctas Pesetas
0.82 .12 1.67 1.97 0.45 0.45 0.16 0.6 0.02 0,04 0.06
0.65 0.85 1.88 2.1)8 0.45 0.45 11.10 1).O 0,02 0.04 0.06
-3 1.61 1.30 1.79 0,44 0.44 0.27 0.27 0,02 0,04 0.06
0.49
.43
0.5 2
0.23
0.84
.1 0
0,5 2
.66
.29
0,6
1 .1 1 7
0.76
0.76
1.92
.92
.89
.82
1.00
0.79
2. 5
2.69
2.28
2,06
0.66
0.47
O,4
0.44
1 1 .54
0.42
1 1 .4 7
0 .4 1
1 1 .4 4
0.42
0.47
0.47
1 1 .2 4
1 1 .2
11,04

1)02
0.4 7
0.24
0. 2
13.04
0.
0.02
0.06
0.08
0. 1
0.09
<1 ,1 )9
0.1)5
0. 3
0. 5
0.2 1
8
1)18
O
..1 1 ,9
0.23
1)32
0.27
0.27
Re a d i n g F r e q u e u c y
Ev c r y Da y
5 6 DaysNcck
3 4 Da ys /X Ve e k
2 Da ys /We e k
0.70
. O
1 1 .94
0.95
0.72
.69
.60
2.21
.33
0,46
1.62
2.23
.44
.20
2.27
<1,42
0.35
0.40
0,5 5
0,42
0.35
1 1 .4 0
0.5 5
0.21 0.21
1 1 .7 ...Q7
0.18 0.18
75 .5 3
69.27
69.82
68.94
72.5 3
2.98
.23
2.29
5 ,1)8
5 .99
1 4 .4 1 )
7.94
9.5 2
19.43
20.94
7.08
1 .56
8.37
6.5 5
0.5 4
Se x
To ta
Me n
Nu m e n
Ag e
8 24 Ye a r s
25 34 Ycars
35 44 Years
45 5 4 Years
5 5 64 Ye a r s
=65 Years
(1.1)9
0.22
0,1)5
0. 4
1 ). [ 7
0.45
0.09
1 1 .2 8
1 1 .2 6
0.67
0.4
1 1 .4 3
Ony Su u d a ys 0.60 1.18 .5 6 2.5 0.5 4 0.5 4 0.1)1 <> 111 0.5 11.3 11.46
(2ONJOINT 3RAND EQUTY
b e r e Ih e p a r [i a u i i y o f b r a u d D wa s n e g a l i v e (wi h r e s p c c [
te Ih e o tb e r b r a n d s ) ; b u t, o f c o ti r s e , tUi s d o c s u o t m e a n th a t
al! the money invesled so lar iii ihe markeling of uewspaper
D has been wasted, Que xvay el obtaining an assessrnent of
abselute braud equity woud be te include a generie produc
( j.c. a newspaper without a name) in Ihe auaysis.
( 2 enjoiu[ auaysis can also he used to assess <be importancc
of braud fer differcu< segrnents of lite market (i.e., groups of
censttmcrs). To [ I,s cud, we divided he sarnpe mb greups
oc tite basis of a series of sociodemographic characteristies,
ncuditg scx, age, aud uewspaper reading frequency. Within
eac characterislie, wc [ hen perlermed separate conjoin
analyses, with he resuls .summarized iii TaNes 4 aud 5. Table
4 lisIs he imperlauce of eacb altribute in each subgroup.
The auribute braud was <he mos impertaut in alt
participaut greups. l-lowever, a number of betwecn-group
difcrences ate appareul. Firs., braud was more importaul
lo mcii thau lo wotneu. Seceud, brand was more irnpor.aut
lo elder (25 - 64 years) Ihan lo youncr (18 - 24 ycar.s)
parlicipants. Third. braud wa.s more importaul te patlicipauts
who read a newspaper daiy litan lo participanCs wbo eny
read a uewspaper once or .wice a week.
Iaite 5 jsts partial utilities (for eacb participan group)
of eacb of tite evels el each of be four attributes. These
dala oller detailed information en the ljne tuning of
readcrs prefereuces. For example, Ihe preseuce of a daiy
pulleul. was raled most itighly by younger \VOrnen, aud xvas
ess important lo older parlicipauts.
Cousidering Ihe levels of [ he altribute braud, men
preferred braud (2, aud wemcn brand E. When participants
werc classified Uy age. braud (2 was preferred u ah age
groups, aldiough its pardal utiity vvith respec lo tite otiter
newspapers varied consideraby. Witen participans were
classiftcd Uy xvceky freqitency of ncwspaper reading, daiy
readcrs prelerred braud (2, whereas participau[ s who read
newspapers 3 - 4 limes a week preferred itraud 13; jn [ he
remainiug groups Itere was no marked diflerence in
prefercuce fer brads 13 aud (2.
Cenclusiens
Gur results cleary intlicate <bat braud is el majer
importance in tite Spanish natienal daily rnarket. l3rand was
tbe most higbly valued attribule in al participan groups,
rcgardess of age, sex, aud frequency of newspaper reading.
Titese resuls suggest that tite rnanagers of newspapers witb
reativcy wcak brands (itere, A and D) sheuld no< assign
significant reseurces te sucit extras as pullouts, and should
uet alemp te compele on price; ralher, <bey sitould
cencentate en creating braud cquity.
bis cen[ ra importance of braud is luy in iue wi<b
previous siudies. Park, .lawerski, aud Maclunis (1986) have
empitasized [ le importance of s[ ratcgic management el Ibe
coucept uf mage of tite braud. iuvelving tite definitiou of
<he Urajd image and ils introduction, elaberation,
fortifjcation, aud control over time, always in accordance
with ceusumer demands.
Siudies such as diese by Keller (1993), Cobb-Walgren,
Ruble, and Douthu (1995 ), Oreen & Krieger (1995 ), Varela,
Ria, Garca, Braa, aud Olea (1996), aud Rial, Varela,
Braa, aud Lvy (2(100) itave stressed <he irnportance of
bui!ding brand value in Ihe ighl of brand u~ ilities for
differcnt consuttner subgrottps. Ilie underlyiug assumptieu
of diese studies is thai dic braiid will oaly be of value lo
<he owuer of <be product il it is are ~ idy vauablc te <he
consutneti (2ouscquently, it is fundamenal te assess <he way
bbc value of lite braud is crca[ ed u consurncrs minds, aud
[ be way <bis turus into beitavion u tius ceuncc[ on, conjoint
aualysis is of particular value: <he utilities es<imated u chis
auaysis are a measure el Ihe subjective value litat a given
produc itas for a given participant, aud alse provide
iufurmaiou abeul lhe atributes en witicit Ihis value is based
(Webcr & Borcberding, 993).
Our results also provide evideuce of lite value of cenjoint
analysis fer market scgtcutatieu. Tite itnporanee assigned
by participan<s <o eacb atribute varied markedy witb sez,
age, and frequeucy of newspaper readiug. Thus, tite ideal
predutel for ene sector of tite market may be cnlirely different
frem titat br anetiter sector of tite market. Tbe panial utihities
eblained u cenjoiut analysis can be used as tools for
investigating these quesdeus ir detall, alone or in cenjunetien
wilb uller lecbniques stich as cluster anaysis er discriminaut
analysis.
In stmmary, yve wisit lo peint out sone advautages of
using [ be cenjoi it analysis method iii tUis conlext: (a) it
offers vadees pessibilities of data collectien, making it a
simple task fer <be subjecls, as wel as bigbly realislie;
(b) from preference data (nonmetrie), we ebtaincd mettic
data dial represens lite uti[ y or lite perceived importance
of each level of attribute, or of tite total prodoel, witicb
revcas tite relative value of tite different altributes in lite
competitive markct; (e) tite iuformaieu previded by cenjeint
analysis is capabc of predicting tite reactions of sutbjccts in
lite market, as Lite tuilities reiaby reproduce [ heir
preferences; aud (d) Ihis methedeogy provides resuts br
differen. consunter groups, and, titerefete, is a geed teol ir
market segmeutation. Furthermore, as seen ir <his paper,
cenjoint aualysis can be vcry useftsl te solve specific
prebems in market research and in erder te make strateic
e
decisious.
References
Aaketu 0. (199 ). Menagiag bread equity: Capitaldiag un tit
za/nc cf bread neme, Ne~v York: hce Prcss.
Aaker, 0, (1996). Buihiiug smttg brttuds. Ne w Yo r k: F r e e Pr e s s .
Audersou. N -1. (974). Nfbrazatio;; integratica heorv, A tre!
s,rvev. San Francisco, (2A: Freeman.
54
VARELA, RIAL, ANO BRAA
(2ubb-Nagren, (2.J,, Rube, (2,A., & Doutitu. N. (995 ). Braud
equily. braud prelereuce, aud purcitase inten Icurual uf
Advrtishtg, 24, 25-40.
Fishitein, M. (967). A beitavioral heory appruach tu relaious
itetw eeu iteiet?s abeul an objec[ aud lite attitude lowards tite
objecl. u Nl. Fishheiu (Ed.). Reudings tu atiltude uad dturv
nesurentat ( pp. 389-399). New York: Niey.
Farquhar, PH. (1989). Managiug braud equily. Merkbzg Researel.
1, 2 4 -33.
Oreen. RE., & Wind, Y. (1975 ). New ways te measure consurner
judgmeuts. f1arz.~erd Bu,, ,ass Rcview, 53. < 17 -1 7.
Gree,i, RE., & Kriegcr, AM. (1995) . Atlribure intpurtanc ~ vcgitts
modification i u assessiug a brauds cotnpe[ ilive poleulial.
Marketing Seicace, 14, 2 53-2 70 .
Huber, E. (2000). (2uujuiat nteesuretnent. Methuds ami upplientiuas.
New York: Springer-Verlag.
Kapfcrer, J .N. (1992). Les marques, capital le lnterpri.se. Paris:
Les diions dCrgauisaiou.
Maplerc; IN., & 1ttneuig. (2. (1991). La tzuaqae. Patis: McGraw-
Hill.
Keer, KL. (1993). (2onceptualiziug. measuring, aud mauagiug
cuslemer-hased braud equity. Juurnal ufMarlcziag,.57 , 2 2 .
Kirk, 1. (982). E.vperimnentel desigus Prucdures br tite bhaviural
~ ( 2 ~ cd.), Monterrey. (SA: Brueks-(2ue.
Laucaster, Kl. (1966). A new approaeh tu consumer theory. Juu,nel
of Political Eeununty, 74 , 1 32 -1 57,
Leuden. D.. & Della Bitia, Al. (988). (2unsumer hehaviur:
(2oneepts and applieetiu;ts. New York: McGraw-hill.
Low, OS., & Fullerlen, RA. (994). Brands. brand management
aud lite braud manager system: A critical-itisurical evatuation.
laurnal ufMarketing Rsarelz, .?l. 2 71 -2 88.
Norusis, Mi. (994). SPSS, Prufessioua Staistics 6.1. (Siticago:
SpSS,
Park. (2W., Jewerski. 13.1.. & Macnnis, 1),]. (1986). Sirazegie
braud eotteep< image managemeul. Juvnal uf Marketi,,g, .50.
35 -145 ,
Park, (2.N.. & Srinivasau, V, (994). A servey-based uietitud for
measunng aud uuderstaudiug Utaud equiy aud lIs exeudibiitv.
Juarnul uf Murktiag ke.vearch .3/, 271-288.
Rial, A.. Varela. i.. Braa, T.. & Levy, iP. <2001)). III valor de la
marca a partir (e su rel:tcion con el consut nt i don PsP uftierna,
12. 2 4 7-2 54 .
Raseaberg, Al.]. (95 6). (2ugniti ve srucutre aud aliuciual afret
luarual ufAbnu,,aa aud Sucia! Psveholugy 5 3, 367-372.
Sitocker. Ab.. Srivastava, V.. & Ruckerl. R.W. (994). Uhallenges
aud uppotluu ti es fac i ng braud managemeul: A u ut roduction tu
lite speeia issue. ,Iuurnel uf Murketiug Researc,, /6, 49 15 8.
Varela. 1., & Braa, 1. (996). Anlisis conjuz fu aplicado a la
urestigacicu, t comercial. Madrid: Pi rulide.
Varca, .1.. Rial. A., Garca. (7., Braa, T.. & Gica, .1. (996). Una
medida de a implicacin de os consumidores con Las mareas
eomerciaes, EsiculPeine, 8. 3. 5 43-5 5 1.
Neber, M.. & Burciterding. K. (1993). Behavioral intiucuces un
w eigitl judgrnens u mui altribute decisiun making. Puropeat
Joarna! ofOpzratonal Rescarch, 67 . 1 1 2 .
Nill=ie, NL., & Pcssemier. EA. (973). ssues u markeliugs use
of mulli attribute model s. Juurael of Marketing Reseereh, 10.
4 2 84 4 1 .
Received September 13, 1999
Revtson received June 19, 2000
Acceped Sepember II, 2000