Using a Model’s Apparent Ethnicity to Influence Viewer Responses to Print Ads: A Social Identity Theory Perspective

Journal of Current Issues and Research in Advertising,Volume 31, Number 2 (Fall 2009).

Jeremy J. Sierra (Ph.D., New Mexico State University), AssistantProfessor of Marketing, Texas State University, San Marcos, TX.(email: [email protected])

Michael R. Hyman (Ph.D., Purdue University), Stan Fulton Chair ofMarketing, New Mexico State University, Las Cruces, NM. (email:[email protected])

Ivonne M. Torres (Ph.D., University of Houston), Assistant Profes-sor of Marketing, New Mexico State University, Las Cruces, NM.(email: [email protected])

Using a Model’s Apparent Ethnicity to InfluenceViewer Responses to Print Ads: A Social

Identity Theory Perspective

Jeremy J. Sierra, Michael R. Hyman, and Ivonne M. Torres

Although the efficacy of ads with ethnically resonant cues is well known, further examination ofviewers’ responses to print ads, based on the apparent ethnicity of models used in the ads, is needed.From a Social Identity Theory perspective, we examine a determinant of, as well as the direct andindirect effects of, ethnic identification with a print ad—created by a model’s apparent ethnicity—on intentions to buy the featured brand. Our empirical results, which rely on four extant scales anda new multi-item measure of ethnic identification with an ad, validate the positive relationshipsproposed in the structural model.

To help break through ad clutter, a company mayrun ads meant to persuade targeted consumers toidentify with those ads and featured brands. One ger-mane yet under-researched aspect of consumer iden-tification with ads pertains to models’ or actors’appearance (Brumbaugh and Grier 2006; Qualls andMoore 1990). For example, do readers’ responses to aprint ad or the featured brand depend on the admodel’s skin color (Eberhardt et al. 2004; White andHarkins 1994)? Utilizing a Social Identity Theoryframework, our research probes this question.

Consumers cognizant of a company’s efforts to tar-get their ethnic group tend to respond favorably tothat company’s ads (Cui 1997) and may develop posi-tive brand-related attitudes (Grier and Brumbaugh1999; Grier and Deshpandé 2001). Ads that depictethnic characters (e.g., Orthodox Jewish males in tra-ditional black attire) may stir targeted consumers—either rationally or emotionally—to affiliate positivelywith the featured brand (Koslow, Shamdasani, andTouchstone 1994). Clear ethnic cues in an ad—trans-mitted by special terms (e.g., idioms like average Joe,mosey along, or Holy cow!; well-known Yiddish wordslike kosher, chutzpah, or nosh), values (e.g., an extendedItalian family happily sharing a dinner), and images(e.g., a Mexican flag; cartoon characters like Charlie

Brown or Dagwood)—encourage ethnically resonantconsumers (i.e., people who embrace their ethnicity)to better recall and develop more positive attitudesabout that ad, which in turn may induce more favor-able attitudes and stronger purchase intentions forthe featured brand (Appiah 2001a; Elliott andWattanasuwan 1998; Forehand and Deshpandé 2001).

The effects of ethnic cues in ads on targeted con-sumers is consistent with Social Identity Theory (SIT)(Tajfel and Turner 1985); that is, when membership ina social group enhances people’s self-image (e.g.,higher levels of self-esteem as a result of ethnic groupmembership), SIT posits that people assign themselvesto social categories (e.g., Hispanic ethnic group), whichhelps them to define their self identity and may leadthem to respond favorably to stimuli (e.g., ads) thatrecognize their social category distinctiveness (Tajfel1978a, 1981). Although SIT clearly pertains to adver-tising scholarship and practice (e.g., Forehand,Deshpandé, and Reed 2002; Maldonado and Muehling2006), and is ideal for explaining viewer responses toads that use ethnic cues, it has not been applied tosuch studies. Thus, we use SIT to ground a study onthe direct and indirect effects of ethnic identificationwith a print ad, created by a model’s apparentethnicity, on intentions to buy the featured brand.

Our exposition proceeds as follows. First, we providean extensive review of the theoretical frameworks andexisting research regarding ethnicity and advertising. Next,we discuss Social Identity Theory and its applicability toethnicity and advertising studies, followed by the justi-

42 Journal of Current Issues and Research in Advertising

fication for the hypotheses. Then, the survey methodol-ogy and the statistical results are delineated. Finally, weoffer conclusions, implications for advertisers, studylimitations, and future research opportunities.

Theoretical Frameworks for StudyingEthnicity and Advertising

Table 1 summarizes eight psychological theories thatgrounded previous empirical studies on advertisingand ethnicity: Cultural Script, Distinctiveness, Ac-commodation, In-Group Bias, Identification, PolarizedAppraisal, Heuristic-Systematic Persuasion Model,and Elaboration Likelihood Model. Although research-ers have used these theories to explain how viewersprocess ad messages, identify with ethnic cues in ads,and respond to ads, SIT provides a more completefoundation for studying viewers’ responses to ethni-cally resonant ads because it posits that people iden-tify with groups based on an amalgam of their selfconcept (e.g., self identity, social identity, ethnic iden-tification with ads and/or ad actors), attitudes (e.g.,attitudes toward ads with ethnic cues, attitudes to-ward brands featured in those ads), and behaviors(e.g., dispositions toward buying brands featured inads with ethnic cues) (Reed 2002; Tajfel and Turner1985). In contrast, self concept is superfluous to theCultural Script, Heuristic-Systematic Persuasion, andElaboration Likelihood theories, and it is tangential tothe other five theories. Regardless of the consumerresponse precursor, such as ethnic majority/minority(as noted in Distinctiveness Theory) or type of ethniccue (as noted in Cultural Script Theory), SIT positsthat favorable attitudinal and behavioral responses toethnic-related stimuli stem from peoples’ understand-ing of their self concept. Thus, SIT should be able toprovide extensive and novel insights into viewers’responses to ethnically resonant ads.

Previous Empirical Studies on Ethnicityand Advertising

To inform a theory-based structural model onethnicity and advertising effects, we reviewed 25 em-pirical studies on ethnicity in an advertising context.Table 2 shows the most common research questionaddressed by these studies was response differencesamong ethnic groups—Blacks, Whites, Hispanics, andAsians—to ads with various ethnic cues, and the mostcommon ethnic cues in test ads were targeted atWhites, Blacks, and Hispanics, respectively. For allstudies, only two ethnic groups were depicted in thetest ads, which hinders the generalizability of reported

results (Brumbaugh and Grier 2006). Print and televi-sion ads were studied most frequently.

Most of the studies were grounded in a theoreticalfoundation (e.g., Distinctiveness, Identification) (seeTable 1 for full description of each theory). These frame-works certainly help to explain consumer responses toethnic-embedded ads, such as greater responsivenessto ads that reflect analogous cultural values, increasedtrustworthiness of actors of similar ethnicity, and en-hanced attitudes toward ads that imply advertisers’ cul-tural sensitivity to ethnic minorities. SIT implies thatconsumers’ identification with a marketplace stimulusstems from their self-concept, which in turn may in-duce favorable responses toward the source of thatstimulus; as none of the theoretical frameworks men-tioned in Table 1 focus on this phenomenon, an SIT-grounded model may explain consumers’ responsestoward ads with embedded ethnic cues.

Regarding constructs studied, a fuller understand-ing of consumer responses to ethnic-cued ads is ap-parent, as collectively the studies examined a plethoraof consumer behavior-related constructs (e.g., attitudetoward the ad, message comprehension, and purchaseintentions). However, no study has examined the re-lationship between ethnic identification with the ad,based on a model’s apparent ethnicity, and purchaseintentions of the advertised product.

The findings generally suggest that non-White con-sumers, relative to White consumers, were aware ofand responded more favorably to ads with ethnic cues(e.g., Forehand, Deshpandé, and Reed 2002). Regardingweak versus strong ethnic identifiers, same-ethnicityactors in ads were favored more by the latter group(e.g., Deshpandé, Hoyer, and Donthu 1986). Also, re-garding the effect of ads with ethnic language on same-ethnicity consumers, results indicate that Hispanics weremore likely (1) to use Spanish-language media, (2) tohave positive attitudes towards ads with Spanish ver-biage, and (3) to buy a product advertised with Spanishverbiage (e.g., Roslow and Nicholls 1996).

Findings about responses to ethnic actors in adswere mixed. Some research showed that Whites(Blacks) identified more with White (Black) actors thanBlack (White) actors, responded more favorably toads with White (Black) actors, and were more likelyto buy the featured brand when White (Black) actorswere used (Whittler 1989). In contrast, other researchshowed that Whites responded similarly to point-of-purchase displays with all Black models, all Whitemodels, or a combination of White and Black models(Bush, Gwinner, and Solomon 1974). Asian models inads favorably affected ad-related attitudes and pur-chase intentions of the featured brand for both Asians

Fall 2009 43

Table 1Previously Used Theoretical Frameworks for Advertising and Ethnicity Studies

Application toTheory Brief Description/Definition Ethnicity and Advertising Explanatory Power

Cultural Script Social communication (e.g., Because ads can convey Positive attitudes toward ads with(Triandis, ads) that portrays and/or the cultural values of ethnic resonant cultural scripts.Marin, communicates cultural groups, members of ethnicLisansky, and themes/values common to groups may be moreBetancourt 1984) a unique ethnic group. responsive to ads that

reflect the cultural values oftheir scripts (e.g., Spanglishin ads for barrio Hispanics).

Distinctiveness A person’s distinctive traits Relative to ethnic-majority Responses of minority ethnic groups(McGuire 1984) (e.g., Hispanic) are more Whites living in Beverly Hills, (e.g., Hispanics in Beverly Hills,

salient to him/her than the ethnic-minority Hispanics CA) to ads with resonant ethnicmore common traits of other living in Beverly Hills are cues.people in his/her environment. more likely to trust a

Hispanic actor in an ad.

Accommodation If person X becomes more Hispanics will respond Identification with resonant ethnic(Byrne 1971) analogous to person Y, then favorably to an ad (e.g., more actors/characters in ads.

it is more probable that positive AAD

, AB, and PI

B) with

person Y will like person X Spanish verbiage because(i.e., people tend to like other they attribute the use ofpeople with similar traits). Spanish to the advertiser’s

respect and compassionfor Hispanic culture.

In-Group Bias Bias towards a person in a Hispanics (Blacks) will Identification with resonant ethnic(Brewer 1979) similar social group shows evaluate Hispanic (Black) actors/characters in ads.

favoritism toward in-group actors in ads moremembers rather than favorably than Blackresentment toward out- (Hispanic) actors in ads.group members.

Identification People consider their degree Consumers will identify more Ethnic consumers who more strongly(Kelman 1961) of similarity with sources with ads that show an identify with resonant ethnic actors

(during interactions) and then actor of similar ethnicity in an ad also identify more stronglymake similarity judgments. than with ads that show an with that ad.

actor of dissimilar ethnicity.

Polarized Because cognitive schemas Whites will exaggerate the Responses of one ethnic groupAppraisal are more elaborate for in- positive characteristics of a to ad actors/characters of other(Linville and group members than out- positive Black role model in ethnic groups.Jones 1980) group members, out-group an ad; thus, Whites will

members are less likely to evaluate a Black spokes-fit these schemas when being person more favorably thanassessed; thus, in-group a White spokesperson ofevaluations are less extreme. similar character.

(continued)


Heuristic- Message credence is When exposed to an ad Heuristic or systematic evaluations,Systematic assessed either heuristically embedded with Hispanic stemming from embedded ethnicPersuasion (i.e., message claims (Asian) cues, Hispanic cues, used to assess credibilityModel assessed casually, as basic viewers may systematically of ethnically resonant ads.(Chaiken 1980) decision rules based on (heuristically) examine the

previous experiences message because theyare used) or systematically (dis)trust an ethnically(i.e., message claims (dis)similar source.scrutinized carefully).

Elaboration Attitude formation and Peripheral or central Some ethnic cues in ads, dependingLikelihood change follow one of two processing of an ad on the viewer’s ethnicity, mayModel (Petty routes. Central route message is influenced by encourage central or peripheraland Cacioppo processes require considerable the ad model’s and message examination.1996) thought and high elaboration; viewers’ ethnicity (e.g.,

peripheral route processes Chinese viewers maylead to message assessment peripherally [centrally]without thorough examination. examine an ad for a

Chinese product when anAsian [Hispanic] actor is used).

Table 1 (continued)Previously Used Theoretical Frameworks for Advertising and Ethnicity Studies

Application toTheory Brief Description/Definition Ethnicity and Advertising Explanatory Power

and Whites (Martin, Lee, and Yang 2004). Asians, Blacks,and Hispanics identified most with ads that used Blackactors, and Asians, Blacks, Hispanics, and Whites evalu-ated ads that used Black actors more favorably than adsthat used White actors (Appiah 2001b).

Although the extant literature offers many insightsinto the effects of ad model ethnicity on viewers’ atti-tudes, intentions, and behaviors, the psychologicalmechanism through which these effects occur remainsuncertain (Brumbaugh and Grier 2006; Eberhardt etal. 2004). Also, no published research on ad modelethnicity has examined a system of psychological con-structs suggested by a theoretically robust SIT frame-work. To close this research lacuna, we formulatedand tested a SIT-based structural model that delin-eates the direct and indirect (through attitude towardthe ad (AAD) and attitude toward the brand (AB)) ef-fects of ethnic identification with a print ad (EthIdenAD),via a model’s apparent ethnicity, on intentions to pur-chase the advertised brand (PIB). We also examinestrength of ethnic identity (SEI) as an antecedent ofEthIdenAD. A brief introduction to SIT follows.

Social Identity Theory

By addressing how people perceive and categorizethemselves and others, SIT describes how group af-

filiations may influence peoples’ behaviors (Tajfel1981). For example, how does being Hispanic influ-ence responses to ads with Hispanic actors? SIT relieson three premises: (1) people define and evaluatethemselves in terms of social groups; (2) the subjec-tive status of a social group determines if a person’ssocial identity is positive or negative; and (3) non-group members stipulate the frame of reference forevaluating another group’s status (Tajfel 1978b).

Social identity is determined by two socio-cogni-tive processes: social categorization and self-enhance-ment (Hogg, Terry, and White 1995). Socialcategorization differentiates between groups withincultures by producing group-distinctive perceptions andtastes (Tajfel and Turner 1985), such as when viewersrespond more favorably to television ads that depictmembers of their own ethnicity. Self-enhancementguides the social categorization processes so that groupor cultural norms and values substantially favor within-group members (Hogg, Terry, and White 1995).

SIT predicts that people form self concepts based onboth their social identity and their self identity. Socialidentity is derived from recognizing and acceptingmembership in a society, culture, or group (Tajfel 1981;Tajfel and Turner 1985). Such memberships, whichcontribute to self image and self satisfaction, help todefine a person’s self identity (Tajfel 1978b, 1981). Self

Fall 2009 45T

able

2S

um

mar

y o

f E

thn

icit

y an

d A

dve

rtis

ing

Lit

erat

ure

Ove

rall

Eth

-n;

Dat

aId

en.

Sta

tistic

alA

rtic

leR

esea

rch

Que

stio

nT

heor

yni

city

Sub

ject

sS

timul

iC

olle

cted

w/ A

dP

IM

etho

dsF

indi

ngs

Mus

eH

ow a

re a

ds w

ithN

o fr

ame-

Bla

cks,

233;

Prin

t ads

for

Per

sona

lN

oN

oT

-tes

tA

ds w

ith o

nly

Bla

ck(1

971)

only

Bla

ck m

odel

sw

ork

used

Whi

tes

colle

geci

gare

ttes,

inte

rvie

ws;

mod

els

do n

ot a

ffect

perc

eive

d by

Whi

test

uden

tsvo

dka,

ads

rank

edW

hite

con

sum

ers’

audi

ence

s?na

pkin

s, b

eer

on e

ffect

ive-

perc

eptio

ns o

f ad

ness

and

effe

ctiv

enes

s or

app

eal

over

all a

ppea

l

Bus

h,H

ow d

o W

hite

No

fram

e-B

lack

s,13

443;

End

-of-

aisl

e,O

bser

vers

No

Yes

Chi

-W

hite

s re

spon

d si

mila

rlyG

win

ner,

cons

umer

s re

spon

dw

ork

used

Whi

tes

Whi

teP

OP

dis

play

sre

cord

edsq

uare

to p

oint

-of-

purc

hase

and

to B

lack

mod

els

inad

ults

with

eith

er a

llsh

oppi

ng(P

OP

) di

spla

ys w

ith a

llS

olom

onin

-sto

refr

omB

lack

, all

beha

vior

Bla

ck m

odel

s, a

ll W

hite

(197

4)pr

omot

iona

lS

outh

-W

hite

, or

mix

inm

odel

s, o

r a

mix

of

mat

eria

ls?

wes

tern

of B

lack

and

supe

rmar

ket

Bla

ck a

nd W

hite

mod

els

US

Whi

te m

odel

s

Des

hpan

d é,

Do

diffe

rent

gro

ups

No

fram

e-H

is-

425;

Non

e; m

ail

Eth

nici

ty;

No

Yes

Dis

crim

-R

elat

ive

to w

eak

His

pani

cH

oyer

,of

His

pani

cs d

iffer

wor

k us

edpa

nics

Reg

iste

red

ques

tionn

aire

attit

udes

inan

tid

entif

iers

, str

ong

and

in th

eir

ethn

icT

Xto

war

dA

naly

sis

His

pani

c id

entif

iers

are

Don

thu

iden

tific

atio

n?vo

ters

,in

stitu

tions

mor

e lik

ely

(1)

to u

se(1

986)

278

and

ad;

Spa

nish

lang

uage

med

ia,

Whi

tes,

bran

d(2

) to

hav

e m

ore

147

loya

lty;

posi

tive

attit

udes

tow

ard

His

pani

cspu

rcha

sead

vert

isin

g, a

nd (

3) to

Influ

ence

spu

rcha

se p

rodu

cts

adve

rtise

d to

His

pani

cs

Pitt

s,H

ow d

o B

lack

s an

dN

o fr

ame-

Bla

cks

271;

Fou

r 60

-P

erso

nal

No

No

MA

N-

•Rel

ativ

e to

Whi

tes,

Wha

len,

Whi

tes

resp

ond

tow

ork

used

unde

r-se

cond

valu

es; a

dO

VA

Bla

cks

resp

ond

mor

eO

’Kee

fe,

TV

ads

with

grad

s fro

mT

V a

dsra

ting;

pro

duct

favo

rabl

y to

TV

ads

and

Mur

ray

cultu

ral/e

thni

cW

. and

S.

targ

eted

usag

e ra

tew

ith B

lack

act

ors

(198

9)cu

es?

US

; 82

at B

lack

s•C

ompa

red

to B

lack

s,B

lack

s, 1

89W

hite

s fa

il to

not

ice

Whi

tes

som

e cu

ltura

l val

ues

(e.g

., be

long

ing,

sel

f-fu

lfillm

ent,

acco

mpl

ish-

men

t) d

epic

ted

inco

mm

erci

als (c

ontin

ued)


Whi

ttler

Do

race

of a

d ac

tor

No

fram

e-B

lack

s,34

0;F

ull-c

olor

Mes

sage

No

Yes

AN

OV

A,

•Whi

tes

(Bla

cks)

iden

tify

(198

9)(i.

e., W

hite

ver

sus

wor

k us

edW

hite

sun

der-

stor

yboa

rdco

mpr

ehen

sion

;M

ultip

lem

ore

with

Whi

te (

Bla

ck)

Bla

ck)

and

view

ers’

grad

s fro

mad

s w

ith a

raci

al a

ttitu

des

Reg

res-

acto

rs th

an B

lack

raci

al a

ttitu

des

US

Bla

ck a

ctor

sion

Whi

te) a

ctor

saf

fect

ad

Mid

wes

t;an

d a

Whi

te•B

lack

s (W

hite

s) r

epor

tev

alua

tions

?18

0 W

hite

,ac

tor

mor

e lik

ely

to b

uy th

e16

0 B

lack

adve

rtis

ed b

rand

and

eval

uate

ads

with

aB

lack

(W

hite

) ac

tor

mor

e fa

vora

bly

•Whi

tes

do n

ot r

eact

nega

tivel

y to

Bla

ckac

tors

in a

ds

Web

ster

Do

His

pani

c gr

oups

No

fram

e-n/

a18

0;N

one;

Ext

ent o

fN

oN

oA

NO

VA

,•H

ispa

nics

mor

e w

illin

g(1

992)

cate

goriz

ed b

yw

ork

used

wom

enad

min

iste

red

info

rmat

ion

MA

N-

to s

earc

h m

edia

with

sub-

cultu

ral e

thni

cfr

om S

anqu

estio

nnai

rese

arch

inO

VA

whi

ch th

ey e

thni

cally

iden

tific

atio

n se

arch

Ant

onio

,va

rious

med

iaid

entif

ydi

ffere

ntly

for

adT

X(e

.g.,

ads,

•Rel

ativ

e to

non

-Spa

nish

-in

form

atio

n?ye

llow

pag

es)

spea

king

His

pani

cs,

Spa

nish

-spe

akin

gH

ispa

nics

are

influ

ence

d le

ss b

ybr

ochu

re a

ds a

ndm

agaz

ine

ads

Ros

low

Are

His

pani

csN

o fr

ame-

His

-64

8; a

dult

Spa

nish

-A

d pe

rsua

sive

-N

oN

oT

-tes

tH

ispa

nics

mor

ean

dm

ore

pers

uade

d by

wor

k us

edpa

nics

His

pani

csla

ngua

gene

sspe

rsua

ded

by S

pani

sh-

Nic

holls

Spa

nish

-lang

uage

from

and

lang

uage

TV

ads

(199

6)or

Eng

lish-

Mia

mi,

Eng

lish

embe

dded

in S

pani

sh-

lang

uage

TV

ads

?H

oust

on,

lang

uage

lang

uage

TV

pro

gram

sLA

, and

TV

ads

than

by

anal

ogou

sN

YC

Eng

lish-

lang

uage

TV

ads

embe

dded

inE

nglis

h-la

ngua

ge T

Vpr

ogra

ms

Tab

le 2

(co

nti

nu

ed)

Su

mm

ary

of

Eth

nic

ity

and

Ad

vert

isin

g L

iter

atu

re

Ove

rall

Eth

-n;

Dat

aId

en.

Sta

tistic

alA

rtic

leR

esea

rch

Que

stio

nT

heor

yni

city

Sub

ject

sS

timul

iC

olle

cted

w/ A

dP

IM

etho

dsF

indi

ngs

(con

tinue

d)

Fall 2009 47T

able

2 (

con

tin

ued

)S

um

mar

y o

f E

thn

icit

y an

d A

dve

rtis

ing

Lit

erat

ure

Ove

rall

Eth

-n;

Dat

aId

en.

Sta

tistic

alA

rtic

leR

esea

rch

Que

stio

nT

heor

yni

city

Sub

ject

sS

timul

iC

olle

cted

w/ A

dP

IM

etho

dsF

indi

ngs

Dim

ofte

,C

an a

d ta

rget

ing

No

fram

e-H

is-

82;

Tw

o T

V a

dsE

thni

city

; ad

No

No

Chi

-•U

nusu

al v

oice

-ove

r/F

oreh

and,

inco

ngru

ent w

ithw

ork

used

pani

cs,

empl

oyee

sw

ith a

His

-sc

hem

asq

uare

,su

btitl

ing

augm

ents

and

exis

ting

adW

hite

sof

maj

orpa

nic

acto

r;co

ngru

ity;

AN

OV

Aet

hnic

sel

f-aw

aren

ess

Des

hpan

désc

hem

ata—

base

dC

alifo

rnia

one

ad w

ithat

titud

esan

d in

crea

ses

ad re

call

(200

4)on

unu

sual

use

of

univ

ersi

tyE

nglis

h vo

ice-

tow

ard

the

•Ad

sche

ma

cong

ruity

iden

tity

cues

—ov

er, a

nd o

nesp

okes

pers

onm

oder

ates

the

effe

ctaf

fect

the

salie

nce

ad w

ithan

d ad

;of

targ

et m

arke

tof

sel

f-id

entif

icat

ion

Spa

nish

voi

ce-

cogn

itive

affil

iatio

n on

atti

tude

san

d re

spon

ses

ofov

er a

ndre

spon

ses

tow

ard

the

ad a

ndta

rget

con

sum

ers?

Eng

lish

spok

espe

rson

subt

itles

Will

iam

sD

o m

iddl

e cl

ass

Cul

tura

lB

lack

s16

0; a

dults

TV

ads

Atti

tude

sN

oN

oD

iscr

im-

•Stro

ng a

nd w

eak

Bla

ckan

d Q

ualls

Bla

cks

and

Whi

tes

Scr

ipt

from

feat

urin

gto

war

d th

ein

ant

iden

tifie

rs r

espo

nd(1

989)

resp

ond

diffe

rent

lyD

enve

,rce

lebr

itysp

okes

pers

on,

Ana

lysi

s,si

mila

rly to

ads

with

to a

ds fe

atur

ing

CO

, 80

endo

rser

sad

, and

AN

OV

Ace

lebr

ity e

ndor

sers

cele

brity

end

orse

rs?

Bla

cks,

80

prod

uct

•Stro

ng B

lack

iden

tifie

rsW

hite

san

d W

hite

s ho

ldsi

mila

r fa

vora

ble

attit

udes

tow

ard

ads

with

cel

ebrit

yen

dors

ers

Des

hpan

d éD

o m

ajor

ity a

ndD

istin

c-H

is-

205;

adu

ltsR

adio

-scr

ipt

Eth

nic

No

No

AN

OV

A,

Rel

ativ

e to

maj

ority

gro

upan

dm

inor

ity g

roup

tiven

ess

pani

cs;

from

San

ads

with

eith

ersa

lienc

e;Lo

gist

icm

embe

rs, m

inor

ity g

roup

Sta

yman

mem

bers

res

pond

Whi

tes

Ant

onio

a W

hite

or

spok

espe

rson

Reg

res-

mem

bers

are

mor

e(1

994)

diffe

rent

ly to

rad

ioan

d A

ustin

,H

ispa

nic-

trus

twor

thy-

sion

,Pat

hlik

ely

(1)

to d

eem

thei

rad

s w

ith W

hite

or

TX

nam

edne

ss; b

rand

Ana

lysi

set

hnic

ity im

port

ant,

and

His

pani

c-na

med

spok

espe

rson

attit

udes

(2) t

o tru

st s

poke

sper

sons

spok

espe

rson

s?of

sim

ilar e

thni

city

, whi

chin

duce

s m

ore

posi

tive

attit

udes

tow

ard

the

adve

rtis

ed b

rand

(con

tinue

d)

48 Journal of Current Issues and Research in AdvertisingT

able

2 (

con

tin

ued

)S

um

mar

y o

f E

thn

icit

y an

d A

dve

rtis

ing

Lit

erat

ure

Ove

rall

Eth

-n;

Dat

aId

en.

Sta

tistic

alA

rtic

leR

esea

rch

Que

stio

nT

heor

yni

city

Sub

ject

sS

timul

iC

olle

cted

w/ A

dP

IM

etho

dsF

indi

ngs

Grie

r an

dW

hat m

eani

ngs

doD

istin

c-B

lack

s,62

; MB

As;

Ad

broc

hure

Tho

ught

s an

dN

oN

oC

hi-

Rel

ativ

e to

non

-targ

eted

Bru

mba

ugh

targ

eted

and

non

-tiv

enes

sW

hite

s,35

Whi

te,

that

mim

ics

feel

ings

abo

utsq

uare

grou

ps, t

arge

ted

grou

ps(1

999)

targ

eted

gro

ups

gay/

20 B

lack

,m

agaz

ine

ads

the

ad(1

) vi

ew a

ds m

ore

deriv

e fr

om a

ds?

lesb

ian

7 un

know

n;po

sitiv

ely,

and

(2) b

ette

r40

stra

ight

,un

ders

tand

cul

tura

l cue

s22

oth

erin

ads

Aak

er,

Wha

t effe

ct d

o ad

sD

istin

c-B

lack

s,63

; MB

As

Pam

phle

t of

Atti

tude

tow

ard

No

No

AN

OV

A•R

elat

ive

to th

e ta

rget

edB

rum

baug

h,ta

rget

ed a

t one

tiven

ess

Whi

tes,

from

US

colo

r prin

t ads

the

ad, f

elt

grou

p, n

on-t

arge

ted

and

Grie

rau

dien

ce h

ave

onga

y/M

idw

est;

targ

etin

gta

rget

edne

ssgr

oups

vie

w a

ds le

ss(2

000)

non-

targ

eted

lesb

ian

123;

Bla

cks,

posi

tivel

yau

dien

ces?

stud

ents

,W

hite

s, o

r•W

hen

targ

etin

g cu

es39

Bla

ck,

gay/

lesb

ians

are

abse

nt, v

iew

er84

Whi

tedi

stin

ctiv

enes

s al

one

isun

rela

ted

to A

AD

•Fel

t sim

ilarit

y an

d fe

ltta

rget

edne

ss m

edia

teA

AD fo

r int

ende

d ta

rget

s

For

ehan

dD

oes

ethn

ic s

elf-

Dis

tinc-

Asi

ans,

109

and

Ser

ies

of T

VA

ttitu

de to

war

dN

oN

oLo

gist

icW

hen

ethn

ic c

ues

are

and

awar

enes

s af

fect

tiven

ess

Whi

tes

175;

Asi

anan

d pr

int a

dsth

e ad

and

Reg

res-

anal

ogou

s to

one

’s s

elf-

Des

hpan

d éco

nsum

ers’

and

Whi

tesp

okes

pers

on;

sion

,co

ncep

t, th

ey in

crea

se(2

001)

resp

onse

s to

unde

rgra

dsst

reng

th o

fA

NO

VA

the

rate

at w

hich

peo

ple

targ

eted

ads

?fro

m W

. US

ethn

ic id

entit

ym

entio

n th

eir

ethn

icity

in s

elf-

desc

riptio

ns,

lead

ing

to m

ore

favo

rabl

ere

spon

ses

to a

sam

e-et

hnic

ity s

poke

sper

son

and

ads

targ

etin

g th

eir

ethn

icity

Grie

r an

dD

oes

soci

al a

ndD

istin

c-B

lack

s,17

6; S

.O

ne p

rint a

dR

acia

l/eth

nic

No

No

AN

OV

A,

•Usi

ng s

ocia

l dim

ensi

ons

Des

hpan

d énu

mer

ic s

tatu

s of

tiven

ess

Whi

tes

Afr

ican

with

eith

er a

salie

nce

and

Uni

varia

te/t

o ta

rget

ads

may

be

(200

1)S

outh

Afr

ican

wom

enB

lack

or W

hite

bran

d at

titud

eM

ultip

leef

fect

ive

even

whe

nco

nsum

ers

affe

ct a

dsp

okes

pers

onR

egre

s-ta

rget

ed g

roup

is a

effe

ctiv

enes

s?si

on,

num

eric

maj

ority

(con

tinue

d)

Fall 2009 49T

able

2 (

con

tin

ued

)S

um

mar

y o

f E

thn

icit

y an

d A

dve

rtis

ing

Lit

erat

ure

Ove

rall

Eth

-n;

Dat

aId

en.

Sta

tistic

alA

rtic

leR

esea

rch

Que

stio

nT

heor

yni

city

Sub

ject

sS

timul

iC

olle

cted

w/ A

dP

IM

etho

dsF

indi

ngs

(con

tinue

d)

Pat

h•S

ocia

l sta

tus

pred

icts

Ana

lysi

set

hnic

sal

ienc

e•G

reat

er (

less

er)

ethn

icsa

lienc

e du

e to

min

ority

(maj

ority

) st

atus

lead

sto

mor

e (le

ss)

posi

tive

attit

ude

tow

ard

the

bran

d

For

ehan

d,F

or W

hite

s an

dD

istin

c-W

hite

s,28

4 an

dE

xper

imen

t 1:

Spo

kesp

erso

nN

oN

oLo

gist

ical

Acr

oss

both

exp

erim

ents

,D

eshp

andé

,A

sian

s, h

ow d

otiv

enes

sA

sian

s38

4; A

sian

Fou

r T

V a

ds,

likin

g; a

ttitu

deR

egre

s-A

sian

s (W

hite

s)an

d R

eed

ethn

ic p

rimes

and

and

Whi

te1

for

Nok

iato

war

d th

e ad

;si

on,

resp

onde

d m

ore

(200

2)so

cial

dis

tinct

ive-

unde

rgra

dsan

d 3

fille

r ads

cogn

itive

MA

N-

favo

rabl

y (n

egat

ivel

y) to

ness

affe

ct id

entit

yfr

om W

.E

xper

imen

t 2:

resp

onse

sO

VA

an A

sian

spo

kesp

erso

nsa

lienc

e an

dU

S2

prin

t ads

for

and

Asi

an-r

elat

ed a

dsre

spon

ses

toN

orth

wes

tw

hen

they

wer

e bo

thta

rget

ed a

ds?

Airl

ines

prim

ed a

nd s

ocia

llydi

stin

ctiv

e (i.

e., a

num

eric

min

ority

gro

up, i

n a

cert

ain

soci

alen

viro

nmen

t)

Lee,

How

do

ethn

icD

istin

c-A

sian

s,17

8;F

ull-c

olor

Sel

f-N

oY

esA

NO

VA

,C

onsu

mer

s ex

pose

d to

Fern

ande

z,m

inor

ity m

odel

s in

tiven

ess

Whi

tes

colle

gepr

int a

ds fo

rre

fere

ncin

g;M

AN

-ad

s co

nsis

tent

with

thei

ran

d M

artin

ads

affe

ct th

est

uden

tsw

atch

es a

ndet

hnic

iden

tity;

OV

Aet

hnic

ity s

pont

aneo

usly

(200

2)ev

alua

tions

of

faci

al ti

ssue

attit

udes

self-

refe

renc

e th

e ad

,et

hnic

min

ority

and

with

eith

erto

war

d th

e ad

,w

hich

lea

ds to

(a) m

ore

ethn

ic m

ajor

ityA

sian

or

bran

d, a

ndpo

sitiv

e at

titud

es to

war

dsco

nsum

ers?

Whi

te m

odel

sm

odel

;th

e ad

and

the

mod

elco

gniti

vede

pict

ed in

the

ad, (

b)re

spon

ses

stro

nger

inte

ntio

ns to

buy

the

adve

rtis

edbr

and,

and

(c)

mor

efa

vora

ble

impr

essi

ons

of th

e ad

vert

ised

bra

nd


Mar

tin,

Doe

s co

nsum

erD

istin

c-A

sian

s,12

2T

ea a

ndS

elf-

No

Yes

AN

OV

A•S

elf-

refe

renc

ing

Lee,

and

self-

refe

renc

ing

tiven

ess

Whi

tes

colle

gesh

ampo

o pr

int

refe

renc

ing;

med

iate

s et

hnic

ity e

ffect

sY

ang

med

iate

the

effe

ctst

uden

ts;

ads

with

Whi

teat

titud

eson

(a)

ad-

rela

ted

(200

4)of

eth

nici

ty o

n66

Whi

tes,

and

Asi

anto

war

d th

eat

titud

es,

and

(b)

ad/b

rand

-rel

ated

56 A

sian

sfe

mal

e m

odel

sad

,in

tent

ions

to b

uy th

eat

titud

es a

ndm

odel

, and

adve

rtis

ed b

rand

inte

ntio

ns to

buy

bran

d•F

or A

sian

s, s

elf-

the

adve

rtis

edre

fere

ncin

g re

late

sbr

and?

posi

tivel

y to

atti

tude

sto

war

d th

e ad

and

inte

ntio

ns to

buy

the

adve

rtis

ed b

rand

•For

Asi

ans

and

Whi

tes,

Asi

an m

odel

s in

ads

for

atyp

ical

pro

duct

sbo

oste

d (a

) se

lf-re

fere

ncin

g, (

b) a

d-re

late

d at

titud

es, a

nd(c

) in

tent

ions

to b

uyth

e ad

vert

ised

bra

nd•R

elat

ive

to W

hite

s,A

sian

s ex

hibi

t gre

ater

self

refe

renc

ing

ofA

sian

mod

els

in a

ds

App

iah

Doe

s st

reng

th o

fD

istin

c-B

lack

s,17

3; H

ST

hree

full-

Per

ceiv

edY

esN

oM

ean

•Rel

ativ

e to

wea

k B

lack

(200

1a)

ethn

ic id

entit

ytiv

enes

s;W

hite

sst

uden

tsco

lor

prin

tsi

mila

rity

with

Com

-id

entif

iers

, stro

ng B

lack

affe

ct B

lack

and

Iden

tifi-

from

CA

,ad

s, w

ithac

tors

; bel

ieve

paris

onid

entif

iers

see

them

-W

hite

ado

lesc

ents

’ca

tion

81 B

lack

s,ra

ce o

fad

inte

nded

for

selv

es a

s m

ore

sim

ilar

resp

onse

s to

prin

t92

Whi

tes

char

acte

rssu

bjec

t; id

enti-

to a

nd id

entit

y m

ore

ads

with

mod

els

ofan

d cu

ltura

lfic

atio

n w

ithst

rong

ly w

ith B

lack

diffe

rent

rac

es?

cues

acto

rs; a

ttitu

de to

war

dch

arac

ters

in a

dsdi

gita

lly•W

hite

s’ id

entif

icat

ion

with

man

ipul

ated

char

acte

rs o

r at

titud

e

Tab

le 2

(co

nti

nu

ed)

Su

mm

ary

of

Eth

nic

ity

and

Ad

vert

isin

g L

iter

atu

re

Ove

rall

Eth

-n;

Dat

aId

en.

Sta

tistic

alA

rtic

leR

esea

rch

Que

stio

nT

heor

yni

city

Sub

ject

sS

timul

iC

olle

cted

w/ A

dP

IM

etho

dsF

indi

ngs

(con

tinue

d)

Fall 2009 51

tow

ard

the

ad is

un-

rela

ted

to th

eir

stre

ngth

of e

thni

c id

entit

y

App

iah

How

do

Whi

te,

Dis

tinc-

Bla

cks,

349;

HS

Fou

r fu

ll-P

erce

ived

Yes

No

AN

OV

A,

•Bla

cks

are

mor

e(2

001b

)B

lack

, His

pani

c,tiv

enes

s;W

hite

sst

uden

tsco

lor

prin

tsi

mila

rity

and

Mea

nin

fluen

ced

by ra

ce o

f the

and

Asi

an-

Iden

tifi-

from

CA

,ad

s, w

ithid

entif

icat

ion

Com

-ad

act

or th

an e

thni

cA

mer

ican

catio

n81

Bla

cks,

Bla

ck o

rw

ithpa

rison

cues

in a

dad

oles

cent

s84

Asi

an,

Whi

tech

arac

ters

;•I

dent

ifica

tion

with

an

resp

ond

to a

ds92

His

-ch

arac

ters

,be

lieve

ad

ad b

y H

ispa

nics

,w

ith B

lack

or

pani

cs, 9

2fo

rin

tend

ed fo

rB

lack

s, a

nd A

sian

s is

Whi

te a

ctor

s?W

hite

sco

nsum

ersu

bjec

t; ad

influ

ence

d by

rac

eno

n-ra

ting

(i.e.

, Bla

ck)

of th

e ad

dura

bles

acto

r•B

lack

s, W

hite

s,H

ispa

nics

, and

Asi

ans

rate

Bla

ck-a

ctor

ads

mor

e fa

vora

bly

than

Whi

te-a

ctor

ads

Kos

low

,H

ow d

o H

ispa

nics

Acc

om-

His

-36

7;Fo

ur p

rint a

ds:

Atti

tide

tow

ard

No

No

OLS

Spa

nish

-lang

uage

ads

Sha

m-

resp

ond

to e

thni

cm

odat

ion

pani

cs,

Spa

nish

1 E

nglis

hth

e ad

; adv

er-

Reg

res-

enha

nce

His

pani

cs’

dasa

ni,

lang

uage

in a

ds?

Whi

tes

spea

kers

only

, 1tis

er’s

sen

sitiv

itysi

on,

belie

fs a

bout

the

and

Spa

nish

onl

y,to

His

pani

cA

NC

OV

Aad

vert

iser

’s s

ensi

tivity

Touc

hsto

nean

d 2

mix

edcu

lture

;to

His

pani

c cu

lture

, whi

ch(1

994)

Eng

lish

and

ethn

icbo

osts

His

pani

cs’

Spa

nish

iden

tity

attit

ude

tow

ard

the

ad

Gre

enA

re e

thni

c id

entit

y,A

ccom

-B

lack

s,31

3; B

lack

Ful

l-col

orE

thni

cN

oY

esA

NO

VA

Rel

ativ

e to

wea

k et

hnic

(199

9)m

edia

pla

cem

ent,

mod

atio

n;W

hite

sA

mer

ican

prin

t ads

,id

entit

y;id

entif

iers

, str

ong

ethn

ican

d ra

ce o

f ad

In-G

roup

fem

ales

with

eith

erat

titud

eid

entif

iers

(1)

eva

luat

eac

tor(

s) r

elat

ed to

Bia

sW

hite

or

tow

ard

ads

that

feat

ure

Bla

cks

Bla

cks ’

AA

D a

ndB

lack

mod

els,

the

adm

ore

posi

tivel

y, a

nd (

2)bu

ying

inte

ntio

ns?

for

a fic

titio

usha

ve s

tron

ger

purc

hase

bran

din

tent

ions

for b

rand

s w

ithad

s th

at fe

atur

e B

lack

s

(con

tinue

d)

Tab

le 2

(co

nti

nu

ed)

Su

mm

ary

of

Eth

nic

ity

and

Ad

vert

isin

g L

iter

atu

re

Ove

rall

Eth

-n;

Dat

aId

en.

Sta

tistic

alA

rtic

leR

esea

rch

Que

stio

nT

heor

yni

city

Sub

ject

sS

timul

iC

olle

cted

w/ A

dP

IM

etho

dsF

indi

ngs


(con

tinue

d)

Tab

le 2

(co

nti

nu

ed)

Su

mm

ary

of

Eth

nic

ity

and

Ad

vert

isin

g L

iter

atu

re

Ove

rall

Eth

-n;

Dat

aId

en.

Sta

tistic

alA

rtic

leR

esea

rch

Que

stio

nT

heor

yni

city

Sub

ject

sS

timul

iC

olle

cted

w/ A

dP

IM

etho

dsF

indi

ngs

Qua

lls a

ndD

oes

race

affe

ctIn

-Gro

upB

lack

s,21

1; b

eer

Exp

erim

enta

lA

ttitu

des

No

No

Mea

n•R

elat

ive

to P

olar

ized

Moo

reB

lack

s’ a

ndB

ias;

Whi

tes

drin

kers

TV

ad

for

ato

war

d th

eC

ompa

r-A

ppra

isal

The

ory,

In-

(199

0)W

hite

s ’ a

dP

olar

ized

from

Mid

-ne

w b

eer

acto

r, a

d, a

ndis

onG

roup

Bia

s T

heor

yev

alua

tions

?A

ppra

isal

wes

t tow

npr

oduc

tbe

tter

expl

ains

the

in U

S; 1

03ef

fect

of r

ace

onB

lack

s, 1

07pe

ople

’s a

d ev

alua

tions

Whi

tes

•Pro

duct

eva

luat

ions

by B

lack

s (W

hite

s)im

prov

e (w

orse

n) fo

rad

s w

ith B

lack

act

ors

and

wor

sen

(impr

ove)

for

ads

with

Whi

teac

tors

Whi

ttler

Wha

t effe

ct d

oes

Heu

ristic

-B

lack

s,16

0; P

aid

Ful

l-col

orT

houg

ht-

No

Yes

AN

OV

A,

•Low

-pre

judi

ce W

hite

san

d D

iMeo

view

er’s

pre

judi

ceS

yste

mat

icW

hite

svo

lunt

eers

stor

y-bo

ard

listin

g, b

rand

Logi

stic

al(a

) be

lieve

equ

ally

(199

1)ha

ve o

n si

mpl

eP

ersu

asio

nad

s fo

r a

fur

awar

enes

s,R

egre

s-si

mila

r to

Whi

te a

ndde

cisi

on r

ules

whe

nM

odel

coat

and

ad r

ecal

l,si

on,

Bla

ck a

ctor

s, a

nd (

b)ex

amin

ing

adliq

uid

attit

udes

tow

ard

AN

CO

VA

iden

tify

equa

lly w

ithef

fect

iven

ess?

laun

dry

Bla

cks,

act

or’s

Whi

te a

nd B

lack

act

ors

dete

rgen

tlik

eabi

lity,

•Hig

h-pr

ejud

ice

Whi

tes

perc

eive

d(a

) be

lieve

less

sim

ilar

sim

ilarit

y an

dto

Bla

ck th

an W

hite

iden

tific

atio

nac

tors

, and

(b)

iden

tify

with

act

or,

mor

e st

rong

ly w

ith W

hite

purc

hase

than

Bla

ck a

ctor

sin

tent

ions

, ad

•Reg

ardl

ess

of th

eir

impr

essi

on,

attit

udes

tow

ard

Bla

cks,

raci

al a

ttitu

des,

Whi

tes

had

low

erva

lidity

of

purc

hase

inte

ntio

nsm

essa

ge c

laim

san

d le

ss fa

vora

ble

attit

udes

tow

ard

ads

and

bran

ds fo

r ad

sfe

atur

ing

a B

lack

act

or

Fall 2009 53T

able

2 (

con

tin

ued

)S

um

mar

y o

f E

thn

icit

y an

d A

dve

rtis

ing

Lit

erat

ure

Ove

rall

Eth

-n;

Dat

aId

en.

Sta

tistic

alA

rtic

leR

esea

rch

Que

stio

nT

heor

yni

city

Sub

ject

sS

timul

iC

olle

cted

w/ A

dP

IM

etho

dsF

indi

ngs

(con

tinue

d)

Whi

ttler

Wha

t are

vie

wer

s’H

euris

tic-

Whi

tes,

160

Stu

dy 1

: Fu

ll-S

ame

asN

oY

esM

ean

•Sou

rce

char

acte

ristic

s(1

991)

psyc

holo

gica

lS

yste

mat

icB

lack

sW

hite

s,co

lor

stor

y-W

hittl

er a

ndC

ompa

r-(e

.g.,

acto

r ’s e

thni

city

)pr

oces

ses

Per

suas

ion

140

boar

d ad

sD

iMeo

(19

91)

ison

,w

ere

mor

e lik

ely

tosu

bseq

uent

of

Mod

el;

Bla

cks;

for

liqui

dpl

usIn

tera

ctio

naf

fect

vie

wer

s w

ithex

posu

re to

ads

Iden

tifi-

unde

rgra

dsla

undr

yad

ditio

nal

Effe

cts

mor

e in

tens

e ra

cial

with

diff

erin

gca

tion

and

160

dete

rgen

tin

form

atio

nat

titud

eset

hnic

act

ors?

Whi

tean

d a

wor

dab

out t

he•H

igh-

prej

udic

e W

hite

sad

ults

proc

esso

rpr

oduc

tid

entif

ied

mor

e w

ithfr

om th

eS

tudy

2:

Ful

l-W

hite

than

Bla

ck a

ctor

sso

uthe

ast

colo

r st

ory-

in a

dsbo

ard

ads

•Bla

cks

wer

e m

ore

likel

yfo

r a

fur

coat

to b

uy th

e ad

vert

ised

and

liqui

dpr

oduc

t whe

n B

lack

laun

dry

acto

rs w

ere

used

dete

rgen

t•H

igh-

prej

udic

e W

hite

sw

ere

(a)

less

inte

rest

edin

lear

ning

mor

e ab

out

the

prod

uct w

hen

the

ad c

onta

ined

a B

lack

acto

r, a

nd (

b) le

sslik

ely

to r

ecog

nize

the

bran

d w

hen

the

adco

ntai

ned

a B

lack

acto

r

Whi

ttler

Wha

t effe

ct d

oes

Ela

bora

-B

lack

s,16

0 B

lack

Ful

l-col

orS

ame

asN

oY

esM

ean

•Ide

ntifi

catio

n w

ith B

lack

and

Spi

rara

ce in

adv

ertis

ing

tion

Whi

tes

adul

ts fr

omst

ory-

boar

dW

hittl

er a

ndC

ompa

r-cu

lture

mod

erat

es(2

002)

have

on

Bla

ckLi

kelih

ood

the

ads

for

aD

iMeo

(19

91)

ison

,B

lack

s ’ re

spon

ses

to a

dsvi

ewer

mes

sage

Mod

elS

E U

Sco

rdle

sspl

us m

essa

geIn

tera

ctio

n•H

igh-

iden

tific

atio

npr

oces

sing

and

doe

sph

one,

com

preh

ensi

on,

Effe

cts

Bla

cks

resp

ond

mor

eit

func

tion

as a

garm

ent b

ag,

iden

tific

atio

nfa

vora

bly

to a

ds,

perip

hera

l cue

?an

d liq

uid

with

Bla

ckpr

oduc

ts, a

nd m

odel

sla

undr

ycu

lture

whe

n th

e ad

mod

elde

terg

ent

is B

lack

54 Journal of Current Issues and Research in AdvertisingT

able

2 (

con

tin

ued

)S

um

mar

y o

f E

thn

icit

y an

d A

dve

rtis

ing

Lit

erat

ure

Ove

rall

Eth

-n;

Dat

aId

en.

Sta

tistic

alA

rtic

leR

esea

rch

Que

stio

nT

heor

yni

city

Sub

ject

sS

timul

iC

olle

cted

w/ A

dP

IM

etho

dsF

indi

ngs

•Ad

mod

el’s

rac

ein

fluen

ces

Bla

cks’

perip

hera

l mes

sage

proc

essi

ng•B

lack

s th

ink

mor

eab

out t

he p

rodu

ct in

ads

with

Whi

te r

athe

rth

an B

lack

mod

els

•Bla

cks’

bet

ter

reca

llad

s w

ith B

lack

act

ors

Pre

sent

Wha

t effe

ct d

oes

Soc

ial

Whi

tes,

207;

Ful

l-col

orE

thni

cY

esY

esS

EM

Eth

nica

lly id

entif

ying

with

Stu

dyet

hnic

iden

tific

atio

nId

entit

yB

lack

s,un

der-

prin

t ads

, with

iden

tific

atio

na

prin

t ad,

bas

ed o

n th

e(2

009)

with

prin

t ads

,H

is-

grad

s fr

omei

ther

Whi

te,

with

the

ad;

mod

el’s

app

aren

tcr

eate

d by

the

pani

csS

W to

wn

Bla

ck, o

rat

titud

es to

war

det

hnic

ity, r

esul

ts in

am

odel

’s a

ppar

ent

in U

S;

93H

ispa

nic

mal

eth

e ad

and

posi

tive

dire

ct a

ndet

hnic

ity, h

ave

onW

hite

s, 6

2m

odel

s, fo

r a

bran

d;in

dire

ct (t

hrou

gh a

ttitu

deco

nsum

er re

spon

ses

Bla

cks,

52

fictit

ious

purc

hase

tow

ard

the

ad a

ndto

the

ad a

nd b

rand

?H

ispa

nics

athl

etic

sho

ein

tent

ions

attit

ude

tow

ard

the

bran

d)br

and

effe

ct o

n pu

rcha

sein

tent

ions

of t

head

vert

ised

bra

nd

Key

:O

vera

ll T

heo

ry=

gen

eral

th

eore

tica

l fra

mew

ork

th

at g

rou

nd

ed s

tud

yE

thn

icit

y=

eth

nic

ity

dep

icte

d in

th

e te

st a

ds

n=

sam

ple

siz

e fo

r ea

ch s

tud

y; n

ote

th

at s

om

e ar

ticl

es d

escr

ibed

mo

re t

han

on

e st

ud

yS

ub

ject

s=

stu

dy

par

tici

pan

tsS

tim

uli

=ad

s o

r co

mm

erci

als

that

par

tici

pan

ts r

ead

or

view

edId

en. w

/ Ad

=d

ata

colle

cted

incl

ud

ed m

easu

re o

f id

enti

fica

tio

n w

ith

th

e ad

(E

thId

enA

D)

PI

=d

ata

colle

cted

incl

ud

ed m

easu

re o

f p

urc

has

e in

ten

tio

ns

(PI B

)

Fall 2009 55

identity is the characteristics or traits that each personbelieves he or she possesses (Reed 2002; Tajfel andTurner 1985), such as knowledge of one’s culturalheritage. Ads with ethnically resonant cues allow view-ers to differentiate themselves from others, thereby so-lidifying their self identity and uniqueness. Becausesocial identity is context-dependent, ads that triggeridentity importance must fit the expectations of in-groupmembers (Maldonado and Muehling 2006).

SIT suggests that members of a social group iden-tify with that group, view themselves as representa-tive of that group, and model their attitudes, emotions,and behaviors accordingly (Maldonado, Tansuhaj, andMuehling 2003; Reed 2002; Tajfel and Turner 1985).Identifying with a social group creates an intransi-gent social identity comprised of three facets: cogni-tive, evaluative, and emotive (Tajfel 1978b, 1981).Knowledge of belonging to a group is the cognitivefacet; the positive or negative connotation of being agroup member is the evaluative facet; and conjectureabout others’ feelings regarding one’s group mem-bership is the emotive facet.

Hypotheses

Ethnic identity—an enduring, underlying sense ofconnection to a social group (Tajfel 1978b)—entailsself-identification as a group member, a sense of be-longing, and favorable attitudes toward one’s group(Phinney 1990). Because strength of ethnic identity(i.e., strength of identity with one’s ethnic origin[Deshpandé, Hoyer, and Donthu 1986], or an endur-ing association between one’s ethnicity and sense ofself [Forehand and Deshpandé 2001]) can affect con-sumers’ responses to marketing activities, shoppingorientations, and purchase decisions (Davis andGandy 1999; Donthu and Cherian 1994; Green 1999;Webster 1992), it is worthy of study (Lee, Fairhurst,and Dillard 2002).

Ethnic consumers are more likely to believe andidentify with ads targeted at them (Appiah 2001a).Relative to Blacks with weaker ethnic identities, Blackswith stronger ethnic identities see themselves as moresimilar to, and identify more strongly with, Black char-acters in ads (Appiah 2001a; Whittler 1989). More-over, Blacks, Hispanics, and Asians with strongerethnic identities tend to identify more strongly withads containing actors with similar ethnicity (Appiah2001b). For example, Blacks who strongly self-iden-tify ethnically tend to identify more strongly with adscontaining Black actors than ads containing Hispanicactors. Also, strong Hispanic identifiers are keen onmedia that use Spanish verbiage (Deshpandé, Hoyer,

and Donthu 1986), which implies that as peoples’strength of ethnic identity increases, their ability tonotice and identify with ads that use similar ethniccues increases. Thus, H1 is tested.

H1: The stronger (weaker) a person’s ethnicidentity, the more (less) that person iden-tifies ethnically with ads containing amodel of the apparently same ethnicity.

To effectively target an ethnic group, advertisersmay benefit from using ethnically resonant themes orcues; specifically, by linking an ad to each groupmember’s self identity, advertisers may favorably af-fect group members’ AAD (Aaker, Brumbaugh, andGrier 2000; Brumbaugh 2002). Published studies onethnicity and advertising offer some insight into thisphenomenon. Product evaluations by Blacks (Whites)improve (worsen) for television ads with Black actorsand worsen (improve) for television ads with Whiteactors (Qualls and Moore 1990). Blacks (Whites) respondmore favorably to ads with Black (White) actors thanads with White (Black) actors (Whittler 1989, 1991).Asians respond favorably to television ads with Asianspokespersons (Forehand, Deshpandé, and Reed 2002).Hispanics who view predominantly English-languageads that include a few Spanish words tend to believethat the advertiser is sensitive to their culture, whichencourages more favorable AAD (Koslow, Shamdasani,and Touchstone 1994). To explore further the effect ofethnicity on AAD, we examine the direct effect of ethnicidentification with a print ad, created by a model’s ap-parent ethnicity, on AAD. Thus, we test H2.

H2: The more (less) a person identifies ethnicallywith ads containing a model of the appar-ently same ethnicity, the more (less) positiveare that person’s attitudes toward the ad.

Generally, consumers’ attitudes strongly influencetheir purchase behaviors (e.g., Holmes and Crocker1987; Priester et al. 2004; Whittler 1989). More specifi-cally to advertising, the preponderance of scholarlyliterature suggests that AAD has a positive effect onAB, which in turn, has a positive effect on PIB, i.e., AAD→ AB → PIB (Brown and Stayman 1992; MacKenzieand Lutz 1989). For example, corporate and endorsercredibility positively affect consumers’ AAD, AB, andPIB (Goldsmith, Lafferty, and Newell 2000; Lafferty,Goldsmith, and Newell 2002). Although the AAD→ AB→ PIB sequence has been examined in various adver-tising contexts, no study has examined this sequencein an ethnic identity and print ad context. Hence, H3and H4 were re-tested.

H3: Peoples’ attitude toward the ad is posi-tively related to their attitudes toward theadvertised brand.


Figure 1Conceptual Model

H4: Peoples’ attitude toward the brand is posi-tively related to their intentions to pur-chase the advertised brand.

Consumers—especially those with strong ethnicidentities—who view ads with ethnically resonantactors or social settings develop stronger PIB. Relativeto weak Hispanic identifiers, strong Hispanic identi-fiers are more likely to purchase products advertisedto Hispanics (Deshpandé, Hoyer, and Donthu 1986).Blacks, Whites, and Asians tend to develop more posi-tive AAD and stronger PIB when ads depict an actor oftheir race (Martin, Lee, and Yang 2004; Whittler 1989,1991; Whittler and Spira 2002). For low involvementproducts, Black females who strongly identify withtheir ethnicity evaluate print ads containing Black fe-males more positively than print ads containing Whitefemales; these more positive assessments lead to agreater PIB (Green 1999). Yet, no study has examinedthe direct effect of ethnic identification with a printad, created by a model’s apparent ethnicity, on PIB.Thus, H5 was tested.

H5: The more (less) a person identifies ethni-cally with ads containing a model of theapparently same ethnicity, the stronger(weaker) are that person’s intentions topurchase the advertised brand.

Methodology

Ethnicities Studied: Hispanics, Blacks,and Whites

In the U.S., Hispanics and Blacks constitute twosubstantial and lucrative ethnic market segments(Torres and Gelb 2002). Hispanics represent the larg-est (13.4%) and most rapidly growing (CNN 2006)(estimated 18.9% by 2020) minority group (U.S. Cen-sus Bureau 2001); their roughly $600 billion of pur-chases in 2003 (7.4% of U.S. purchasing power) isprojected to reach $1.2 trillion by 2012, growing by

9.1% annually from 2002 to 2020 (Dodson 2007; Torresand Gelb 2002). Blacks, the second largest minoritygroup (CNN 2006), comprise 12.7% of the U.S. popu-lation (U.S. Census Bureau 2001); their $723 billion ofpurchases in 2004 is projected to reach $965 billion by2009 and $1.1 trillion by 2012 (Dodson 2007; NationalOrganization for Diversity in Sales & Marketing 2005).

Along with Hispanics and Blacks, Whites–a broadterm denoting Caucasian people as an ethnic descent,especially those with fair skin—can be viewed as anethnic group worthy of study in a multi-ethnic con-text (Brumbaugh and Grier 2006; Webster 1992; Will-iams and Qualls 1989). Whites often appear indiscussions about racial attitudes, particularly in thehumanities, and in fields such as Black Studies, CriticalRace Theory, and Whiteness Studies. At least 30 univer-sities, such as Princeton, UCLA, and University of Mas-sachusetts, now teach Whiteness Studies (Wikipedia2005). (Note: In the U.S., the standard definition ofWhites omits Hispanics and Latinos.)

Pretest

Prior to the main study, a pretest, with 47 under-graduate business students (who did not participatein the main study) from a major research university inthe southwest U.S., was conducted to identify adsthat best elicited ethnic identification via a model’sapparent ethnicity (Brumbaugh and Grier 2006). Fif-teen potential test ads—five ads for each of the threestudied ethnic groups—were tested. To avoid biasassociated with print ads for existing products, a singlefictitious product—Advance athletic shoes—appearedin all test ads. This product was chosen because of itsrelevance to the sample population (MacKenzie andLutz 1989). All test ads were derived from real adsthat were digitally modified. To enhance internal va-lidity, the test ads were polychromatic, without copy,and similar in model and product position; only themodels’ apparent ethnicity—the only ethnic cue used

EthIdenAD

PIB

AAD

AB

H4( +)

H3( +)

H5(+)

H1(+)

SEI

H2(+)

EthIdenAD

Fall 2009 57

in the ad—differed meaningfully. Data were collectedduring a regularly scheduled class.

Students rated the model in all fifteen ads forlikeability, credibility, and apparent ethnicity.Likeability was assessed with the Reysen (2005)Likeability Scale, which contains eleven Likert-typescale items ranging from (1) strongly disagree to (7)strongly agree; credibility was measured with four-teen 6-point semantic differential scale items fromOhanian (1990); and apparent ethnicity was deter-mined by asking a multiple choice question with thechoices White, Black, Hispanic, Asian, and Other. Thead chosen from each set of five ads was rated by allpretest respondents as acceptable on the first two cri-teria and highest on the third criterion. Specifically,for the test ads used in the main study, the meanpretest scores for the White model were 5.78 (out of 7)on likeability, 5.07 (out of 6) on credibility, and 96%correct for ethnic identification; for the Hispanicmodel, the scores were 5.63 on likeability, 4.75 oncredibility, and 82% correct for ethnic identification;and for the Black model, the scores were 5.51 onlikeability, 4.92 on credibility, and 98% correct forethnic identification.

Procedure and Measurement Scales

A sociology professor from a major research uni-versity in the southwest U.S. asked students enrolledin his undergraduate courses to serve as respondents.Hypothesis guessing was minimized by telling themthat the goal of the study was to obtain consumers’reactions to print ads. Exposure to the ads and comple-tion of the questionnaire occurred online within aone-week period. To access the ads and questionnaire,students received the relevant URL and a four-char-acter password. To receive extra credit, a student couldsubmit the last four digits of his or her social securitynumber, which was later matched to a university-issued class roster. Anonymity was ensured by up-loading that four-digit number to a database separatefrom the questionnaire-response database.

Once they logged onto the website, subjects wereexposed to the test ads (see Figures 2, 3, and 4) andtwo actor-and-ethnic-cue-devoid filler ads for icecream and detergent. Each ad appeared on its ownweb page. In order, subjects viewed the ice cream,Black actor, White actor, detergent, and Hispanic ac-tor ads, respectively. After viewing these ads, sub-jects indicated the ad they believed best targeted theirdemographic. (Note: Subjects could not review thepreviously shown ads). Then, for the chosen ad, theyanswered in sequence questions on EthIdenAD, AAD,

AB, SEI (test ad exclusive), and PIB. Finally, subjectsanswered demographic questions, including a closed-ended, self-designating measure of ethnicity (i.e., iden-tify oneself as belonging to one ethnic group) (Staymanand Deshpandé 1989). Regarding the studied con-structs, a new scale was developed to measureEthIdenAD, and previously developed scales were usedto assess AAD, AB, SEI, and PIB. Table 3 summarizes the24 Likert and semantic differential items comprisingthe five scales; these scales are now described.

There are two types of ethnicity: self-designatedethnicity—identifying oneself as belonging to an eth-nic group—and felt ethnicity—how strongly one iden-tifies with an ethnic group (Stayman and Deshpandé1989). Self-designated ethnicity was measured with asingle, close-ended, self-report question that asked re-spondents to indicate if they were Hispanic, Black,White, Asian, or Other. Felt ethnicity—called strength ofethnic identity (SEI) here—was measured using the fivemost context-relevant items from the Multigroup Mea-sure of Ethnic Identification scale developed in Phinney(1992). All items were measured on a seven-point Likertscale, ranging from strongly agree to strongly disagree.

Ethnic identity is an underlying sense of connec-tion to a social group (Tajfel 1978b), and its activationmay stem from peoples’ ability to associate and iden-tify with stimuli that pertain to their ethnicity (Aaker,Brumbaugh, and Grier 2000; Maldonado and Muehling2006). In marketing-related contexts, researchers haveassessed consumers’ identification with a source with itemssuch as I want to be like and My type of person (Kelmanand Eagly 1965) and Black consumers’ identificationwith Black culture with items such as perceived similarityand ability to identify (Whittler 1991). However, no multi-item scale exists for ethnic identity with an ad; thus, onewas developed (i.e., EthIdenAD).

A pretest was run on candidate scale items to exam-ine factor structure, scale reliability, and item word-ing. The same forty-seven undergraduate businessstudents who identified the test ads that best elicitedethnic identification, via a model’s apparent ethnicity,supplied the requisite data during a regularly sched-uled class. After an exploratory factor analysis withmaximum likelihood estimation and pairwise dele-tion, five items that loaded adequately on the samefactor remained. The reliability coefficient for theseitems exceeded the 0.70 threshold for preliminary re-search (Nunnally and Bernstein 1994). Based on re-spondents’ written feedback, minor wording changeswere made to some items for the main study. Theresulting seven-point Likert scales ranged fromstrongly agree to strongly disagree, very strong to veryweak, or absolutely intended to absolutely not intended.


Figure 2Black Model Ad

Figure 3Hispanic Model Ad

Fall 2009 59

Figure 4White Model Ad

AAD is the tendency to respond either favorably orunfavorably to an ad upon exposure (MacKenzie andLutz 1989). AAD was measured with the five-item,seven-point, semantic differential scale in Holmes andCrocker (1987), which was originally used to assessAAD for high and low involvement products.

AB pertains to consumers’ internal assessment of aparticular brand (Mitchell and Olson 1981); it was mea-sured with the four-item, seven-point, semantic differen-tial scale in Grier and Deshpandé (2001), which was meantto assess general attitudes about an advertised brand.

PIB relates to the likelihood that a person will buy aspecific brand (Spears and Singh 2004). PIB for athleticshoes, the product featured in each of the three testads, was measured with a five-item, seven-point, se-mantic differential scale based on three items fromHolmes and Crocker (1987) and two items fromMackenzie, Lutz, and Belch (1986).

Sample Profile

Two hundred and thirty eight undergraduate soci-ology students enrolled at a major research universityin the southwestern U.S. were asked to participate inthis study. Because the model in each test ad wasWhite, Black, or Hispanic, and two of the hypotheses

entail ethnic identification, only data from respondentsself-reporting membership in one of these three ethnicgroups were analyzed. The 31 respondents who did notmeet this requirement were dropped from the dataset;the remaining 207 respondents provided sufficient datafor effective structural equation modeling (Hair et al.2006; McQuitty 2004). Respondents tended to earn lessthan $20,000 annually (76%), be female (54%), and befrom 18 to 30 years old (97%). Their ethnic mix was 45%White (46 males, 47 females), 30% Black (28 males, 34females), and 25% Hispanic (22 males, 30 females).

The use of respondents who are undergraduate stu-dents and, for the Hispanic subgroup, second-or-more-generation ethnic immigrants, could pose a threat toexternal validity (Winer 1999). Because undergradu-ates students often are used in ethnic identity studies(see Table 2), and the product used in the test ads isgermane to the sample, the first threat is dismissible.However, the second threat could be meaningful be-cause second-generation ethnic immigrants may(Laroche et al. 1998; Phinney 1990) or may not(Rosenthal and Feldman 1992; Seitz 1998) acculturateinto their host culture and lose their ethnic identity.Unlike earlier immigrant groups, Hispanics generallyhave resisted acculturation into mainstream U.S. cul-ture (Huntington 2004; Seitz 1998). Due to their promi-


nence, Mexican Americans in the southwestern U.S.—the dominant Hispanic subgroup queried for thisstudy—are especially likely to retain a strong ethnicidentity (Huntington 2004). In other words, MexicanAmericans in the southwestern U.S. feel no pressureto acculturate because they no longer view themselvesas foreigners who must adapt to an indigenous cul-ture; thus, they retain their ethnic identity.

Results

ANOVA results for Whites (n=93), Blacks (n=62),and Hispanics (n=52) and each of the five itemsused to measure EthIdenAD revealed non-signifi-cant mean differences at the P<0.05 level; that is,for the attached item (MWhite=3.49, MBlack=3.58,

MHispanic=3.77, F(2, 204)=0.517, P=NS), the affiliateditem (MWhite=2.96, MBlack=2.98, MHispanic=3.44, F(2,204)=2.291, P=NS), the intended item (MWhite=2.97,MBlack=2.89, MHispanic=3.37, F(2, 204)=1.793, P=NS),the associate item (MWhite=3.02, MBlack=3.00,MHispanic=3.25, F(2, 204)=0.615, P=NS), and the relateitem (MWhite=3.04, MBlack=2.97, MHispanic=3.08, F(2,203)=0.102, P=NS).

Additionally, independent sample t-tests were runto examine mean scores for males (N=96) and females(N=111) for each of the five items used to measureEthIdenAD. Results indicate non-significant mean dif-ferences at the P<0.05 level for all five items; that is,for the attached item (MMale=3.69, MFemale=3.50,t(205)=0.842, P=NS), the affiliated item (MMale=3.13,MFemale=3.05, t(205)=0.364, P=NS), the intended item

Table 3Factor Loadings and Reliabilities

FactorConstruct (α) Indicator Loadings

SEI (0.71)(SEI1) I feel a strong attachment towards my own ethnic group. 0.726(SEI2) I feel good about my cultural or ethnic background. 0.406(SEI3) I have a lot of pride in my ethnic group and its accomplishments. 0.791(SEI4) I am happy that I am a member of the group I belong to. 0.885(SEI5) I have a strong sense of belonging to my own ethnic group. 0.508

EthIdenAD (0.92)(ETHIDEN1) I feel attached to the ad I chose. 0.732(ETHIDEN2) Indicate the degree to which you feel the ad you chose was intended for you. 0.777(ETHIDEN3) How strongly can you associate with the ad you chose? 0.825(ETHIDEN4) How can you relate to the ad you chose? 0.840(ETHIDEN5) I feel affiliated to the ad I chose. 0.848

AAD (0.84)(AAD1) Appealing / Unappealing 0.691(AAD2) Believable / Unbelievable 0.723(AAD3) Impressive / Unimpressive 0.758(AAD4) Attractive / Unattractive 0.691(AAD5) Overall liking / Overall disliking 0.661

AB (0.91)(AB1) Favorable / Unfavorable 0.782(AB2) Good / Bad 0.818(AB3) Pleasant / Unpleasant 0.794(AB4) High quality / Low quality 0.746

PIB (0.90)(PIB1) Would try / Would not try 0.663(PIB2) Would seek out / Would not seek out 0.766(PIB3) Very likely / Not very likely 0.811(PIB4) Probable / Improbable 0.830(PIB5) Would consider / Would not consider 0.724

Fall 2009 61

Table 4Confirmatory Factor Analysis

Constructs SEI ETHIDEN AB AAD PIB Item Reliabilities Delta (δ)

SEI1 0.61 0.372 0.628SEI2 0.35 0.123 0.877SEI3 0.68 0.462 0.538SEI4 0.99 0.980 0.020SEI5 0.35 0.123 0.877ETHIDEN1 0.77 0.593 0.407ETHIDEN2 0.68 0.462 0.538ETHIDEN3 0.93 0.865 0.135ETHIDEN4 0.95 0.903 0.097ETHIDEN5 0.87 0.757 0.243AB1 0.88 0.774 0.226AB2 0.89 0.792 0.208AB3 0.89 0.792 0.208AB4 0.76 0.578 0.422AAD1 0.84 0.706 0.294AAD2 0.65 0.423 0.577AAD3 0.75 0.563 0.437AAD4 0.79 0.624 0.376AAD5 0.61 0.372 0.628PIB1 0.81 0.656 0.344PIB2 0.84 0.706 0.294PIB3 0.92 0.846 0.154PIB4 0.87 0.757 0.243PIB5 0.65 0.423 0.577Variance Extracted 41.20% 71.60% 73.40% 53.76% 67.76%Phi (φ) MatrixSEI 1.00ETHIDEN 0.22 1.00AB 0.09 0.44 1.00AAD 0.29 0.70 0.51 1.00PIB 0.12 0.47 0.72 0.52 1.00Phi (φ)2 MatrixSEI 1.00ETHIDEN 0.05 1.00AB 0.008 0.19 1.00AAD 0.08 0.49 0.26 1.00PIB 0.01 0.22 0.52 0.27 1.00

(MMale=2.98, MFemale=3.10, t(205)=-0.594, P=NS), the as-sociate item (MMale=3.08, MFemale=3.06, t(205)=0.109,P=NS), and the relate item (MMale=3.03, MFemale=3.03,t(204)=0.021, P=NS).

Also, independent sample t-tests were run to ex-amine mean scores for males (N=96) and females(N=111) for each of the five items used to measurePIB. Results indicate non-significant mean differ-ences at the P<0.05 level for all five items; that is,

for the would try item (MMale=2.23, MFemale=2.36,t(205)=-0.650, P=NS), the would seek out item(MMale=3.10, MFemale=3.27, t(205)=-0.728, P=NS), thevery likely item (MMale=2.91, MFemale=3.05, t(204)=-0.643, P=NS), the probable item (MMale=2.72,MFemale=2.95, t(205)=-1.103, P=NS), and the would con-sider item (MMale=2.26, MFemale=2.49, t(205)=-1.152,P=NS). Based on the ANOVA and independent samplet-test results, the data were pooled for analysis.


Reliability and Validity of Scales

Principal components analysis with varimax rota-tion was used to confirm the structure of the 24 itemscomprising the five scales: SEI, EthIdenAD, AAD, AB,and PIB. Missing data were handled via pairwise dele-tion. The resulting five-factor solution, in which eachitem loaded robustly on the appropriate factor, ac-counted for 69.41% of the variance. Reliabilities foreach scale exceed the suggested 0.70 threshold forpreliminary research (Nunnally and Bernstein 1994).High factor loadings and alphas provide evidence forreliability and convergent validity. The lack of signifi-cant cross-loadings (not shown) provides evidence ofdiscriminant validity. Factor loadings and coefficientalphas are provided in Table 3.

A measurement model was estimated with LISREL8.50 and the 24 items comprising the five scales. Theaverage variance extracted (AVE) for all but the SEIconstruct (AVE=41.2%) exceeds 50% which providesfurther evidence for convergent validity (see Table 4).The AVE for all but the SEI construct and two valuesexceeds the squared correlations between each con-struct and the other constructs (see Phi and Phi2 ma-trices in Table 4), which provides further evidence fordiscriminant validity (Fornell and Larcker, 1981; Hairet al. 2006). Estimation of the measurement modelproduced the following goodness-of-fit statistics:χ2(242)=353.37 (P=0.001), comparative fit index (CFI)(0.96), non-normed fit index (NNFI) (0.95), goodness-of-fit index (GFI) (0.87), root mean square error of ap-proximation (RMSEA) (0.047), and standardized rootmean square residual (SRMR) (0.060). Collectively, thesefit statistics provide evidence of adequate model fit andvalid construct measures (Hair et al. 2006).

Structural Equation Model

The relationships shown in Figure 1 were testedusing a structural equation model with LISREL 8.50.A covariance matrix and maximum likelihood esti-mation were used to estimate model parameters. Miss-ing data were handled via pairwise deletion. The fiveconstructs—SEI, EthIdenAD, AAD, AB, and PIB—withfive, five, five, four, and five items, respectively, wereincluded in the model.

Model estimation produced the following goodness-of-fit statistics: χ2(247)=368.54 (P=0.001), CFI=0.96,NNFI=0.95, GFI=0.87, SRMR=0.067, and RMSEA=0.049.The ratio of χ2 per degree of freedom is less than two,which indicates an acceptable fit of the model to thedata (Hair et al. 2006). The SRMR and RMSEA figuresmeet the Hu and Bentler (1999) recommendation for

acceptable fit (i.e., RMSEA<0.06, and SRMR<0.09). Thecombination of these two fit statistics minimizes Type Iand II errors for interpretation of model fit (Hu andBentler 1999). The CFI and NNFI statistics also implygood model fit, although the GFI statistic implies mar-ginal model fit (Hair et al. 2006; Hu and Bentler 1999).Because the statistical power associated with theRMSEA statistic approaches 1.0, the goodness-of-fitstatistics are assumed conservative (Kaplan 1995;MacCallum, Browne, and Sugawara 1996; McQuitty2004). Therefore, overall model fit is construed as good,and the model cannot be rejected based on these data.

The path coefficients indicate that all five hypothesesare supported at the P<0.01 level (see Table 5). Specifi-cally, the stronger (weaker) a person’s ethnic identity,the more (less) that person identifies ethnically with adscontaining a model of the apparently same ethnicity(H1); the more (less) a person identifies ethnically withads containing a model of the apparently same ethnicity,the more (less) positive are that person’s attitudes to-ward the ad (H2) and intentions to purchase the adver-tised brand (H5); and peoples’ attitude toward the ad isrelated positively to their attitudes toward the adver-tised brand (H3), which in turn relate to their intentionsto purchase the advertised brand (H4).

Discussion

If consumers’ identification with marketplace stimuliaffect their purchase decisions (Forehand, Deshpandé,and Reed 2002), then related research should advanceadvertising theory and practice (Bhattacharya, Rao,and Glynn 1995). In this vein, we formulated a struc-tural model grounded in Social Identity Theory, whichhad not been attempted previously, and testedwhether or not ethnic identification with a print ad,created by a model’s apparent ethnicity, favorablyaffects viewers’ responses toward the ad and adver-tised brand. Our empirical results support our model.

Our study contributes to advertising scholarship inthree ways. First, SIT has not been applied to studiesof ethnicity and advertising effects. By testing a struc-tural model grounded in a theory that suggests selfconcept—derived from self and social identity—en-courages people to identify with and respond favor-ably to marketing stimuli, our study confirms SITtenets and thus shows the relevance of SIT to ethnicidentity and advertising contexts. Our findings sug-gest that ethnic-related impressions formed by theapparent ethnicity of ad actors can improve targetedconsumers’ responses toward the ad and advertisedbrand. For example, we show that viewers’ ethnicidentification with a print ad, based on a model’s ap-

Fall 2009 63

parent ethnicity, has a positive direct and indirect(through AAD and AB) effect on intentions to purchasethe advertised brand. Second, respondents from threedifferent ethnicities examined three ads with modelsof different ethnicities. Previous studies have reliedon two-group comparisons with an ethnic group thatnumerically dominated the region where the samplewas drawn. As a result, previous findings may be lessgeneralizable (Brumbaugh and Grier 2006). Also, whenonly two ethnic groups are represented in test ads, con-sumers may not identify most with the ad(s) containingactors of their ethnicity (Appiah 2001b). Third, a new,multi-item scale that measures ethnic identification withan ad was developed and shown to be reliable andvalid. This scale may prove useful for practitioners andresearchers in their future research endeavors.

Implications for Advertisers

Relative to other ads, ads that reinforce consumers’beliefs about their ethnic identity are noticed more andmay induce more favorable AAD and AB (Davis andGandy 1999). If ethnic identification with a print ad,created by a model’s apparent ethnicity, induces morefavorable AAD and AB, then companies targeting an eth-nic group might benefit from running ads that stresssame-ethnicity models’ or actors’ skin color, facial fea-tures, and hair styles. In addition to apparent ethnicity,a model’s/actor’s/spokesperson’s attire (e.g., hip-hop at-tire or vintage sports jerseys) or body art may further rein-force ethnic identification with an ad and featured brand.

Table 5Hypotheses Tests

StructuralHypothesis Coefficient t-statistic

H1: The stronger (weaker) a person’s ethnic identity, the more (less) that 0.23 3.05*person identifies ethnically with ads containing a model of the apparentlysame ethnicity.

H2: The more (less) a person identifies ethnically with ads containing 0.71 9.22*a model of the apparently same ethnicity, the more (less) positive arethat person’s attitudes toward the ad.

H3: Peoples’ attitude toward the ad is positively related to their attitudes 0.53 7.07*toward the advertised brand.

H4: Peoples’ attitude toward the brand is positively related to their 0.65 9.12*intentions to purchase the advertised brand.

H5: The more (less) a person identifies ethnically with ads containing 0.19 3.20*a model of the apparently same ethnicity, the stronger (weaker) arethat person’s intentions to purchase the advertised brand.

*Significant at the P<0.01 level

To create more favorable AB and PIB, companiescould use ethnic actors in ads that recognize and cel-ebrate annual ethnic-related occasions. For example,beer producers could promote Saint Patrick’s Day withads showing actors in Irish attire enjoying a lager.Similarly, book publishers could endorse Black His-tory Month with ads showing esteemed Blackspokespeople reading acclaimed Black literature. For-tunately, advertisers need not worry about non-tar-geted group backlash, as ads targeting distinct ethnicgroups should not alienate non-targeted customers(Bush, Hair, Solomon 1979; Whittler 1989).

Many ads now include virtual or animated charac-ters. To encourage customers to ethnically identifywith such ads and featured brands, these characterscould have physical traits representative of a targetedethnic group. For example, firms could use animatedcharacters in commercials that have facial featuresand skin colors similar to targeted ethnic viewers.

Limitations and Suggestionsfor Future Research

This study is not without limitations. First, the stan-dard caveats pertain regarding the use of a southwestU.S. undergraduate student sample, a single advertis-ing medium, a single advertised product, a single eth-nic cue, and a male-only ad model (Winer 1999). Toensure external validity, future research should con-sider various U.S. regions, non-student samples, non-print media, non-attire goods, additional ethnic cues,


and female models. Second, the five scales used inthis study may not be equally valid across all samplesand ethnic-related advertising settings, which mayaffect the measurement properties of the constructsand their relationships with one another. Third, al-though the EthIdenAD scale we developed showed con-vergent and discriminant validity and related aspredicted to SEI, AAD, and PIB, further research isneeded to validate this new measure. Fourth, onlytwo of the three components that comprise social iden-tity were examined (Tajfel 1978b, 1981). To gain afuller understanding of advertising effects in an eth-nic identity context, emotive as well as cognitive andevaluative facets should be studied (Agarwal andMalhotra 2005; Maldonado and Muehling 2006).

Future research could examine other constructs rel-evant to ethnicity and advertising, such as compre-hension, involvement, prejudice, productcountry-of-origin, readability, and recall (Mick 1992;Ryu, Park, and Feick 2006; White and Harkins 1994;Whittler, Calantone, and Young 1991; Whittler andDiMeo 1991; Zaichkowsky 1994). Additional theoreti-cal frameworks (e.g., implicit social cognition and the“match-up” hypothesis) could be applied to help ex-plain viewer responses to ethnic-related advertising(Greenwald and Banaji 1995; Kamins 1990). Also, manywithin-country ethnicities are multifarious becausetheir members have emigrated from different coun-tries and regions (Caetano 1986); thus, companies maychoose ineffective ethnic cues in ads when targetingsuch groups. For example, when targeting Hispanics,should advertisers use South American or Mexicanethnic cues in their ads? To examine subculture het-erogeneity and its effect on ad responses, future re-search could examine both personal ethnicity andcountry-of-origin. If targeted consumers view ads withembedded ethnic cues as exploitive, then such cuesmay induce hostility toward the ad and brand. Fi-nally, a consumer’s gender and the apparent congru-ity of an ad model’s ethnicity may interact; forexample, do women respond more favorably to adsusing male or female ethnic-similar actors?

References

Aaker, Jennifer L., Anne M. Brumbaugh, and Sonya A. Grier (2000),“Nontarget Markets and View Distinctiveness: The Impact ofTarget Marketing on Advertising Attitudes,” Journal of Con-sumer Psychology, 9 (Fall), 127-140.

Agarwal, James and Naresh K. Malhotra (2005), “An Integrated Modelof Attitude and Affect: Theoretical Foundation and an EmpiricalInvestigation,” Journal of Business Research, 58 (April), 483-493.

Appiah, Osei (2001a), “Black, White, Hispanic, and Asian Ameri-can Adolescents’ Responses to Culturally Embedded Ads,”Howard Journal of Communications, 12 (January), 29-48.

–––––––––––– (2001b), “Ethnic Identification on Adolescents’Evaluations of Advertisements,” Journal of Advertising Research,41 (September/October), 7-22.

Bhattacharya, C. B., Hayagreeva Rao, and Mary Ann Glynn (1995),“Understanding the Bond of Identification: An Investigationof its Correlates among Art Museum Members,” Journal ofMarketing, 59 (October), 46-57.

Brewer, Marilynn B. (1979), “In-Group Bias in the Minimal Inter-group Situation: A Cognitive-Motivational Analysis,” Psy-chological Bulletin, 86 (March), 307-324.

Brown, Steven P. and Douglas M. Stayman (1992), “Antecedentsand Consequences of Attitude toward the Ad: A Meta-analy-sis,” Journal of Consumer Research, 19 (June), 34-51.

Brumbaugh, Anne M. (2002), “Source and Nonsource Cues in Ad-vertising and Their Effects on the Activation of Cultural andSubcultural Knowledge on the Route to Persuasion,” Journalof Consumer Research, 29 (September), 258-269.

–––––––––––– and Sonya A. Grier (2006), “Insights from a ‘Failed’Experiment: Directions for Pluralistic, Multiethnic Advertis-ing Research,” Journal of Advertising, 35 (Fall), 35-46.

Bush, Ronald F., Robert F. Gwinner, and Paul J. Solomon (1974),“White Consumer Sales Response to Black Models,” Journal ofMarketing, 38 (April), 25-29.

–––––––––––– , Joseph F. Hair Jr., and Paul J. Solomon (1979), “Con-sumers’ Level of Prejudice and Response to Black Models in Ad-vertisements,” Journal of Marketing Research, 16 (August), 341-345.

Byrne, Donn E. (1971), The Attraction Paradigm, New York: Aca-demic Press.

Caetano, Raul (1986), “Alternative Definitions of Hispanics: Con-sequences in an Alcohol Survey,” Hispanic Journal of BehavioralSciences, 8 (December), 331-344.

Chaiken, Shelly (1980), “Heuristic versus Systematic Information Pro-cessing and the Use of Source versus Message Cues in Persuasion,”Journal of Personality and Social Psychology, 39 (November), 752-766.

CNN (2006), “Minorities Getting Closer to the Majority,” availableat: http://www.cnn.com/2006/US/05/10/hispanics/index.html (accessed March 2, 2008).

Cui, Geng (1997), “Marketing Strategies in a Multi-Ethnic Environment,”Journal of Marketing Theory and Practice, 5 (Winter), 122-134.

Davis, Jessica L. and Oscar H. Gandy, Jr. (1999), “Racial Identityand Media Orientation: Exploring the Nature of Constraint,”Journal of Black Studies, 29 (January), 367-397.

Deshpandé, Rohit, Wayne D. Hoyer, and Naveen Donthu (1986),“The Intensity of Ethnic Affiliation: A Study of the Sociologyof Hispanic Consumption,” Journal of Consumer Research, 13(September), 214-220.

–––––––––––– and Douglas M. Stayman (1994), “A Tale of TwoCities: Distinctiveness Theory and Advertising Effectiveness,”Journal of Marketing Research, 31 (February), 57-64.

Dimofte, Claudiu V., Mark R. Forehand, and Rohit Deshpandé(2004), “Ad Schema Incongruity as Elicitor of Ethnic Self-Awareness and Differential Advertising Response,” Journal ofAdvertising, 32 (Winter), 7-17.

Dodson, David (2007), “Minority Buying Power Report,” availableat: http://www.terry.uga.edu/news/releases/2007/minority_buying_power_report.html (accessed March 2, 2008).

Donthu, Naveen and Joseph Cherian (1994), “Impact of Strength ofEthnic Identification on Hispanic Shopping Behavior,” Journalof Retailing, 70 (Winter), 383-393.

Eberhardt, Jennifer L., Phillip Atiba Goff, Valerie J. Purdie, andPaul G. Davies (2004), “Seeing Black: Race, Crime, and VisualProcessing,” Journal of Personality and Social Psychology, 87 (De-cember), 876-893.

Elliott, Richard and Kritsadarat Wattanasuwan (1998), “Brands asSymbolic Resources for the Construction of Identity,” Interna-tional Journal of Advertising, 17 (Summer), 131-144.

Fall 2009 65

Forehand, Mark R. and Rohit Deshpandé (2001), “What We See MakesUs Who We Are: Priming Ethnic Self-Awareness and AdvertisingResponse,” Journal of Marketing Research, 38 (August), 336-348.

–––––––––––– , ––––––––––– , and Americus Reed II (2002), “Iden-tity Salience and the Influence of Differential Activation of theSocial Self-Schema on Advertising Response,” Journal of Ap-plied Psychology, 87 (December), 1086-1099.

Fornell, Claes and David F. Larcker (1981), “Evaluating StructuralEquation Models with Unobservable Variables and Measure-ment Error,” Journal of Marketing Research, 18 (February), 39-50.

Goldsmith, Ronald E., Barbara A. Lafferty, and Stephen J. Newell(2000), “The Impact of Corporate Credibility and CelebrityCredibility on Consumer Reaction to Advertisements andBrands,” Journal of Advertising, 29 (Fall), 43-54.

Green, Corliss L. (1999), “Ethnic Evaluations of Advertising: Inter-action Effects of Strength of Ethnic Identification, Media Place-ment, and Degree of Racial Composition,” Journal of Advertising,28 (Spring), 49-64.

Greenwald, Anthony G. and Mahzarin R. Banaji (1995), “ImplicitSocial Cognition: Attitudes, Self-Esteem, and Stereotypes,”Psychological Review, 102 (January), 4-27.

Grier, Sonya A. and Anne M. Brumbaugh (1999), “Noticing Cul-tural Differences: Ad Meanings Created by Target and Non-Target Markets,” Journal of Advertising, 28 (Spring), 79-93.

–––––––––––– and Rohit Deshpandé (2001), “Social Dimensions ofConsumer Distinctiveness: The Influence of Social Status ofGroup Identity and Advertising Persuasion,” Journal of Mar-keting Research, 38 (May), 216-224.

Hair Joseph F. Jr., William C. Black, Barry J. Babin, Rolph E. Ander-son, and Ronald L. Tatham (2006), Multivariate Data Analysis,New Jersey: Prentice Hall.

Hogg, Michael A., Deborah J. Terry, and Katherine M. White (1995),“A Tale of Two Theories: A Critical Comparison of IdentityTheory with Social Identity Theory,” Social Psychology Quar-terly, 58 (December), 255-269.

Holmes, John H. and Kenneth E. Crocker (1987), “Predispositionsand the Comparative Effectiveness of Rational, Emotional andDiscrepant Appeals for Both High Involvement and Low In-volvement Products,” Journal of the Academy of Marketing Sci-ence, 15 (March), 27-35.

Hu, Li-tze and Peter M. Bentler (1999), “Cutoff Criteria for Fit Indexes inCovariance Structure Analysis: Conventional Criteria versus NewAlternatives,” Structural Equation Modeling, 6 (Spring), 1-55.

Huntington, Samuel P. (2004), “The Hispanic Challenge,” ForeignPolicy, 141 (March/April), 30-45.

Kamins, Michael A. (1990), “An Investigation into the ‘Match-Up’Hypothesis in Celebrity Advertising: When Beauty May beOnly Skin Deep,” Journal of Advertising, 19 (Spring), 4-13.

Kaplan David (1995), “Statistical Power in Structural Equation Model-ing,” in Structural Equation Modeling: Concepts, Issues, and Applica-tions, Rick H. Hoyle, ed., California: Sage Publications, 100-117.

Kelman, Herbert C. (1961), “Process of Opinion Change,” PublicOpinion Quarterly, 25 (Spring), 57-78.

–––––––––––– and Alice H. Eagly (1965), “Attitude toward theCommunicator, Perception of Communication Content, andAttitude Change,” Journal of Personality and Social Psychology, 1(January), 63-78.

Koslow, Scott, Prem N. Shamdasani, and Ellen E. Touchstone (1994),“Exploring Language Effects in Ethnic Advertising: A SociolinguisticPerspective,” Journal of Consumer Research, 20 (March), 575-585.

Lafferty, Barbara A., Ronald E. Goldsmith, and Stephen J. Newell (2002),“The Dual Credibility Model: The Influence of Corporate andEndorser Credibility on Attitudes and Purchase Intentions,” Jour-nal of Marketing Theory and Practice, 10 (Summer), 1-12.

Laroche, Michel, Chankon Kim, Michael K. Hui, and Marc A.Tomiuk (1998), “Test of a Nonlinear Relationship betweenLinguistic Acculturation and Ethnic Identification,” Journal ofCross-Cultural Psychology, 29 (May), 418-433.

Lee, Eun-Ju, Ann Fairhurst, and Susan Dillard (2002), “Usefulnessof Ethnicity in International Consumer Marketing,” Journal ofInternational Consumer Marketing, 14 (Summer), 25-48.

Lee, Christina Kwai-Choi, Nalini Fernandez, and Brett A.S. Martin(2002), “Using Self-Referencing to Explain the Effectiveness ofEthnic Minority Models in Advertising,” International Journalof Advertising, 21 (Fall), 367-379.

Linville, Patricia W. and Edward E. Jones (1980), “Polarized Ap-praisal of Out-Group Members,” Journal of Personality and So-cial Psychology, 38 (May), 689-703.

MacCallum, Robert C., Michael W. Browne, and Hazuki M.Sugawara (1996), “Power Analysis and Determination ofSample Size for Covariance Structure Modeling,” Psychologi-cal Methods, 1 (June), 130-149.

Mackenzie, Scott B. and Richard J. Lutz (1989), “An Empirical Ex-amination of the Structural Antecedents of Attitude towardthe Ad in an Advertising Pretesting Context,” Journal of Mar-keting, 53 (April), 48-65.

–––––––––––– , Richard J. Lutz, and George E. Belch (1986), “TheRole of Attitude toward the Ad as a Mediator of AdvertisingEffectiveness: A Test of Competing Explanations,” Journal ofMarketing Research, 23 (May), 130-143.

Maldonado, Rachel and Darrel D. Muehling (2006), “The Effect ofSocial Identity Cues Under Incidental Ad Exposure Condi-tions,” Journal of Promotion Management, 12 (Fall), 77-91.

–––––––––––– , Patriya Tansuhaj, and Darrel D. Muehling (2003), “TheImpact of Gender on Ad Processing: A Social Identity Perspec-tive,” Academy of Marketing Science Review, 3, available at: http://www.amsreview.org/articles/maldonado03-2003.pdf.

Martin, Brett A. S., Christina Kwai-Choi Lee, and Feng Yang (2004),“The Influence of Ad Model Ethnicity and Self-Referencingon Attitudes,” Journal of Advertising, 33 (Winter), 27-37.

McGuire, William J. (1984), “Search for the Self: Going BeyondSelf-Esteem and the Reactive Self,” in Personality and the Pre-diction of Behavior, Robert A. Zucker, Joel Aronoff, and A.I.Rabin, eds., New York: Academic Press.

McQuitty, Shaun (2004), “Statistical Power and Structural Equa-tion Models in Business Research,” Journal of Business Research,57 (February), 175-183.

Mick, David Glen (1992), “Levels of Subjective Comprehension inAdvertising Processing and Their Relations to Ad Percep-tions, Attitudes, and Memory,” Journal of Consumer Research,18 (March), 411-424.

Mitchell, Andrew A. and Jerry C. Olson (1981), “Are Product Be-liefs the Only Mediator of Advertising Effect on Brand Atti-tude?” Journal of Marketing Research, 18 (August), 318-332.

Muse, William V. (1971), “Product-Related Response to Use of BlackModels in Advertising,” Journal of Marketing Research, 8 (Feb-ruary), 107-109.

National Organization for Diversity in Sales & Marketing (2005),“Important Facts,” available at: http://www.minoritymarketshare.com/?id=facts.

Nunnally, Jum C. and Ira H. Bernstein (1994), Psychometric Theory,New York: McGraw-Hill.

Ohanian, Roobina (1990), “Construction and Validation of a Scale toMeasure Celebrity Endorsers’ Perceived Expertise, Trustworthi-ness, and Attractiveness,” Journal of Advertising, 19 (Fall), 39-52.

Petty, Richard E. and John T. Cacioppo (1996), Attitudes and Persuasion:Classic and Contemporary Approaches, Colorado: Westview Press.


Phinney, Jean S. (1990), “Ethnic Identity in Adolescents and Adults:Review of Research,” Psychological Bulletin, 108 (November),499-514.

–––––––––––– (1992), “The Multigroup Ethnic Identity Measure–A New Scale for Use with Diverse Groups,” Journal of Adoles-cent Research, 7 (April), 156-176.

Pitts, Robert E., D. Joel Whalen, Robert O’Keefe, and Vernon Murray(1989), “Black and White Response to Culturally TargetedTelevision Commercials: A Values-Based Approach,” Psy-chology & Marketing, 6 (Winter), 311-328.

Priester, Joseph R., Dhananjay Nayakankuppam, Monique A.Fleming, and John Godek (2004), “The A2SC2 Model: The Influ-ence of Attitudes and Attitudes Strength on Consideration andChoice,” Journal of Consumer Research, 30 (March), 574-587.

Qualls, William J. and David J. Moore (1990), “Stereotyping Effectson Consumers’ Evaluation of Advertising: Impact of RacialDifferences between Actors and Viewers,” Psychology & Mar-keting, 7 (Summer), 135-151.

Reed, Americus II (2002), “Social Identity as a Useful Perspectivefor Self-Concept-based Consumer Research,” Psychology &Marketing, 19 (March), 235-266.

Reysen, Stephen (2005), “Construction of a New Scale: The ReysenLikeability Scale,” Social Behavior and Personality: An Interna-tional Journal, 33 (February), 201-208.

Rosenthal, Doreen A. and S. Shirley Feldman (1992), “The Natureand Stability of Ethnic Identity in Chinese Youth: Effects ofLength of Residence in Two Cultural Contexts,” Journal ofCross-Cultural Psychology, 23 (June), 214-227.

Roslow, Peter and J .A. F. Nicholls (1996), “Targeting the Hispanic Mar-ket: Comparative Persuasion of TV Commercials in Spanish andEnglish,” Journal of Advertising Research, 36 (May/June), 67-77.

Ryu, Gangseog, Jongchul Park, and Lawrence Feick (2006), “TheRole of Product Type and Country-of-Origin in Decisions aboutChoice of Endorser Ethnicity in Advertising,” Psychology &Marketing, 23 (June), 487-213.

Seitz, Victoria (1998), “Acculturation and Direct Purchasing Behav-ior among Ethnic Groups in the U.S.: Implications for Busi-ness Practitioners,” Journal of Consumer Marketing, 15 (January/February), 23-31.

Spears, Nancy and Surendra N. Singh (2004), “Measuring Attitudetoward the Brand and Purchase Intentions,” Journal of CurrentIssues & Research in Advertising, 26 (Fall), 53-66.

Stayman, Douglas M. and Rohit Deshpandé (1989), “SituationalEthnicity and Consumer Behavior,” Journal of Consumer Re-search, 16 (December), 361-371.

Tajfel, Henri (1978a), Differentiation between Social Groups, New York:Academic Press, Inc.

–––––––––––– (1981), Human Groups and Social Categories, UnitedKingdom: Cambridge University Press.

–––––––––––– (1978b), The Social Psychology of Minorities, New York:Minority Rights Group.

–––––––––––– and John C. Turner (1985), “The Social IdentityTheory of Group Behavior,” in Psychology of Intergroup Rela-tions, Vol. 2, Stephen Worchel and William G. Austin, eds.,Illinois: Nelson-Hall, 7-24.

Torres, Ivonne M. and Betsy D. Gelb (2002), “Hispanic-targetedAdvertising: More Sales?” Journal of Advertising Research, 42(November/December), 69-75.

Triandis, Harry C., Gerardo Marin, Judith Lisansky, and Hector Betancourt(1984), “Simpatia as a Cultural Script of Hispanics,” Journal of Per-sonality and Social Psychology, 47 (December), 1363-1375.

U.S. Census Bureau (2001), “National Population Estimates,” June, 18.Webster, Cynthia (1992), “The Effects of Hispanic Subcultural Iden-

tification on Information Search Behavior,” Journal of Advertis-ing Research, 32 (September/October), 54-62.

White, Paul H. and Stephen G. Harkins (1994), “Race of SourceEffects in the Elaboration Likelihood Model,” Journal of Per-sonality and Social Psychology, 67 (November), 790-807.

Whittler, Tommy E. (1991), “The Effects of Actors’ Race in Com-mercial Advertising: Review and Extension,” Journal of Adver-tising, 20 (Spring), 54-60.

–––––––––––– (1989), “Viewers’ Processing of Actor’s Race andMessage Claims in Advertising Stimuli,” Psychology & Market-ing, 6 (Winter), 287-309.

–––––––––––– , Roger J. Calantone, and Mark R. Young (1991),“Strength of Ethnic Affiliation: Examining Black Identifica-tion with Black Culture,” Journal of Social Psychology, 131 (Au-gust), 461-467.

–––––––––––– and Joan DiMeo (1991), “Viewers’ Reactions to Ra-cial Cues in Advertising Stimuli,” Journal of Advertising Re-search, 31 (December), 37-46.

–––––––––––– and Joan Scattone Spira (2002), “Model’s Race: APeripheral Cue in Advertising Messages?” Journal of ConsumerPsychology, 12 (Winter), 291-301.

Wikipedia (2005), “Whiteness Studies,” available at: http://en.wikipedia.org/wiki/Whiteness_studies.

Williams, Jerome D. and William J. Qualls (1989), “Middle-ClassBlack Consumers and Intensity of Ethnic Identification,” Psy-chology and Marketing, 6 (Winter), 263-286.

Winer, Russell S. (1999), “Experimentation in the 21st Century: TheImportance of External Validity,” Journal of the Academy ofMarketing Science, 27 (June), 349-358.

Zaichkowsky, Judith Lynne (1994), “The Personal Involvement In-ventory: Reduction, Revision, and Application to Advertis-ing,” Journal of Advertising, 23 (December), 59-70.

Copyright of Journal of Current Issues & Research in Advertising is the property of CTC Press and its content

may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express

written permission. However, users may print, download, or email articles for individual use.

Using a Model’s Apparent Ethnicity to Influence Viewer Responses to Print Ads: A Social Identity Theory Perspective

Documents