This a icle is dis ibu ed unde he e ms o he
C ea i e Commons A ibu ion 4.0 In e nacional License
ins i u ua
enp esa
Ins i u o de Economía Aplicada a la Emp esa
Managemen Le e s / Cuade nos de Ges ión
jou nal homepage: h p://www.ehu.eus/cuade nosdeges ion/ e is a/es/
ISSN: 1131-6837 / e-ISSN: 1988-2157
Managemen Le e s
Cuade nos de Ges ión
Enp esa Ins i u ua, UPV/EHU
Conocimien o en Ges ión/Managemen Knowledge
Volume 23 / Numbe 1 (2023) • ISSN: 1131-6837 / e-ISSN: 1988-2157
h p://www.ehu.eus/cuade nosdeges ion/ e is a/es/
The in luence o Ins ag amme s’ ecommenda ions on heal hy ood pu chase
in en ion: The ole o consume in ol emen
La in luencia de las ecomendaciones de los Ins ag amme s en las in enciones de comp a de
alimen os saludables: El papel de la implicación del consumido
Belem Ba bosaa, Eda Añaña*
a Assis an P o esso a he School o Economics and Managemen , Uni e si y o Po o - Rua D . Robe o F ias - 4200-464 – Po o – Po ugal – Telephone: +351225571 100–
e-mail ma ke ing.be[email p o ec ed]m–h p://o cid.o g/0000-0002-4057-360X
* Co esponding au ho : Associa e P o esso a Uni e sidade Fede al de Pelo as – Rua Gomes Ca nei o, 1 – 96.015-560 – Pelo as – RS – B azil – Telephone: +55 53 98405 1004–
e-mail eda [email protected]– h ps://o cid.o g/0000-0002-8298-9208
ARTICLE INFO
Recei ed 27 Feb ua y 2022,
Accep ed 27 June 2022
A ailable online 18 Janua y 2023
DOI: 10.5295/cdg.221693ea
JEL codes: M31, M37
ABSTRACT
This a icle examines he impac o digi al in luence s´ ecommenda ions, especially Ins ag amme s, on he pu -
chase in en ion o heal hy ood. In addi ion o he di ec in luence o sou ce c edibili y on beha io al in en ion,
he s udy also examines he in luence o sel -b and cong uence and consume s’ in ol emen wi h heal hy ood
on pu chase in en ion. To es esea ch hypo heses, a quan i a i e s udy was conduc ed wi h 221 Po uguese con-
sume s. High and low in ol emen wi h heal hy ood g oups we e classi ied by K-Means Clus e ing, and he
analysis o he s uc u e and he measu emen models was pe o med by using Sma -PLS so wa e. The esul s
con i med ha Ins ag amme s’ c edibili y d i es sel -b and cong uence and pu chase in en ion o heal hy ood.
I was also con i med ha he in ol emen wi h heal hy ood mode a es he in luence o sel -b and cong ui y
and Ins ag amme s’ c edibili y on consume s’ in en ion o pu chase heal hy ood, and ha b and sel -cong uence
pa ially media es he in luence o Ins ag amme s’ c edibili y on pu chase in en ion. O e all, his wo k o e s el-
e an insigh s o bo h ma ke ing manage s and esea che s, as i demons a es he impo ance o conside ing he
indi ec e ec s o sou ce c edibili y on pu chase in en ion o heal hy ood and o compa ing consume s wi h high
and low p oduc in ol emen o e ec i ely e alua e he impac o digi al in luence s’ in heal hy ood endo semen .
Keywo ds: Heal hy ea ing, digi al in luence s, in luence ma ke ing, consume in ol emen , social media, con-
sume beha io .
RESUMEN
Es e a ículo examina el impac o de las ecomendaciones p oceden es de in luence s digi ales, conc e amen e de los
Ins ag amme s, en la in ención de comp a de alimen os saludables. Además de la in luencia di ec a de la c edibilidad
de la uen e sob e la conduc a, el es udio ambién examina la in luencia de la au ocong uencia con la ma ca y la impli-
cación de los consumido es con la alimen ación saludable sob e la in ención de comp a. Pa a con as a un conjun o
de hipó esis de in es igación, se ealiza un es udio cuan i a i o con 221 consumido es po ugueses. La al a y baja
implicación con los dis in os g upos de alimen os saludables se clasi ica median e K-Means Clus e ing, y el análisis
de los modelos es uc u ales y de medida se ealiza median e el so wa e Sma -PLS. Los esul ados con i man que la
c edibilidad de los Ins ag amme s conduce a la au ocong uencia con la ma ca y a la in ención de comp a de alimen os
saludables. También se con i ma que la implicación con los alimen os saludables mode a la in luencia de la au ocon-
g uencia con la ma ca y la c edibilidad de los Ins ag amme s en la in ención de comp a alimen os saludables, y que la
au ocong uencia con la ma ca media pa cialmen e la in luencia de la c edibilidad de los Ins ag amme s en la in ención
de comp a. En gene al, es e abajo o ece in o mación ele an e an o pa a los esponsables de ma ke ing como pa a
los in es igado es, pues demues a la impo ancia de conside a los e ec os indi ec os de la c edibilidad de la uen e
en la in ención de comp a de alimen os saludables y de compa a la al a y baja implicación de los consumido es con
el p oduc o pa a e alua e icazmen e el impac o del apoyo de los in luence s digi ales en la alimen ación saludable.
Palab as cla e: Alimen ación saludable, in luence s digi ales, ma ke ing de in luence s, implicación del consumi-
do , medios sociales, compo amien o del consumido .
Managemen Le e s / Cuade nos de Ges ión 23/1 (2023) 75-86
76 Belem Ba bosa, Eda Añaña
1. INTRODUCTION
The digi al in luence indus y g ew mo e han 100% be-
ween 2019 and 2021, when i s wo ldwide es ima ed alue
eached 13.8 billion U.S. dolla s (S a is a, 2021b). On Ins ag am,
in luence ma ke ing was wo h 1.3 billion US dolla s in 2018
and was expec ed o double by 2020 (S a is a, 2021a). Unlike
adi ional celeb i ies o public igu es, digi al in luence s a e
“ egula people” who became “online celeb i ies” by c ea ing and
publishing con en on social media (Lou & Yuan, 2019). Thei
con en c ea ion and di usion s a egy enables hem o be seen
as au ho i ies and c edible in o ma ion p o ide s and endse -
e s on a ce ain opic, hence in luencing hei ollowe s’ beha -
io s. Fo ins ance, an Ins ag am in luence , hence o wa d iden-
i ied as Ins ag amme , is an Ins ag am use wi h a high numbe
o ollowe s, who has a high as e in opics such as ashion and
li es yle, which enables him/he o mone ize his/he pos ing
ac i i y (Jin e al.,2019). Acco ding o S a is a (2022), in 2020,
47.3 pe cen o all Ins ag am in luence s we e mic o-in luenc-
e s (ha ing be ween 5,000 and 20,000 ollowe s), while mid- ie
(be ween 20,000 and 100,000) and mac o-in luence s (be ween
100,000 and one million) accoun ed o 26.8 and 6.5 pe cen , e-
spec i ely, and only 0.5% o he Ins ag am in luence s had mo e
han one million ollowe s (S a is a, 2022).
Digi al in luence s, including mic o-celeb i ies, in luence
he beha io s o hei ollowe s due o hei appeal and connec-
ion wi h hei audience, as well as hei expe ise and pe cei ed
au ho i y on a pa icula opic (Dja a o a & Rushwo h, 2017;
Masuda e al.,2022; To es e al.,2019).
Schou en e al.,(2020) no e ha mic o-celeb i ies’ endo se-
men s appea o be mo e inspi ing, ela able, and us wo hy
han adi ional celeb i ies, and consequen ly mo e e ec i e
o accomplish ma ke ing objec i es. In his con ex , in luenc-
e ma ke ing comp ises he iden i ica ion o key indi iduals
and opinion leade s (Chop a e al.,2021; Lin e al.,2018) who
can d i e b and awa eness and pu chasing decisions (Chop a
e al.,2021; Lou & Yuan, 2019) wi hin a ce ain a ge audience,
namely by o p omo ing he u ili a ian and hedonic alue o he
b and’s o e ings (Lin e al.,2018).
The g owing impo ance o in luence ma ke ing has been
accompanied by calls o u he esea ch on he opic (Dwi edi
e al.,2021) o help guide managemen s a egies, namely on In-
s ag am (Dja a o a & Rushwo h, 2017), and by u he explo -
ing sou ce cha ac e is ics and hei impac on consume beha io
(Dja a o a & Rushwo h, 2017; V on is e al.,2021). In addi ion,
he ela ionship wi h he b ands ad e ised needs also o be u -
he conside ed, namely in e ms o cong ui y be ween consum-
e s’ alues and he ad e ised b ands. As endo semen p oduces
symbols, images, and meanings, which a e ans e ed om he
celeb i y o he b and, hese elemen s mus be aligned wi h he
consume ’s pe sonali y (Albe e al.,2017; P adhan e al.,2016).
This a icle ocuses on heal hy ood. The heal h and wellness
ood ma ke alue is expec ed o g ow mo e han 36% be ween
2020 and 2026, when i shall o su pass one illion U.S. dolla s
(Shahbandeh, 2021). Mo eo e , despi e ood being one o he
opics mos explo ed by digi al in luence s, esea ch abou digi-
al in luence s’ impac on ood pu chase decisions is s ill sca ce
(Cal o-Po al e al.,2021).
Conside ing he ele ance o he opic and he gaps ound
in he li e a u e, his wo k examines he impac o digi al in lu-
ence s, especially Ins ag amme s, on acili a ing he pu chase
in en ion o heal hy ood hey ecommend. In pa icula , i seeks
o answe whe he he c edibili y o Ins ag amme s, as ecom-
mende s, di ec ly in luences and/o media es he in luence o
he b and cong uence on pu chase in en ion among consume s
wi h high and low in ol emen wi h heal hy ea ing.
O e all, his a icle makes se e al con ibu ions. Fi s ly,
i con ibu es o he li e a u e on heal hy ood, a sec o ha is
g owing apidly bu s ill is insu icien ly s udied ega ding he
ac o s ha explain consume beha io , especially in digi al
con ex s and ela ed o digi al in luence s. Al hough he li e -
a u e on celeb i y endo semen is pa icula ly ich, his a icle
explo es he combina ion o endo se cha ac e is ics and con-
sume -b and cong ui y, which helps unco e he impac o dig-
i al in luence s’ endo semen s o heal hy ood b ands. Secondly,
his a icle demons a es he ele ance o conside ing media ion
e ec s in analyzing he impac o sou ce c edibili y on con-
sume beha io , pa icula ly in pu chase in en ion. As i will be
demons a ed along hese pages, his may be essen ial o accu-
a ely e alua e in luence ma ke ing e icacy. Thi dly, his a icle
demons a es he ele ance o segmen ing consume s acco ding
o hei high o low in ol emen wi h he endo sed p oduc , in
his case heal hy ood. Indeed, his a icle shows ha p oduc
in ol emen is essen ial o unde s and no only consume be-
ha io , bu pa icula ly he ou comes o in luence ma ke ing,
which is o u mos impo ance o bo h manage s and esea ch-
e s in e es ed in in luence s’ oles in os e ing consume beha -
io s, namely ega ding heal hy ood.
The nex sec ion p esen s he mos ele an con ibu ions in
he li e a u e ha guide he de elopmen o a se o hypo heses
included in a quan i a i e s udy ha is also included in he ol-
lowing pages.
2. BACKGROUND
2.1. Sou ce c edibili y and sel -b and cong ui y
Sou ce c edibili y is poin ed ou as an essen ial de e mi-
nan o he success o digi al in luence s (Schou en e al.,2020;
V on is e al.,2021; Zhu e al.,2022), as hey a e conside ed ex-
pe s who sha e hones (E dogan, 1999), us wo hy (Hudde s
e al.,2021), and high quali y (Sokolo a & Ke i, 2020) in o ma-
ion wi h hei ollowe s. Sou ce c edibili y is a e m gene ally
used o e e o he posi i e cha ac e is ics o he sende o a
message, which in luence he accep ance o he communica ed
message by he ecei e (Ohanian, 1990). As explained by E do-
gan (1999), he sou ce c edibili y model assumes ha , whene e
he in o ma ion sou ce is deemed c edible, he ecei e s accep
i and in e nalize he message in e ms o belie s, opinions, a -
i udes, and beha io s. Ohanian (1990) concep ualized i as a
h ee-dimensional cons uc , comp ising expe ise, us wo hi-
ness, and a ac i eness.
This a icle a gues ha unde s anding he impac o in lu-
ence s’ messages on he pu chase in en ions o hei audiences
can bene i o he conside a ion o bo h he pe cei ed cha ac-
Managemen Le e s / Cuade nos de Ges ión 23/1 (2023) 75-86
The in luence o Ins ag amme s’ ecommenda ions on heal hy ood pu chase in en ion: The ole o consume in ol emen 77
e is ics o he in luence , pa icula ly sou ce c edibili y, and he
cha ac e is ics o he audience ela ed o he b and being ad e -
ised. As no ed by Zhu e al.,(2022), sel -b and cong ui y s ands
ou as a possible explana ion o he lack o success o some in lu-
ence ma ke ing ini ia i es, as he e ec i eness depends on how
well he p oduc being ad e ised i s he in luence and i s au-
dience. While se e al au ho s explo ed he cong uence be ween
he b and and he in luence (e.g., Cal o-Po al e al., 2021;
Schou en e al.,2020; Zhu e al.,2022), his a icle ocus on he
consume s’ pe spec i e in e ms o sel -b and cong ui y.
Sel -b and cong ui y is poin ed ou by Yang and Jiang (2021)
as a ele an s ill unde - esea ched opic in social media li e a-
u e, s essing ha i is de e minan o ad e ising e ec i eness.
Si gy e al.,(1997) de ine sel -b and cong ui y as he pe cei ed
ma ch be ween he b and’s a ibu es and he iden i y, cha ac-
e is ics, and sel -concep o he indi idual. As explained in de-
ail by Islam e al.,(2018), cong ui y heo y pos ula es ha con-
sume s end o minimize dissonance and o be mo e inclined
o b ands ha hey pe cei e as consis en wi h hei sel -image
and hei alues. As a consequence, b ands should adop com-
munica ion s a egies ha help cus ome s iden i y a high le el o
sel -b and cong ui y. This can be done by highligh ing he al-
ues, pe sonali y, and o e all cha ac e is ics ha he b and has
in common wi h i s a ge audiences, bu also os e in e ac ion
ac oss b and- ela ed ouchpoin s ha may ein o ce he cong u-
i y pe cep ions (Islam e al.,2018). Wi h his ega d, being en-
do sed by in luence s ha he a ge audiences conside as c ed-
ible sou ces is one e ec i e way o posi i ely a ec sel -b and
cong ui y (Dwi edi e al.,2015).
By endo sing ce ain b ands, Ins ag am celeb i ies ans e
emo ional bonds wi h he b and o hei audiences. Celeb i ies’
symbolic meanings a e pa icula ly ele an o hei ollowe s
(Dwi edi e al.,2015) and hey end o be pe cei ed as ha ing
high s anda ds in picking hei endo sed b ands o which emo-
ional a achmen and meanings a e assigned (Jin e al.,2019;
Schou en e al.,2020). The aspi a ional na u e o he connec-
ions be ween in luence s and hei ollowe s also os e s a-
o able b and e alua ions (Dwi edi e al.,2015) and ollowe s’
a achmen wi h he b ands being ad e ised. In luence s’ sou ce
c edibili y is he e o e an e ec i e igge o sel -b and connec-
ions among hei ollowe s (Dwi edi e al.,2015). Li and Peng
(2021) s ess ha digi al in luence s os e consume s’ sel -b and
cong ui y p ocesses, by p omo ing and displaying he use o he
p oduc s and se ices o hose b ands, namely due o he digi-
al in luence s’ pe cei ed c edibili y and us wo hiness. Addi-
ionally, consume s c ea e s ong posi i e associa ions ega ding
c edible sou ces wi h which hey aspi e o iden i y wi h. In line
wi h hese con ibu ions, i is expec ed ha :
H1 – The Ins ag amme s’ c edibili y has a posi i e impac on
consume s’ sel -b and cong ui y.
2.2. Expec ed impac s on pu chase in en ion
Since he seminal wo k by Ohanian (1990), sou ce c edibili y
has ecei ed con inuous a en ion by schola s s udying he im-
pac o celeb i ies on consume beha io , including due o he
expec ed impac s on pu chase in en ion. Indeed, he li e a u e
a gues ha he expec ed posi i e impac o Ins ag amme s on
pu chase in en ions o hei ad e ised b ands depends on how
c edible hei ollowe s pe cei ed hem (Dja a o a & Rushwo h,
2017; Schou en e al.,2020), especially ega ding hei us wo -
hiness and pe cei ed expe ise (Masuda e al.,2022). By adop -
ing in luence ma ke ing, b and manage s aim o ans e he
image o he in luence and his/he s a o able cha ac e is ics
on o he b and (Schou en e al.,2020), and o achie e posi i e
b and esponses, namely in e ms o u u e pu chases. Sou ce
c edibili y builds consume us (Kim & Kim, 2021), and ul-
ima ely explains message accep ance (Ohanian, 1990) and be-
ha io al in en ions. As a consequence, he li e a u e on digi al
in luence s s esses pu chase in en ion as one o he ele an
ou comes o sou ce c edibili y (AlFa aj e al.,2021; Dja a o a
& Rushwo h, 2017; Masuda e al.,2022; Pick, 2021; Schou en
e al.,2020; Sokolo a & Ke i, 2020; Weismuelle e al.,2020; Zhu
e al.,2022). Al hough he opic o heal hy ood is unde s ud-
ied, he e is e idence o he posi i e impac o sou ce c edibili y
on pu chase in en ion namely ega ding ood p oduc s (Cal o-
Po al e al.,2021). Hence, and in line wi h hese con ibu ions,
i is expec ed ha :
H2 – The Ins ag amme s’ c edibili y posi i ely a ec s con-
sume s’ in en ion o buy he heal hy ood hey ecommend.
Cong ui y wi h he endo sed b and enables he consume o
app op ia e symbols and meanings o enhance his/he sel -con-
cep (Si gy, 1982) and suppo his/he iden i y cons uc ion
(Albe e al.,2017). As explained by Belk (1988), b ands ha e
pe sonal and social meanings, which a e used by consume s o
c ea e, enhance, o communica e hei own iden i y. Sel -b and
cong ui y acili a es he ini ia ion and de elopmen o cus om-
e -b and ela ionships (Gho bani e al., 2022), posi i ely a -
ec s b and e alua ions (Agui e-Rod iguez e al.,2012; Islam
e al.,2018), and os e s consume p e e ence (Li e al.,2021).
Consequen ly, sel -b and cong ui y is associa ed wi h a g ea e
willingness o choose hose p oduc s when making pu chas-
es, hence inc easing pu chase in en ion (Agui e-Rod iguez
e al.,2012; Michel e al.,2022; Phua & Kim, 2018; Si gy, 1982).
In line wi h hese con ibu ions, i is expec ed ha :
H3 – Sel -b and cong ui y inc eases consume s’ in en ion o
buy heal hy ood.
The in eg a ion o he h ee p e iously s a ed hypo heses an-
icipa e a media ion e ec o sel -b and cong ui y, and conse-
quen ly he conside a ion o bo h di ec and indi ec e ec s o
sou ce c edibili y on pu chase in en ion. Howe e , he assump-
ions o a media ion e ec should be con i med o make such an
assump ion. Acco ding o Ba on and Kenny (1986), a media ion
e ec exis s when: (a) a ia ions in he le els o an independ-
en a iable a e signi ican ly accounTable o a ia ions in he
p esumed media o (i.e., Pa h a), (b) a ia ions in he media o
a e signi ican ly accounTable o a ia ions in he dependen
a iable (i.e., Pa h b), and (c) when Pa hs a and b a e con olled
o , a p e iously signi ican ela ionship be ween he independ-
en and dependen a iables is no longe signi ican . As such, a
ull media ion occu s when he di ec e ec c is no signi ican ,
whe eas he indi ec e ec a × b is signi ican , which is an indica-
ion ha only he indi ec e ec ia he media o exis s. All o he
si ua ions unde he condi ion ha bo h he di ec e ec c and
he indi ec e ec a × b a e signi ican ep esen pa ial media-
ion (Ca ión e al.,2017). By assuming he media ion e ec o
Managemen Le e s / Cuade nos de Ges ión 23/1 (2023) 75-86
78 Belem Ba bosa, Eda Añaña
sel -b anding, his a icle posi s ha he expec ed impac s o dig-
i al in luence s’ sou ce c edibili y on hei ollowe s’ pu chase in-
en ion (e.g., AlFa aj e al.,2021; Dja a o a & Rushwo h, 2017;
Masuda e al.,2022; Pick, 2021; Schou en e al.,2020; Sokolo a
& Ke i, 2020; Weismuelle e al.,2020; Zhu e al.,2022) is also
explained by he ac ha sou ce c edibili y a ec s sel -b and
cong uence (Agui e-Rod iguez e al.,2012; Michel e al.,2022;
Phua & Kim, 2018; Si gy, 1982). Consequen ly, he indi ec e -
ec s o sou ce c edibili y on pu chase in en ion ough o be con-
side ed o e alua e he impac s o celeb i y endo semen s. As
such, i is expec ed ha :
H4 – The sel -b and cong ui y media es he in luence o Ins-
ag amme s’ c edibili y in he in en ion o buy heal hy ood.
2.3. The in ol emen wi h heal hy ood (IWHF)
In ol emen is used o desc ibe he le el o consume in e -
es , sea ch in ensi y, and he complexi y o he decision-mak-
ing p ocess owa d an objec (e.g., heal hy ood). The objec
o in ol emen may be a p oduc , a se ice, a si ua ion, o an
ad e isemen . Low in ol emen implies ine ia, and high in-
ol emen implies a g ea deal o ac i a ion and i is ound o
in luence consume decision making and in e ac i e communi-
ca ions (Zaichkowsky, 2010). Acco ding o Zaichkowsky (2010),
low in ol emen implies ine ia, and high in ol emen implies
a g ea deal o ac i a ion and i is ound o in luence consume
decision making and in e ac i e communica ions.
In ol emen ’s in luence on consump ion expe iences is bes
illus a ed by he psychological consequences e oked by a p od-
uc ’s heigh ened ele ance o he consume .
P oduc in ol emen is he in e es a consume inds in a
p oduc class. Such in e es s ems om he consume ’s pe cep-
ion ha he p oduc class mee s impo an alues and goals.
P oduc in ol emen and pu chase-decision in ol emen may
occu as sepa a e en i ies (Mi al & Lee, 1989). Because o i s
psychological implica ions, a mo e in ol ing p oduc is likely o
lead o a mo e delibe a i e e alua ion.
As such, in ol emen is no an e alua i e dimension pe se;
a he , i me ely enhances e alua ion and subsumes e alua ion’s
dimensions (Mano & Oli e , 1993). Acco ding o K ugman
(1965), as ci ed by G eenwald and Lea i (1984), o he high-in-
ol emen consume s, a communica ion should ac mos di ec -
ly o modi y belie s (i.e., e balizable p oposi ions), and among
he less in ol ed ones, he impac should be mo e on pe cep ions
(such as b and logos o package con igu a ions) and should oc-
cu mo e g adually, being e ec i e only wi h epea ed exposu es.
The exp ession “heal hy ea ing” may ake di e en meanings,
depending on he belie s and in e es s o hose who de ine i . The
ood can be associa ed wi h he ype (e.g., ui , ege ables, ani-
mal oods, sa e oods, unc ional oods), he nu ien s (e.g., gen-
e al nu ien s, ibe , i amins, mine als, a s, ca bohyd a es), o
o he componen s (e.g., glu en, addi i es, oxins). And oods can
also be classi ied as heal hy by he way hey a e p ocessed (e.g.,
na u al, homemade, o ganic) (Bisogni e al.,2012).
Hence, in ol emen is he ele ance pe cei ed by a pe son
conce ning an objec , based on his needs, alues, and in e es s.
The g ea e he in ol emen wi h an objec , he g ea e he en-
gagemen wi h i (Co ea e al.,2020). When consume s ha e a
s ong in e es in he con en , hey ind he in o ma ion on he
p oduc ele an o he con en mo e in e es ing, and in u n gi e
mo e a o able esponses o an ad e (Choi & Ri on, 2002). Fo
his eason, his a icle elec s he ac onym IWHF as a su oga e
o he in ol emen wi h heal hy ood, ega dless o wha one
unde s ands by “heal hy ood”. Indeed, people’s in e p e a ions
o heal hy ea ing a e no a simple se o belie s ha can be judged
as co ec o inco ec acco ding o he ways he scien is s de-
sc ibe i . Ins ead, people hold complex, mul i ace ed belie s and
eelings abou his opic ha may be connec ed o o he pa s o
hei li es and be based on pe sonal knowledge gained h ough
hei li e expe iences (Bisogni e al.,2012).
In line wi h hese con ibu ions, his s udy a gues ha in-
ol emen is a use ’s s a e o mind ha esul s in a pe sonally
meaning ul bene i , a use ’s pe cep ion ha his/he ole is im-
po an o mee his/he needs (Di Gangi & Wasko, 2016), and o
ha eason, i is expec ed ha :
H5 – Consume s’ IWHE mode a es he in luence o sel -
b and cong ui y on consume s’ in en ion o buy heal hy ood.
H6 – Consume s’ IWHE mode a es he in luence o Ins a-
g amme s’ c edibili y on he sel -b and cong ui y.
H7 – Consume s’ IWHE mode a es he impac o Ins ag am-
me s’ c edibili y in he consume s’ in en ion o buy he heal hy
ood hey ecommend.
By combining his se o hypo heses, his a icle sugges s
s udying he in luence o Ins ag amme s’ c edibili y on he pu -
chase in en ion o heal hy ood by hei ollowe s, conside ing
consume s’ in ol emen wi h ha ype o ood, and cong uence
be ween he ad e ised b and and consume s’ alues. Mo eo e ,
Ins ag amme s’ c edibili y is concep ualized as a second-o de
ac o (a ac i eness, expe ise, and us wo hiness), in line
wi h ex an li e a u e (Cal o-Po al e al.,2021; Ohanian, 1990;
O azi & New on, 2018). Figu e1 ou lines he concep ual model
p oposed o his a icle. To imp o e eadabili y, he lines co e-
sponding o he second-o de ela ionships ha e been dashed,
and he numbe s o he hypo heses we e placed nex o he co -
esponden s uc u al ela ionship.
Figu e1
Ou line o he concep ual model p oposed o his a icle
Sou ce: Own elabo a ion.
Managemen Le e s / Cuade nos de Ges ión 23/1 (2023) 75-86
The in luence o Ins ag amme s’ ecommenda ions on heal hy ood pu chase in en ion: The ole o consume in ol emen 79
3. METHOD
To es he esea ch hypo heses de ined o his s udy, a quan-
i a i e and explo a o y esea ch app oach was adop ed.
3.1. Ma e ials and measu es
Da a we e collec ed online, h ough a ques ionnai e con-
aining 25 i e-poin Like - ype scala ques ions. The ques ion-
nai e was o ganized in o wo pa s. The i s comp ised ques-
ions o sample cha ac e iza ion, such as demog aphics and
hei Ins ag am usage. Only esponden s ha ollowed a leas
one Ins ag amme ela ed o heal hy ood and li es yle quali ied
o his s udy. Responden s we e asked o iden i y one o such
Ins ag amme s hey ollowed, who would become associa ed
wi h he emainde ques ions posed (e.g., Ins ag amme c edi-
bili y). Mo eo e , his s udy does no go in o he me i s o wha
should o should no be conside ed heal hy, being based on he
esponden s’ in e p e a ions o he exp ession “heal hy ood”
(Bisogni e al.,2012). As such, he ques ionnai e did no p o ide
sugges ions o heal hy ood and ocused on he gene al concep ,
om he pa icipan s’ pe spec i e.
The ques ionnai e ea u ed measu emen scales adap -
ed om o he au ho s o each o he a iables in he con-
cep ual model. The in ol emen o consume s wi h heal hy
ood —IWHF— was assessed by a i e-poin in e al scale,
adap ed om O’Cass (2000), anging om s ongly disa-
g ee o s ongly ag ee. Fou o he 12 ques ions on he scale,
ep esen ed in ol emen wi h he p oduc , ou ep esen ed
he in ol emen wi h consump ion, and ano he ou ep e-
sen ed he in ol emen in pu chasing decision o ha ype
o p oduc . The scale ep esen ing he sel -b and cong ui y
( ou i ems) was bo owed om Si gy e al.,(1997), acco d-
ing o he adap a ion by Taylo e al.,(2011), and pu chase
in en ion was e alua ed by h ee i ems adap ed om Dodds
e al.,(1991). The Ins ag amme s’ c edibili y was ep esen -
ed by nine a iables bo owed om Ohanian (1990), h ee o
hem ep esen ing hei a ac i eness, h ee e e ing o hei
expe ise, and h ee o he s e e ing o hei us wo hiness.
I should be no ed ha , despi e i s age, Ohanian’s scale is a
eliable and alid measu emen o celeb i y c edibili y, a opic
ha has shi ed o he online con ex in ecen yea s (Halde
e al.,2021). This scale is an essen ial ool o cha ac e ize ce-
leb i ies and endo se s, and is equen ly conside ed by digi al
in luence s udies (e.g., Chung & Cho, 2017; Jin & Phua, 2014;
J.-S. Lee e al.,2022; J. A. Lee e al.,2022; Lee e al.,2021; Li &
Peng, 2021; Ma, 2021; O azi & New on, 2018; Shi e al.,2021)
o e ing a ele an con ibu ion o u he unde s and online
consume beha io .
The ques ionnai e was subjec o p e es wi h 20 indi iduals
wi h cha ac e is ics simila o esponden s, who con i med he
cla i y o he ins uc ions and he ques ions posed.
3.2. Pa icipan s
Da a we e collec ed in No embe 2019. Indi iduals we e in-
i ed o pa icipa e in he s udy using a snowball me hod. One
o he Au ho s app oached a o al o 25 indi iduals om he
social ne wo ks and asked hem o sha e he in i a ion wi h 10
indi iduals o hei ne wo k, who we e eques ed o epea his
p ocedu e.
E hical p ocedu es we e applied o his s udy, ollowing he
p ocedu es usually applicable o esea ch in social sciences (B y-
man, 2021). Pa icipa ion was olun a y, con iden ial, and anon-
ymous. Comple e in o ma ion on he na u e and objec i es o
he s udy we e p o ided be o e posing he ques ions. Only e-
sponses om Ins ag am use s ha ollowed a leas one Ins a-
g amme ela ed o heal hy ood we e conside ed.
As a esul o his collabo a ion p ocess, a o al o 221 alid
esponses we e ob ained. The sample comp ises 221 Po u-
guese adul s, p eponde an ly emale (75%). Rega ding edu-
ca ion, 32% o he esponden s had up o seconda y school,
37% we e g adua es, and 32% we e pos -g adua es (including
mas e s’ deg ee and PhD holde s), which is cohe en wi h he
medium o high le el o educa ion in he coun y. As expec ed
among Ins ag am use s, he pa icipan s ended o be young
adul s: 30% we e 18-26 yea s-old, 20% we e 23-26 yea s old,
12% we e 27-30 yea s old, 26% we e 31-40, and 12% we e old-
e han 40. The majo i y o he pa icipan s had a p o essional
ca ee (63%) and he emaining 37% we e s ill s udying. Fi-
nally, pa icipan s had di e se p o iles o Ins ag am usage,
spending up o one hou (43%), be ween 1 and 2 hou s (36%)
and mo e han 2 hou s (21%) pe day on Ins ag am. As expec -
ed, he IWHF did no show signi ican dependence on gende
(χ2=1.175; DF=1; Sig=0.278) o on educa ional le el (χ2=0.034;
DF=2; Sig=0.983). Table1 p esen s he dis ibu ion o he sam-
ple by gende and IWHF.
Table1
Sample by gende and le el o in ol emen wi h heal hy ood
Le el o IWHF Gende To al
Female Male
Low IWHF Coun 56 23 79
Expec ed Coun 59.3 19.7 79
High IWHF Coun 110 32 142
Expec ed Coun 106.7 35.3 142
To al Coun 166 55 221
Expec ed Coun 166 55 221
Sou ce: Own elabo a ion.
The p ocedu es o iden i ying hese le els o IWHF a e ex-
plained below.
3.3. P ocedu es
Once he in ol emen wi h heal hy ood was cap u ed as
me ically scaled indica o s, and gi en ha consume ’s le el o
in ol emen can be used as a segmen ing a iable (Zaichkowsky,
2010), and aking in o conside a ion ha he mechanism o
communica ion impac o low in ol emen is di e en om
ha o high in ol emen (G eenwald & Lea i , 1984), be o e
analyzing he main da a, he a iables ep esen ing IWHF we e
g ouped, dicho omized, and ans o med in o a ca ego ical a i-
able o high and low in ol emen (Hensele e al.,2016). Taking
Managemen Le e s / Cuade nos de Ges ión 23/1 (2023) 75-86
80 Belem Ba bosa, Eda Añaña
in o conside a ion ha da a ep esen ing he IWHF we e cap-
u ed by 12 a iables, he iden i ica ion o consume s o high
and low in ol emen was done by K-Means Clus e ing Analysis
(Assake & Hallak, 2013), which iden i ied a majo i a ian g oup
o 142 consume s (64%) o high in ol emen , and ano he o low
in ol emen comp ising 79 consume s (36%). Jus in case, he
IWHF o ming a iables we e compa ed be ween he high and
low in ol emen consume s (Figu e2) and e alua ed by - es ,
and all o hem showed signi ican di e ences in means (p<0.00)
amongs he wo g oups.
Figu e2
Means o a iables ac oss he g oups o low and
highin ol emen wi h heal hy ood
Sou ce: Own elabo a ion.
The analysis in ol ed h ee s eps: (a) Clus e Analysis, by
K-Means Clus e ing, o iden i y consume s o low and high in-
ol emen wi h heal hy ood; (b) analysis o he measu emen
model, h ough Con i ma o y Fac o Analysis (CFA), o assess
he eliabili y and alidi y o he ac o s; (c) s uc u al model
analysis, o assess he hypo hesized ela ionships; and (d) com-
pa a i e analysis ( - es ), o assess he in a iance o he in luence
ela ionships, be ween he high and low in ol emen g oups
(Chin, 2000). Bo h he measu emen model and he s uc u al
model we e designed and analyzed in Sma -PLS 3.3.3 so wa e
(Ringle e al.,2015).
4. RESULTS
Bo h he measu emen and he s uc u al models we e e al-
ua ed wi h he Sma PLS-3.3.3 package (Ringle e al., 2015),
ecommended by Hai e al.,(2014) conside ing he sui abili y
o ha algo i hm o es ima e models wi h small samples and
non-no mal da a, as in his case. Con e gen alidi y was a es -
ed by he alue o he ac o loadings o he measu emen model
( he ou e model), all o which we e g ea e han 0.7, and by he
A e age Va iance Ex ac ed-AVE g ea e han 0.5 in all he ac-
o s (Hai e al.,2014), as shown in Table2.
Table2
Fac o s i and eliabili y
Fac o s C onbach’s
Alpha
Composi e
Reliabili y
A e age
Va iance
Ex ac ed
(AVE)
R2
Ins ag amme
A ac i eness 0.858 0.912 0.776 0.306
Ins ag amme
C edibili y 0.903 0.923 0.589 0.042
Ins ag amme
Expe ise 0.932 0.957 0.881 0.833
Ins ag amme
T us wo hiness 0.949 0.967 0.908 0.884
Pu chase In en ion 0.915 0.946 0.855 0.457
Sel -B and Cong ui y 0.932 0.952 0.831 —
Sou ce: Own elabo a ion.
The Composi e Reliabili y indices and C onbach’s Alphas
exceeded he expec ed minimum o 0.7 in all ac o s, and he
Coe icien s o De e mina ion - R2 o he Pu chase In en ion, he
mo e impo an endogenous a iable, is a ound 0.5, which can
be classi ied as mode a e explana o y powe (Hai e al.,2019;
Hai e al.,2014). The disc iminan alidi y was assessed by he
c i e ion p oposed by Fo nell and La cke (1981), which com-
pa es he co ela ions be ween he ac o s o he squa e oo s o
he co esponding AVEs. By his c i e ion, he e is disc iminan
alidi y when each AVE squa e oo su passes he co esponding
co ela ions be ween he ac o and all he o he s. As shown in
Table3, he squa e oo s o he AVE (in bold on he main diag-
onal), exceed he co ela ions in all he co esponding ows and
columns, excep o he ac o s e e ing o he Ins ag amme
C edibili y (b), he highe -o de componen o which hey a e a
pa o Sa s ed e al.,(2019, p. 200), which a es s he exis ence
o disc iminan alidi y in he model.
Table3
Fac o s’ Co ela ions and AVE’ Squa e Roo s
Fac o s (a) (b) (c) (d) (e) ( )
Ins ag amme
A ac i eness (a) 0.881
Ins ag amme
C edibili y (b) 0.553 0.768
Ins ag amme
Expe ise(c) 0.290 0.913 0.938
Ins ag amme
T us wo hiness (d) 0.385 0.940 0.805 0.953
Pu chase In en ion (e) 0.045 0.350 0.366 0.357 0.925
Sel -B and Cong ui y ( ) 0.024 0.205 0.210 0.215 0.638 0.912
Sou ce: Own elabo a ion.
Managemen Le e s / Cuade nos de Ges ión 23/1 (2023) 75-86
The in luence o Ins ag amme s’ ecommenda ions on heal hy ood pu chase in en ion: The ole o consume in ol emen 81
The mode a ion e ec was e alua ed by Mul i-G oup Anal-
ysis (MGA), a ype o analysis whe e he mode a o a iable
is ca ego ical and is assumed o po en ially a ec all ela ion-
ships in he inne model (Hai e al.,2014). In gene al e ms, a
mode a o is a quali a i e (e.g., sex, ace, class) o quan i a i e
(e.g., le el o ewa d) a iable ha a ec s he di ec ion and/o
s eng h o he ela ionship be ween an independen and a de-
penden o c i e ion a iable (Ba on & Kenny, 1986). Con a y
o he s anda d mode a ion, he MGA o e s a mo e comple e
pic u e o he mode a o ’s in luence, once he analysis emb ac-
es all he ela ionships in he model, and no a speci ic ela-
ionship be ween an exogenous and an endogenous a iable
(Cheah e al.,2020).
The con igu al and composi ional in a iances we e assessed
by he pe mu a ion algo i hm a ailable in Sma -PLS 3.3.3 (Rin-
gle e al., 2015). The pe mu a ion algo i hm allows o es i
p e-de ined da a g oups ha e s a is ically signi ican di e ences
in hei g oup-speci ic pa ame e es ima es (e.g., ou e weigh s,
ou e loadings, and pa h coe icien s) (Ringle e al.,2015). The
con igu al in a iance equi es ha he same basic ac o s uc-
u e exis s in all he g oups (in e ms o he numbe o cons uc s
and i ems associa ed wi h each cons uc ); and composi ional in-
a iance means ha he p esc ip ion o condensing he indica-
o a iables in o composi es is he same o all g oups (Hensele
e al.,2016).
The con igu al in a iance en ails ha a composi e, which
has been speci ied equally o all he g oups, eme ges as a un-
idimensional en i y in he same nomological ne ac oss all he
g oups. The composi ional in a iance (MICOM) is accep ed i
he co ela ion be ween he sco es o each cons uc does no
di e signi ican ly om he uni y, wi hin he ou e weigh s o
each g oup. As can be seen in Table4, he e a e no di e ences
in he composi ion o bo h g oups, once he co ela ions o in-
dica o s wi hin g oups do no di e om one.
Table4
Resul s o MICOM (pe mu a ion) P ocedu e
MICOM O iginal
Co ela ion
Co ela ion
Pe mu a ion
Mean
5% Pe mu a ion
p-Values
Ins ag amme
A ac i eness 0.996 0.993 0.992 0.156
Ins ag amme
C edibili y 0.998 0.997 0.989 0.362
Ins ag amme
Expe ise 1.000 1.000 1.000 0.306
Ins ag amme
T us wo hiness 1.000 1.000 1.000 0.388
Pu chase
In en ion 1.000 1.000 0.999 0.346
Sel -B and
Cong ui y 1.000 1.000 0.999 0.567
Sou ce: Own elabo a ion.
A e es ing he in a iance o assu e ha indica o s we e
app op ia e o bo h g oups o esponden s, he nex s ep was o
e alua e he ela ionships e e ing o he main se o hypo he-
ses. The s uc u al model analysis a es ed o he signi icance o
all he s uc u al ela ionships ( he inne model). As shown in
Table5, he ela ionships o Ins ag amme s’ c edibili y wi h he
ac o s i ela es o as a second-o de ac o a e all signi ican
(p < 0.00). The in luence o Ins ag amme s’ c edibili y on sel -
b and cong ui y and pu chase in en ion is signi ican , which
con i ms he hypo heses H1 and H2; and he in luence o sel -
b and cong ui y in pu chase in en ion is also signi ican , which
con i ms H3.
Table5
S uc u al Rela ionships
S uc u al Rela ionships Sample
Mean
S and.
De ia . P
Ins . C edibili y
Ins .A ac i eness 0.553 0.082 6.72 0.000
Ins . C edibili y
Ins .Expe ise 0.913 0.015 61.3 0.000
Ins . C edibili y
Ins .T us wo hiness 0.940 0.009 109 0.000
Ins . C edibili y
Pu chaseIn en ion (H2) 0.229 0.063 3.62 0.000
Ins . C edibili y
Sel -B andCong ui y (H1) 0.205 0.081 2.53 0.014
Sel -B and Cong ui y
Pu chase In en ion (H3) 0.591 0.066 8.89 0.000
Speci ic Indi ec E ec s
Ins ag amme C edibili y
Sel -B and Cong ui y
Pu chase In en ion (H4)
0.122 0.047 2.596 0.01
No e: (1) “Ins ” = “Ins ag amme s”; (2) “S and. De ia .” = “S anda d
De ia ions”
Sou ce: Own elabo a ion.
Resul s also con i med ha Ins ag amme s’ c edibili y ans-
e s 12.2% o i s load indi ec ly o he pu chase in en ion, by he
sel -b and cong ui y. This indi ec e ec a es s ha sel -b and
cong ui y is a pa ial media o be ween Ins ag amme s’ c edibil-
i y and pu chase in en ion, which con i ms H4 (p=0.01).
The di e ences be ween pa hs in he models es ima ed o
consume s o low and high IWHF we e e alua ed by - es , ac-
co ding o he o mula p oposed by (Chin, 2000). As pe Ebe l
(2010, p. 497), he s a is ic o his app oach shall be cons uc ed
acco ding o he s anda d e o s o he pa h es ima o s in he
wo subg oups, o assess whe he hey a e equal o no . The es
s a is ic is calcula ed as ollows:
Managemen Le e s / Cuade nos de Ges ión 23/1 (2023) 75-86
82 Belem Ba bosa, Eda Añaña
Compa ing he pa h in he models es ima ed o g oups
o high and low in ol emen , he s udy iden i ied signi ican
di e ences ( > 1.96) in he in luence o Ins ag amme s’
c edibili y and he sel -b and cong ui y, on he pu chase in-
en ion o heal hy ood. As shown in Table6, he Ins ag am-
me s’ c edibili y ans e s 11.8% o i s weigh o pu chase
in en ion in he low-in ol emen g oup, and 32.9% among
he high-in ol emen consume s; and he sel -b and con-
g ui y inc eases 71.4% o he pu chase in en ion among he
low-in ol emen consume s and 45.6% among he high-in-
ol emen ones.
Table6
Compa ison o in luences in high and low IWHF g oups
S uc u al
Rela ionships
High IWHF Low IWHF
O iginal
Sample
S anda d
De ia ion
O iginal
Sample
S anda d
De ia ion
Ins ag amme
C edibili y
Ins ag amme
A ac i eness
0.503 0.109 0.611 0.109 –1.339
Ins ag amme
C edibili y
Ins ag amme
Expe ise
0.926 0.012 0.898 0.035 1.081
Ins ag amme
C edibili y
Ins ag amme
T us wo hiness
0.948 0.011 0.929 0.015 1.712
Ins ag amme
C edibili y
Pu chase
In en ion(H7)
0.329 0.083 0.118 0.08 3.565
Ins ag amme
C edibili y
Sel -B and
Cong ui y (H4)
0.235 0.097 0.116 0.158 1.018
Sel -B and
Cong ui y
Pu chase
In en ion (H5)
0.456 0.09 0.714 0.08 –4.359
Sou ce: Own elabo a ion.
In o he wo ds, he in luence o Ins ag amme s’ c edibili y
on he pu chase decision was ound o be almos h ee imes
highe among consume s in ol ed wi h heal hy ood han in he
low in ol emen g oup, which con i ms H7. On he o he hand,
he impac o sel -b and cong uence on he pu chase in en ion
is almos wice as impo an o he low-in ol emen consum-
e s, han o he mo e in ol ed ones, which con i ms H5. The
hypo hesis H6, ha IWHF should mode a e he in luence o
Ins ag amme s’ c edibili y on he sel -b and cong ui y was no
con i med as he di e ence did no achie e he minimum cu -
o poin ( > 1.96).
To acili a e he in e p e a ion he au ho s eplica ed
manually he es ima es o high and low-in ol emen g oups,
abo e and below he co e es ima es in Figu e3.
Figu e3
S uc u al model
Sou ce: Own elabo a ion.
As shown in Figu e 3, when using he comple e sample
( he co e loads) all he s uc u al ela ionships a e signi ican
(p<0.00), and he same occu s o he high-in ol emen g oup.
Howe e , he in luence o Ins ag amme s’ c edibili y on he sel -
b and cong ui y, and he in luence o he o me on he pu chase
in en ion, did no achie e he s a is ical signi icance (p < 0.05)
among he low-in ol emen g oup.
In summa y, i can be said ha :
a) Ins ag amme s’ c edibili y s eng hens he pu chase in en ion
among consume s wi h high le els o IWHE, bu i does no
a ec ha decision among consume s o low in ol emen .
b) Sel -b and cong ui y signi ican ly imp o es consume s’ pu -
chase in en ion ega dless o hei le els o in ol emen , bu
i s e ec is no iceable highe among he consume s o low
IWHE, he ones whose pu chase in en ions a e no a ec ed
by Ins ag amme s’ c edibili y.
These indings a e in line wi h Co ea e al.,(2020) o whom
he sole engagemen wi h an in luence does no necessa ily
mean ha a ollowe will buy he b ands he/she ecommends
since a close iden i ica ion o he consume is necessa y. In-
deed, he compa ison o he means o ac o s among g oups
con i ms ha consume s mo e in ol ed wi h heal hy ood a e
mo e cong uen wi h he b ands associa ed wi h his class o
p oduc s and mo e p one o pu chase hem, which could be ob-
ious and is in line wi h Mi al and Lee (1989). Howe e , i does
cause some su p ise he lack o di e ences in he e alua ions o
Ins ag amme s by high and low in ol emen consume s, gi en
ha hei in luence on he subsequen cons uc s di e s be ween
one g oup and he o he .
The appa en con o e sy ha is es ablished, when he e-
sul poin s ou di e ences in he impac s o Ins ag amme s’
c edibili y on o he cons uc s, despi e no exp essi e di e enc-
es exis ing in he e alua ion o he o me , sugges s ha he
c edibili y o hese in luence s a ec s pu chase in en ion mo e
by he way i co ela es o he choices o ones and o he s, a he
han by c edibili y e alua ion i sel . Table7 p esen s he means
Managemen Le e s / Cuade nos de Ges ión 23/1 (2023) 75-86
The in luence o Ins ag amme s’ ecommenda ions on heal hy ood pu chase in en ion: The ole o consume in ol emen 83
and di e ence signi icances o ac o s, amongs consume s by
in ol emen wi h heal hy ood. Fo easy in e p e a ion, he di -
e ences ound o be non-signi ican a e in i alics.
The esul s con i med ha he Ins ag amme s’ c edibili-
y d i es he sel -b and cong ui y and he pu chase in en ion,
hus con i ming hypo heses H1 and H2. And he hypo hesis
ha sel -b and cong ui y should imp o e consume s’ in en ion
o buy heal hy ood (H3) was also con i med. The hypo heses
ha in ol emen wi h heal hy ood - IWHF, should mode a e
he ela ionships be ween Ins ag amme s’ c edibili y, he sel -
b and cong ui y, and he pu chase in en ion, was con i med
only in wo —among h ee— es ed ela ionships. The IWHF
does mode a e he in luence o sel -b and cong ui y (H5) and
Ins ag amme s’ c edibili y (H7), on he consume s’ in en ion o
buy heal hy ood, bu he in luence o Ins ag amme s’ c edibili y
in sel -b and cong ui y (H6) was no con i med. And he hy-
po hesis ha sel -b and cong ui y media es pa ially he in lu-
ence o Ins ag amme s’ c edibili y on pu chase in en ion (H4),
was also con i med.
As impo an as analyzing he a ows connec ing ac o s, is
o analyze ac o s’ means ac oss di e en g oups, once segmen s
usually di e in a i udes, choices, and in he way hey in e p e
he ecommenda ions. To e alua e he opinion p e alen in each
segmen ac oss he cons uc s, he ac o s we e ans o med in o
mani es a iables by he a e ages o hei indica o s and e alu-
a ed by a - es (assuming equi alen a iances).
The esul iden i ied signi ican di e ences be ween he wo
le els o in ol emen , in he sel -b and cong ui y and he pu -
chase in en ion (Table7).
Table7
Fac o s’ means and di e ences be ween g oups
o lowandhigh in ol emen
Le el o IWHF N Mean S anda d
De ia ion DF Sig.
Ins ag amme
A ac i eness
Low
IWHF 79 2.49 1.06
0.82 219 0.414
High
IWHF 142 2.38 0.98
Ins ag amme
T us wo hiness
Low
IWHF 79 2.08 1.03
1.84 219 0.067
High
IWHF 142 1.84 0.92
Ins ag amme
Expe ise
Low
IWHF 79 2.15 1.08
1.55 219 0.122
High
IWHF 142 1.93 1.01
Pu chase
In en ion
Low
IWHF 79 2.62 1.08
–5.77 219 0.000
High
IWHF 142 3.27 1.03
Sel -B and
Cong ui y
Low
IWHF 79 2.33 0.90 –4.46 219 0.000
High
IWHF
142 3.06 0.97
Sou ce: Own elabo a ion.
T us wo hiness and expe ience a e he ac o s ha bes e-
lec he Ins ag amme s’ c edibili y o ecommend heal hy ood
(Figu e3). And sel -b and cong uence is he mos impo an
d i e o pu chase in en ion in he model, gi en ha i s e ec s
a e signi ican in bo h segmen s, while he Ins ag amme s’ c edi-
bili y in luences ha beha io only among he high-in ol emen
consume s.
The posi i e e ec o digi al in luence c edibili y on sel -
b and cong ui y sugges ed by ex an li e a u e (e.g., Dwi edi
e al.,2015; Li & Peng, 2021) was con i med in he pa icula
con ex o Ins ag amme s’ endo semen o heal hy ood p od-
uc s and b ands. The same applies o he expec ed impac s o
bo h sou ce c edibili y (e.g., AlFa aj e al.,2021; Dja a o a &
Rushwo h, 2017; Masuda e al., 2022; Pick, 2021; Schou en
e al.,2020; Sokolo a & Ke i, 2020; Weismuelle e al.,2020; Zhu
e al.,2022) and sel -b and cong ui y (e.g., Agui e-Rod iguez
e al.,2012; Michel e al.,2022; Phua & Kim, 2018; Si gy, 1982).
Hence, his s udy p o ides u he empi ical suppo o ex an
li e a u e, pa icula ly in he con ex o heal hy ood pu chase
in en ion. Fu he mo e, his a icle p o ides wo impo an con-
ibu ions ha should be s essed. The model p oposed by his
s udy includes a media ion e ec ha was shown o be de e mi-
nan in unde s anding he impac o Ins ag amme s on heal hy
ood pu chase in en ions. Despi e being a esul o he combina-
ion o well-es ablished hypo heses and cohe en wi h he heo-
e ical assump ions o a media o e ec (Ba on & Kenny, 1986),
o he bes o he au ho s’ knowledge, his e ec has no been
es ed in p e ious s udies. As demons a ed in he p e ious pag-
es, his e ec was impo an o p o ide a wide unde s anding
o he impac s o sou ce c edibili y on pu chase in en ion, espe-
cially when segmen ing he consume s acco ding o hei le el o
in ol emen wi h heal hy ood. Aligned wi h he p e alence o
ma ke segmen a ion p ac ices by manage s, his s udy demon-
s a es ha compa ing ele an g oups o consume s may un-
co e in e dependencies be ween a iables and, ul ima ely, help
iden i y mo e e ec i e ways o in luence consume beha io .
5. CONCLUSION
Following he g owing s eam o esea ch on he impac o
Ins ag amme s on consume beha io (Dwi edi e al., 2021;
V on is e al.,2021), his a icle ocuses on heal hy ood pu -
chase in en ions, by shedding ligh on how consume in ol e-
men wi h heal hy ood a ec s he impac o Ins ag amme s on
hei ollowe s’ pu chase in en ions and cong ui y ega ding he
ad e ised b ands. The consequen heo e ical and manage ial
implica ions a e p esen ed below.
5.1. Theo e ical implica ions
F om he heo e ical s andpoin , he s udy ad ances knowl-
edge by demons a ing ha he impac o Ins ag amme s’ c edibil-
i y on in en ion o pu chase he ad e ised p oduc s a ec s di e -
en ly he segmen s o consume s wi h high and low in ol emen
wi h he consump ion o he p oduc s unde s udy, in his case
heal hy ood. Hence, al hough his s udy con ibu es o he li e a-
u e s essing he impo ance o in luence s’ c edibili y o explain
Managemen Le e s / Cuade nos de Ges ión 23/1 (2023) 75-86
