This is a sel -a chi ed e sion o an o iginal a icle. This e sion
may di e om he o iginal in pagina ion and ypog aphic de ails.
Au ho (s):
Ti le:
Yea :
Ve sion:
Copy igh :
Righ s:
Righ s u l:
Please ci e he o iginal e sion:
CC BY 4.0
h ps://c ea i ecommons.o g/licenses/by/4.0/
In e naliza ion o nega i e socie al iews on old age in o sel -pe cep ions o aging :
explo ing ac o s associa ed wi h sel -di ec ed ageism
© 2023 Ishikawa.
Published e sion
Ishikawa, Mo oko
Ishikawa, M. (2023). In e naliza ion o nega i e socie al iews on old age in o sel -pe cep ions o
aging : explo ing ac o s associa ed wi h sel -di ec ed ageism. F on ie s in Sociology, 8, A icle
1291325. h ps://doi.o g/10.3389/ soc.2023.1291325
2023
F on ie s in Sociology 01 on ie sin.o g
In e naliza ion o nega i e socie al
iews on old age in o
sel -pe cep ions o aging:
explo ing ac o s associa ed wi h
sel -di ec ed ageism
Mo okoIshikawa
1,2*
1 Facul y o Social Sciences, Uni e si y o Helsinki, Helsinki, Finland, 2 Depa men o Social Sciences and
Philosophy, Uni e si y o Jy äskylä, Jy äskylä, Finland
In oduc ion: A g owing numbe o esea ch has p o ided e idence o he nega i e
impac o ageism on olde people’s heal h and well-being. Among he h ee di e en
mani es a ions o ageism, namely ins i u ional, in e pe sonal and sel -di ec ed ageism,
signi ican ageism-heal h associa ions ha e been p o ed o bes onges o sel -
di ec ed ageism. This suppo s s e eo ype embodimen heo y, which main ains ha
li e ime exposu e o nega i e age s e eo ypes leads o he in e naliza ion o ageism
as a o m o nega i e a i udes owa ds own aging and i ad e sely a ec s heal h and
well-being in old age. Howe e , li le is known abou how people in e nalize nega i e
age s e eo ypes held in he socie y in o sel -pe cep ions o aging.
Me hods: This s udy aimed o explo e how socially sha ed belie s abou old age a e
in e nalized in o sel -pe cep ions o aging ocusing on unco e ing ac o s ela ed o
sel -di ec ed ageism. Da a we e de i ed om he su ey ha had examined ci izen’s
a i udes owa ds old age and aging in Finland. Mul inominal logis ic eg ession
models we e pe o med o examine he associa ion o sociodemog aphic and
con ex ual ac o s wi h di e en combina ions o socie al age s e eo ypes and wo
indica o s o sel -pe cep ions o aging: subjec i e iews on old age and pe sonal
eelings o own old age.
Resul s: The analyses showed ha being emale, a aining e ia y educa ion,
e alua ing poo quali y o li e and awa eness o ins i u ional old age disc imina ion
we e ela ed o holding nega i e iews on aging owa ds bo h socie y and onesel .
Discussion: The indings om uni a ia e and mul i a ia e models sugges ha i is
no age pe se, bu s uc u al and cul u al ci cums ances shaped wi h g owing olde
ha u ns socially sha ed nega i e age s e eo ypes in o nega i e sel -pe cep ions o
aging. E en hough he s udy add essed si ua ions in one coun y, he indings ha e
an impo an implica ion o o he apidly aging socie ies ega ding how social and
cul u al con ex s a e closely linked o he o ma ion o sel -di ec ed ageism.
KEYWORDS
ageism, sel -di ec ed ageism, nega i e age s e eo ypes, in e naliza ion, sel -pe cep ions
o aging
1 In oduc ion
Ageism is a mul i ace ed social phenomenon ha is de ined as he s e eo ypes, p ejudice
and disc imina ion di ec ed owa ds o he s o onesel based on age (Wo ld Heal h O ganiza ion,
2021). The h ee dimensions o ageism signi ies ha s e eo ypes a ec how we hink, p ejudice
a ec s how we eel, and disc imina ion a ec s how weac owa ds people on he basis o hei
OPEN ACCESS
EDITED BY
Todd L. Ma hews,
Sac ed Hea Uni e si y, Uni edS a es
REVIEWED BY
Anna Wanka,
Goe he Uni e si y F ank u , Ge many
Gio gio Picci o,
Bocconi Uni e si y, I aly
*CORRESPONDENCE
Mo oko Ishikawa
[email p o ec ed]
RECEIVED 09 Sep embe 2023
ACCEPTED 06 No embe 2023
PUBLISHED 17 No embe 2023
CITATION
Ishikawa M (2023) In e naliza ion o nega i e
socie al iews on old age in o sel -pe cep ions
o aging: explo ing ac o s associa ed wi h
sel -di ec ed ageism.
F on . Sociol. 8:1291325.
doi: 10.3389/ soc.2023.1291325
COPYRIGHT
© 2023 Ishikawa. This is an open-access a icle
dis ibu ed unde he e ms o he C ea i e
Commons A ibu ion License (CC BY). The
use, dis ibu ion o ep oduc ion in o he
o ums is pe mi ed, p o ided he o iginal
au ho (s) and he copy igh owne (s) a e
c edi ed and ha he o iginal publica ion in his
jou nal is ci ed, in acco dance wi h accep ed
academic p ac ice. No use, dis ibu ion o
ep oduc ion is pe mi ed which does no
comply wi h hese e ms.
TYPE O iginal Resea ch
PUBLISHED 17 No embe 2023
DOI 10.3389/ soc.2023.1291325
Ishikawa 10.3389/ soc.2023.1291325
F on ie s in Sociology 02 on ie sin.o g
age (O ice e  al., 2020). Ageism can ope a e ei he explici ly
(consciously) o implici ly (unconsciously), and i mani es s i sel a
h ee di e en le els as ins i u ional, in e pe sonal and sel -di ec ed
ageism, being in e wined and mu ually ein o cing (Ma ques e al.,
2020; Wo ld Heal h O ganiza ion, 2021). Ageism is ega ded o
bep e alen , deeply ing ained and mo e socially accep ed han o he
o ms o bias, and olde people a e mos ulne able o ageism hough
i can a ec any age g oup (Ayalon e al., 2019; O ice e al., 2020;
Wo ld Heal h O ganiza ion, 2021).
O e he pas wo decades, a g owing body o esea ch has p o ided
e idence o he nega i e impac o ageism on olde people’s heal h and
well-being. A i s global sys ema ic e iew on he impac s o ageism on
heal h, which included bo h s uc u al- and indi idual-le el s udies om
45 coun ies, ound ha ageism led o signi ican ly wo se heal h
ou comes: 95.5% o he 422 s udies and 74.0% o he 1,159 associa ions
be ween ageism and heal h showed e idence o he ad e se e ec s o
ageism (Chang e al., 2020). Signi ican ageism-heal h associa ions we e
obse ed ac oss 11 heal h domains: exclusion om heal h esea ch,
de alued li es o olde pe sons, lack-o -wo k oppo uni ies, denied access
o heal hca e and ea men s, educed longe i y, poo quali y-o -li e and
well-being, isky heal h beha io s, poo social ela ionships, physical
illness, men al illness and cogni i e impai men . These domains ep esen
a b oad ange o heal h ou comes, which con o ms wi h Wo ld Heal h
O ganiza ion’s (2020) de ini ion o heal h as a s a e o comple e physical,
men al and social well-being and no me ely he absence o disease o
in i mi y. The sys ema ic e iew also ound ha he associa ion be ween
ageism and heal h ou comes was s onges o sel -di ec ed ageism, ha
is, when ageism was u ned agains onesel and ope a ionalized as sel -
pe cep ions o aging measu e. Fo ins ance, nega i e sel -pe cep ions o
aging signi ican ly educed longe i y (Zhao e al., 2017), inc eased isk o
ha m ul obacco and alcohol use (Villie s-Tu hill e  al., 2016), and
impeded eco e y om se e e disabili y (Le y e al., 2012). Mo eo e , a
ecen s udy e ealed ha he connec ion be ween COVID-19 heal h
wo ies and anxie y symp oms was s onge among olde adul s wi h
high le els o sel -di ec ed ageism (Be gman e al., 2020).
O he sys ema ic e iew and me a-analysis o longi udinal s udies
co obo a e he indings o ad e se heal h e ec s o ageism among
olde adul s. A sys ema ic e iew, which examined he longi udinal
consequences o sel -pe cep ions o aging in pa icipan s 50 yea s o
olde by syn hesizing 21 s udies ac oss i e wes e n coun ies, e ealed
ha ha ing mo e posi i e sel -pe cep ions o aging was consis en ly
associa ed wi h be e sel - a ed heal h and less obesi y, g ea e
longe i y, be e pe o mance o he ac i i y o daily li ing, less
dep ession and be e cogni i e unc ioning (Tully-Wilson e al., 2021).
Sel -pe cep ions o aging (SPA) e e o a pe son’s gene al e alua ion o
own aging p ocess and e lec in e nalized s e eo ypes o aging. SPA
and i s eliable scale used in nume ous longi udinal s udies― he
A i ude Towa ds Own Aging (ATOA) scale (Law on, 1975; Liang and
Bollen, 1983; Tully-Wilson e al., 2021) ―a e ega ded as a measu e o
app aise sel -di ec ed ageism. Thus, he esul s o his sys ema ic e iew
sugges ha main aining posi i e sel -pe cep ions o aging p o ides a
bu e agains he nega i e impac o ageism on a ious heal h ou comes
o olde people. Likewise, a me a-analysis o 19 longi udinal s udies
ound a small signi ican o e all e ec o subjec i e aging―
ope a ionalized ei he as subjec i e age o sel -pe cep ions o aging―
on heal h, heal h beha io s and su i al o e ime (Wes e ho e al.,
2014). The analyses, howe e , e ealed he e ogenei y, wi h s onge
e ec s on he mo e p oximal ou come o heal h compa ed o su i al,
o e a sho e ime pe iod, in younge pa icipan s in hei second hal
o li e, and in wel a e egimes wi h less s a e p o isions. Taken oge he ,
p e ious esea ch including he ecen sys ema ic e iews as well as
me a-analysis highligh s ha ageism is a global heal h issue, which
cons i u es an impo an , and hi he o neglec ed, social de e minan o
heal h (O ice e al., 2020; Wo ld Heal h O ganiza ion, 2021).
1.1 Theo y on ageism-heal h associa ions
The ample e idence o signi ican ageism-heal h associa ions
p o ides suppo o s e eo ype embodimen heo y (SET). D awing on
esea ch showing ha posi i e and nega i e age s e eo ypes held by olde
indi iduals can ha e bene icial and de imen al e ec s, espec i ely, on
a a ie y o cogni i e and physical ou comes, SET p oposes ha age
s e eo ypes a e embodied when hei assimila ion om he su ounding
cul u e leads o sel -de ini ions ha , in u n, in luence unc ioning and
heal h (Le y, 2009). The e a e ou componen s cons i u ing he heo y.
Fi s , age s e eo ypes (i.e., belie s abou olde people in gene al) ypically
begin o de elop in childhood and a e ein o ced and in e nalized ac oss
he li e span. Second, hese s e eo ypes can ope a e unconsciously. Thi d,
as one g ows olde , age s e eo ypes become inc easingly ele an o he
aging pe son when iden i ying onesel wi h old pe son. Fou h and he
inally, hese age s e eo ypes gained salience om sel - ele ance shape
sel -s e eo ypes o aging (Le y, 2003; Wu m e al., 2017) and hey exe
in luence on an a ay o heal h ou comes h ough mul iple pa hways:
psychologically, beha io ally, and physiologically. Acco dingly, SET
o e s a li espan pe spec i e on he in e naliza ion and impac o bo h
posi i e and nega i e age s e eo ypes on indi idual aging p ocess (Wu m
e al., 2017; Tully-Wilson e al., 2021), which occu s in a op-down way
om socie y o he indi idual as well as o e ime om childhood o old
age (Le y, 2009).
Signi ican associa ions be ween ageism and heal h as well as
longi udinal heal h consequences o sel -pe cep ions o aging iden i ied
in a o emen ioned s udies speci ically alida e he hi d and ou h
componen s o SET, namely he ad e se e ec o in e nalized nega i e
age s e eo ypes o sel -di ec ed ageism on heal h and well-being in old
age. In con as , much ewe s udies ha e ocused on he ways in which
age s e eo ypes, as socially sha ed belie s abou olde people and old
age, a e inco po a ed in o sel -s e eo ypes. A comp ehensi e sys ema ic
e iew ound a o al o 14 de e minan s o ageism agains olde people,
o which 13 associa ed obus ly wi h o he -di ec ed o ms o ageism,
and only one had an e ec on sel -di ec ed ageism (Ayalon and Tesch-
Röme , 2017; Ma ques e al., 2020). Sel -di ec ed ageism was mos ly
de e mined by olde adul s’ men al and physical heal h s a us. This
inding implies bi-di ec ional na u e o he associa ion be ween sel -
di ec ed ageism and heal h. Meanwhile, ano he s udy ound lowe
heal hy li e expec ancy and ewe p opo ion o olde people wi hin a
coun y as explana o y ac o s o an indi idual o a coun y being
ageis (O ice e al., 2020). Ne e heless, i is again ele an o ageism
on in e pe sonal le el bu no o sel -di ec ed ageism.
To sum i up, li le is known abou p ocesses be ween age
s e eo ypes and sel -s e eo ypes, especially in e ms o how nega i e
age s e eo ypes gene ally held in he socie y de elop sel -di ec ed
ageism ia sel - ele ance. To da e, no obus de e minan s o sel -
di ec ed ageism ha e been ound a he in e pe sonal/in e g oup and
ins i u ional/cul u al le el o analysis (Ma ques e  al., 2020). The
sca ce knowledge abou de e minan s o sel -di ec ed ageism p esen s
Ishikawa 10.3389/ soc.2023.1291325
F on ie s in Sociology 03 on ie sin.o g
a majo esea ch gap, while he heo y main ains ha indi iduals
assimila e age s e eo ypes om he su ounding cul u e (Le y, 2009),
and speci ically esea ch has inc easingly demons a ed de imen al
e ec o sel -di ec ed ageism on heal h and well-being. A sociological
app oach ha sheds ligh on socio-cul u al in luences on indi iduals
in shaping sel -di ec ed ageism is now called o .
1.2 The cu en s udy
To add ess he iden i ied knowledge gap, he cu en s udy aims o
examine unde wha condi ions, including sociodemog aphic
cha ac e is ics and socio-cul u al con ex s, people a e suscep ible o
in e nalizing socially pe cei ed belie s abou he p ocess o aging and
abou olde people. In o he wo ds, i in es iga es how socially sha ed
belie s abou old age a e linked o indi iduals’ sel -pe cep ions o aging.
The esea ch ques ions a e as ollows: (1) Does inc easing age cause
socially sha ed nega i e age s e eo ypes o shape nega i e sel -pe cep ions
o aging among olde adul s? (2) I olde age is no he majo p edic o o
sel -di ec ed ageism, wha o he ac o s a e associa ed wi h in e naliza ion
o nega i e socie al age s e eo ypes in o sel -pe cep ions o aging?
This s udy u ilizes da ase om he su ey in which pa icipan s
wi h a b oad age ange we e asked abou hei pe cep ions o age
s e eo ypes sha ed in he socie y and sel -pe cep ions o aging. Inqui ing
iews on aging on bo h socie al and indi idual le els has an ad an age
in unde aking he esea ch ques ions o he p esen s udy. Besides, wo
ques ions in oduced in he su ey o assess sel -pe cep ions o aging
cap u e mul i ace ed na u e o ageism: one is abou esponden s’
nega i e a i udes owa ds aging and he o he e e s o ea o aging.
Acco ding o SET, li e ime exposu e o nega i e age s e eo ypes leads o
he in e naliza ion o ageism as a o m o nega i e a i udes owa ds own
aging. Meanwhile, e o managemen heo y main ains ha h ea s o
dea h, physical de e io a ion, and insigni icance e oked by he
encoun e wi h olde adul s mani es in ageism as a ea o aging di ec ed
owa ds o he s as well as sel es (Ma ens e al., 2005; Le e al., 2018).
Thus, he wo ques ions a e bo h ele an o he mani es a ion o ageism,
bu i p esumably eme ges h ough di e en pa hways. This sugges s
ha hese ques ions should beanalyzed sepa a ely, since de e minan s
o sel -di ec ed ageism may di e be ween he wo.
Answe ing o he esea ch ques ions by analyzing his su ey da a
ills a gap ha has been iden i ied in bo h he heo y (SET) and
empi ical s udies. Gi en he accumula ed knowledge on ad e se heal h
e ec s o ageism, he mos no able is ha esul s o his s udy abou
ac o s associa ed wi h sel -di ec ed ageism will highligh po en ial
ulne able popula ion g oups and li e si ua ions o which social- and
heal h policy as well as in e en ions should pay mo e ca e ul a en ion.
2 Ma e ials and me hods
2.1 Da a
The su ey examining ci izens’ a i udes owa ds old age, aging and
olde popula ion in Finland was he da a sou ce o his s udy. VTKL
– The Finnish Associa ion o he Wel a e o Olde Adul s commissioned
he su ey o Aula Resea ch Oy. The a ge s o he su ey we e Finnish
ci izens aged 16 yea s and olde and sampling was conduc ed o
co espond o he dis ibu ion o a ge popula ion acco ding o
esiden ial egion, age, and gende . The su ey was conduc ed be ween
Janua y and Ma ch 2022 h ough online ques ionnai e o all age g oups
and by elephone in e iews o hose who we e 85+ yea s old (Vaa ama,
2022). The esponse a e o he online su ey was 19%, whe eas 46% o
people con ac ed by elephone esponded o he ques ions.
The su ey ecei ed 2,056 esponden s. Howe e , we emo ed 30
cases wi hou in o ma ion on gende o bi h yea and 29 cases who
did no espond o he ques ion on socie al iews on old age, 10 cases
who did no answe he ques ion on subjec i e iews on old age, and
3 cases wi hou esponse o he ques ion on pe sonal eelings o own
old age. Addi ionally, we emo ed one case whose bi h yea was 1900
and 2 cases whose bi h yea was 2007 o la e . Thus, wehad 1,981
esponden s in ou inal s udy coho .
Gi en ha ecen ise in popula ion aging seems o beassocia ed
wi h nega i e a i udes owa ds aging and olde people in a global
con ex (No h and Fiske, 2015), Finland is an in e es ing coun y o
in es iga ing how age s e eo ypes sha ed in he socie y and people’s
a i udes owa ds own aging a e ela ed. In Finland, as a coun y
belonging o he No dic wel a e egime, public au ho i ies ha e been
esponsible o uni e sal p o isions o income secu i y and ca e
se ices o olde people. Howe e , he coun y has been aging mos
apidly in Eu ope (Eu os a , 2023a), which poses a se ious challenge
o upkeep he public p o isions. Hence, he ecen demog aphic
change and i s possible nega i e consequences in old age social
policies may in luence bo h socie al and sel -pe cep ions o aging.
2.2 Ou come a iables
To examine he esea ch ques ions on how socially sha ed belie s
abou old age a e linked o indi iduals’ sel -pe cep ions o aging, his
s udy c ea ed new ou come a iables by me ging alues o wo
ques ions. The ollowings explain he o iginal a iables on socie al age
s e eo ypes and sel -pe cep ions o aging and how hese we e
in eg a ed in o he new a iables.
2.2.1 Age s e eo ypes sha ed in he socie y
Age s e eo ypes we e assessed in he su ey as people’s pe cep ions
o socie al iews on old age by asking pa icipan s as ollows: ‘How
posi i ely o nega i ely do yousee old age is ega ded in Finnish
socie y in gene al?’ As he p esen s udy ocuses on in es iga ing
nega i e pe cep ions, esponse al e na i es wi h a six-le el Like scale
we e dicho omized as nega i ely (‘ e y nega i ely’ and ‘somewha
nega i ely’) and posi i ely o neu ally (‘ e y posi i ely’, ‘somewha
posi i ely’, ‘nei he posi i ely no nega i ely’ and ‘canno say’).
2.2.2 Sel -pe cep ions o aging
Sel -pe cep ions o aging we e measu ed in he su ey using wo
ques ions: one asked pa icipan s’ a i udes owa ds aging in a
ela i ely gene al manne and he o he explo ed mo e explici ly
pe sonal eelings o own old age. The i s ques ion, ‘How posi i ely o
nega i ely do yousee old age?’ was in ended o assessing subjec i e
iews on old age and his was inqui ed be o e he p e iously
men ioned ques ion on socie al iews on old age. The same six
esponse al e na i es as socie al age s e eo ypes ( om ‘ e y posi i ely’
o ‘ e y nega i ely’) we e ecoded simila ly in a dicho omous manne .
The second ques ion ocused mo e on pa icipan s’ pe sonal eelings
ega ding ea o aging: ‘When you hink o you own old age, wha
Ishikawa 10.3389/ soc.2023.1291325
F on ie s in Sociology 04 on ie sin.o g
does i eel like?’ To examine nega i e pe cep ions, i e-le el Like
scale o esponse was dicho omized as eel ea (‘ e y ea ul’ and
‘somewha ea ul’) and eel sa e o neu al (‘ e y sa e’, ‘somewha sa e’
and ‘nei he sa e no ea ul’).
2.2.3 C ea ing wo ou come a iables:
combina ions o socie al age s e eo ypes and
sel -pe cep ions o aging
Nex , he wo ou come a iables we e c ea ed o he p esen s udy
by me ging p e iously desc ibed dicho omous a iables. The a iable on
socie al age s e eo ypes was me ged wi h each a iable on sel -
pe cep ions o aging: one wi h subjec i e iews on aging (a i udes
owa ds aging) and ano he wi h pe sonal eelings o own old age ( ea
o own old age). The combined new a iables had ou ca ego ies: (1)
hose who ake a posi i e/neu al s ance owa ds bo h socie al iews on
old age and sel -pe cep ions o aging, (2) hose who ha e nega i e socie al
iews on old age, while sel -pe cep ions o aging a e posi i e/neu al, (3)
hose who ha e posi i e/neu al socie al iews on old age, while sel -
pe cep ions o aging a e nega i e and (4) hose who ake a nega i e s ance
owa ds bo h socie al iews on old age and sel -pe cep ions o aging.
Responden s loca ed in he las ca ego y e e o hose who in e nalize
nega i e age s e eo ypes in he socie y in o sel -pe cep ions.
2.3 Explana o y a iables
Explana o y a iables included he ollowing demog aphic and
socioeconomic cha ac e is ics: gende , age in en-yea g oups, li ing
a angemen s (‘li ing wi h a pa ne ’ o ‘no li ing wi h a pa ne ’),
highes a ainmen in educa ion (‘basic educa ion’, ‘seconda y educa ion’
o ‘ e ia y educa ion’) and income le el (‘enough money o daily needs’
o ‘no enough money o eluc ance o answe ’). The cumula i e
ad an age –disad an age pe spec i e has heo e ically and empi ically
explained he in e indi idual di e gence in a gi en cha ac e is ic (e.g.,
money, heal h, o s a us) wi h he passage o ime, which ep oduces a
g ea e he e ogenei y and inequali y in la e li e (Danne e , 2003). Thus,
hese explana o y a iables we e expec ed o be ela ed o sel -di ec ed
ageism shaped o e he cou se o one’s li e. Besides, gende is especially
impo an in s udies on ageism, as he so-called ‘double s anda d o
aging’, e e ing o disc imina ion agains olde women bo h because o
hei gende and hei age, has gene a ed di e en meanings o aging o
women and men (Son ag, 1978; A be and Ginn, 1995).
Sel - a ed quali y o li e (‘ e y good’, ‘good’, ‘nei he good no bad’,
‘bad’ o ‘ e y bad’) was also included as an explana o y a iable. Gi en
he socio-cul u al in luence on shaping age s e eo ypes (Le y, 2009),
pa icipan s’ opinions abou whe he ollowing ins i u ions a e
disc imina ing agains olde people we e inco po a ed in he analysis:
poli icians, media, heal h ca e se ices (e.g., doc o ’s appoin men ),
social se ices (e.g., home ca e, se ice housing, meal se ice), he
Social Insu ance Ins i u ion o Finland, p i a e companies (e.g., shops,
pha macies, es au an s), anspo and employe s.
2.4 Analysis
Mul inomial logis ic eg ession models we e pe o med o
examine he associa ion o pa icipan s’ sociodemog aphic
backg ounds, quali y o li e and opinions on old age disc imina ion
wi h he wo ou come a iables (Kleinbaum and Klein, 2002). Fi s ,
sepa a e logis ic models we e i ed o each explana o y a iable o
analyze hei unadjus ed associa ions wi h he ou come a iables
(Model 1). Since all explana o y a iables we e ca ego ical wi h k
classes, hey we e ans o med as k-1 dummy a iables. Fo example,
o educa ion he equa ion is de ined as:
logi x
xx
seconda y e ia y
πβ
ββ
()
()
=+ +
01 2 (1)
whe e logi (π(x)) is he loga i hmic ans o ma ion o he
expec ed alue o bina y ou come a iable, β
0
is he in e cep o linea
p edic o and β
1
and β
2
a e he eg ession coe icien s o he seconda y
and e ia y educa ion dummy a iables, espec i ely. The eg ession
coe icien s we e ans o med as ela i e isks (RR) hus desc ibing he
p obabili y o ha ing he alue 1 o ou come a iable o seconda y
and e ia y educa ion as compa ed o basic educa ion used as he
e e ence ca ego y. Fu he , 95% con idence in e als (CI) o RRs
we e calcula ed using s anda d e o s.
A e ha a ull logis ic eg ession model was i ed including all
explana o y a iables simul aneously (Model 2) o s udy how he
mu ual adjus men o o he explana o y a iables a ec s hese
associa ions. In he case o p explana o y a iables, he equa ion is:
logi x xx
xx
pp
πβββ
ββ
()
()
=+ +++…
0112233 (2)
Mul inomial logis ic eg ession model can beex ended o ha e
an ou come a iable wi h k ca ego ies es ima ing k-1 se s o es ima es.
Since he ou come a iables had 4 classes, h ee se s o eg ession
es ima es we e es ima ed. In his case, RRs desc ibe he p obabili y o
belong in a ce ain ca ego y o ou come a iable as compa ed o he
e e ence ca ego ies o bo h he ou come and explana o y a iables.
The models we e es ima ed using S a a/SE 17.0 o Windows s a is ical
so wa e (S a aCo p, College S a ion, TX, Uni edS a es).
3 Resul s
3.1 Sample cha ac e is ics and sha e o
ha ing nega i e socie al iews on old age
and ha ing nega i e sel -pe cep ions o
aging
Table1 displays cha ac e is ics o he o al sample acco ding o
ca ego ized g oups and sha e o ha ing nega i e socie al iews on old
age and ha ing nega i e sel -pe cep ions o aging in espec i e g oups
o he sample. These a e backg ound in o ma ion o he mul inominal
logis ic eg ession models explained la e . The sha e o hose who had
nega i e subjec i e iews on old age and hose who el ea abou own
old age was smalle han he sha e o hose who nega i ely pe cei ed
socie al iews on old age in almos all g oups. This suppo s he
a gumen ha people endo se ce ain bu no all socio-cul u al age
s e eo ypes (Wu m e al., 2017). I is also no ewo hy ha while olde
pa icipan s, excep he oldes old, had mo e nega i e pe cep ions on
how old age was ega ded in he socie y, hose who had nega i e
subjec i e iews on aging and we e ea ul abou own old age ended
o dec ease as age o he esponden s inc eased.

Ishikawa 10.3389/ soc.2023.1291325
F on ie s in Sociology 05 on ie sin.o g
3.2 Associa ion o di e en ac o s wi h
combina ions o socie al age s e eo ypes
and subjec i e iews on old age
Table 2 p esen s he esul s o mul inominal logis ic eg ession
analyses ha in es iga ed he associa ion o explana o y a iables wi h
combina ions o socie al age s e eo ypes and subjec i e iews on old age.
Re e ence ca ego y was hose who ook a posi i e/neu al s ance owa ds
bo h socie al iews on old age and subjec i e iews on old age. The esul s
show he likelihood o being placed in o he h ee ca ego ies (nega i e
socie al iews–nega i e subjec i e iews, nega i e socie al iews–
posi i e/neu al subjec i e iews, posi i e/neu al socie al iews–nega i e
subjec i e iews) in espec o esponden s’ cha ac e is ics compa ed
wi h he e e ence ca ego y. Model 1 e e s o uni a ia e analysis in which
each a iable was included in o he model sepa a ely, whe eas
mul i a ia e analysis con olled o all a iables is shown in Model 2.
The p ima y ocus o his s udy was o examine who had nega i e
pe cep ions o socie al iews on old age and hold nega i e subjec i e
iews on old age as well. Mul i a ia e analysis con olled o all
a iables e ealed ha esponden s who we e emale, had he highes
educa ional le el, a ed quali y o li e as wo se, and saw poli icians,
social se ices, p i a e companies and employe s as pe pe a o s o
ageism we e mo e likely o hink ha old age was ega ded nega i ely
in Finland and we e hemsel es liable o iew aging in a nega i e ligh .
I can bea gued ha people wi h such cha ac e is ics a e suscep ible
o in e nalizing nega i e socie al age s e eo ypes in o hei own
pe cep ions abou aging. Howe e , as he able p esen s, o he
ca ego ies o he ou come a iable also had many p edic o s wi h
s a is ical signi icance. Hence, he ollowing subsec ions desc ibe
indings o highligh he ole o each ac o in he associa ions.
3.2.1 Gende
Being emale inc eased he isk o conside ing bo h socie al and
subjec i e iews on old age as nega i e, as well as pe cei ing nega i e
socie al a i udes owa ds aging while holding posi i e/neu al
subjec i e iews on old age. Mul i a ia e analysis con olled o all
a iables e en s eng hened he gende e ec o hese ca ego ies.
Howe e , gende was no ela ed o he likelihood o iewing old age
nega i ely on a pe sonal le el despi e ha ing posi i e/neu al socie al
iews on aging. The indings indica e ha nega i e a i udes among
emales we e di ec ed mainly o socie al belie s abou aging a he
han o hei own iews on old age.
3.2.2 Age
Responden s aged be ween 56 and 75 yea s compa ed wi h he
younges age g oup we e mo e likely o pe cei e aging nega i ely on
bo h socie al and indi idual le els. Age- ela ed changes in one’s li e
cou se such as e i emen , app oaching old age, and ca e needs o
close ela i es and/o onesel migh make esponden s hink in his
manne . Howe e , adjus ing o o he a iables a enua ed associa ion
e ec o age. Olde age also inc eased he isk o ha ing nega i e
socie al iews on aging while subjec i ely iewing old age in a posi i e/
neu al ligh . This applied o he esponden s aged 46 yea s and olde
excep he oldes old. Mul i a ia e analysis con olled o all a iables
kep he signi ican associa ion o age be ween 66 and 85. Con e sely,
bo h uni a ia e and mul i a ia e analyses e ealed ha olde
esponden s we e less likely o nega i ely iew old age on a pe sonal
le el despi e holding posi i e/neu al socie al iews on old age. Age
e ec s on all ca ego ies o he ou come a iable sugges ha olde
people we e inclined o shape nega i e pe cep ions o aging owa ds
socie y and much unlikely owa ds hemsel es.
3.2.3 Li ing a angemen s
Whe he he esponden s li ed wi h a pa ne o no was gene ally
no ela ed o di e en combina ions o socie al age s e eo ypes and
subjec i e iews on old age.
3.2.4 Educa ion
Compa ed o hose who had only basic educa ion, esponden s wi h
e ia y educa ion we e mo e likely o pe cei e old age as nega i e on
bo h socie al and pe sonal le els, as well as o ha e nega i e socie al iews
on old age while subjec i e iews we e posi i e o neu al. Adjus ing o
o he a iables e en inc eased he associa ion e ec o he highes
educa ion o hese ca ego ies. Howe e , educa ional le el was no ela ed
o he ca ego y o ha ing posi i e/neu al socie al iews on old age and
simul aneously holding nega i e subjec i e iews on old age. The esul s
indica e ha nega i e a i udes owa ds aging among people wi h e ia y
educa ion we e di ec ed mainly o he socie y a he han o hemsel es.
3.2.5 Income le el
When income le el was analyzed sepa a ely om o he a iables,
li ing unde insu icien economic si ua ion inc eased he isk o
belonging o he wo ca ego ies: nega i e socie al iews on old age–
nega i e subjec i e iews on old age, and nega i e socie al iews on old
age–posi i e/neu al subjec i e iews on old age. Howe e , adjus ing
o o he a iables a enua ed associa ion e ec o income le el.
3.2.6 Quali y o li e
Among di e en explana o y a iables, poo quali y o li e appea ed
o bemos obus ly associa ed wi h esponden s’ nega i e a i udes
owa ds aging on bo h socie al and indi idual le els. The mul i a ia e
analysis e ealed ha he isk o ha ing such way o hinking was 15
imes highe o hose who e alua ed quali y o li e as bad and 21 ime
highe o hose wi h e y bad quali y o li e, compa ed wi h people
wi h e y good li e quali y. Poo e sel - a ed quali y o li e was
signi ican ly ela ed o he o he wo ca ego ies oo, bu he isk was
highe o nega i ely pe cei ing subjec i e iews on old age while
ha ing posi i e/neu al socie al iew on aging. This indica es ha hose
who a ed wo se quali y o li e we e mo e likely o shape nega i e
pe cep ions o aging owa ds hemsel es a he han owa ds he socie y.
3.2.7 Ins i u ions in ol ed in old age
disc imina ion
The uni a ia e analysis showed ha seeing each social ins i u ion
as disc imina ing agains olde people inc eased he isk o nega i ely
pe cei ing aging on bo h socie al and indi idual le els, as well as
holding nega i e socie al iews on aging while subjec i e iews we e
posi i e/neu al. Though adjus ing o o he a iables a enua ed
mos o he associa ion e ec s, ega ding poli icians, social se ices,
p i a e companies and employe s as pe pe a o s o ageism emained
o besigni ican ly associa ed wi h ha ing nega i e pe cep ions o
bo h socie al and subjec i e iews on old age. Meanwhile, opinions
on old age disc imina ion we e no ela ed o he likelihood o
iewing old age nega i ely on a pe sonal le el despi e ha ing posi i e/
neu al socie al iews on aging. These indings sugges ha whe he
each ins i u ion was disc imina ing agains olde people was mo e
Ishikawa 10.3389/ soc.2023.1291325
F on ie s in Sociology 06 on ie sin.o g
TABLE1 Sample cha ac e is ics o he o al sample and sha e o ha ing nega i e socie al iews on old age and ha ing nega i e sel -pe cep ions o
aging in espec i e g oups.
Dis ibu ion in he
sample (%), N=  1,981
Sha e o ha ing
nega i e socie al
iews on old age (%)
Sha e o ha ing
nega i e subjec i e
iews on old age (%)
Sha e o eeling
ea abou own old
age (%)
Gende
Male 47.7 36.7 14.2 22.1
Female 52.3 50.8 15.6 32.9
Age in en-yea g oups
16–25 11.3 30.0 16.1 30.0
26–35 12.4 37.4 17.5 37.8
36–45 13.2 37.0 15.3 36.6
46–55 14.8 42.7 13.7 33.8
56–65 18.9 51.5 18.1 29.1
66–75 17.0 54.2 11.6 18.2
76–85 8.9 54.8 11.3 10.2
86 + 3.5 29.0 14.5 10.1
Li ing a angemen s
Li ing wi h a pa ne 46.9 43.7 12.6 24.9
No li ing wi h a pa ne 53.1 44.4 17.0 30.3
Educa ion
Basic educa ion 19.9 37.2 13.7 26.6
Seconda y educa ion 54.6 43.3 14.7 28.9
Te ia y educa ion 25.4 51.2 16.5 26.2
Income le el
Su icien 48.1 40.4 12.9 18.4
Insu icien / eluc ance o
answe
51.9 47.4 16.8 36.4
Quali y o li e
Ve y good 8.7 28.9 5.8 6.9
Good 46.0 40.4 11.5 21.2
Nei he good no bad 32.2 46.9 15.7 31.4
Bad 11.1 59.4 30.1 53.4
Ve y bad 2.1 63.4 36.6 68.3
Ins i u ions in ol ed in old age disc imina ion
Poli icians
No 71.4 34.6 13.6 23.0
Yes 28.6 67.7 18.2 39.8
Media
No 86.9 40.2 14.8 26.1
Yes 13.1 69.9 16.2 38.6
Heal h ca e se ices
No 76.7 37.4 13.6 24.9
Yes 23.3 66.2 19.5 37.1
Social se ices
No 74.1 36.9 13.2 23.1
Yes 25.9 64.5 19.9 41.1
(Con inued)
Ishikawa 10.3389/ soc.2023.1291325
F on ie s in Sociology 07 on ie sin.o g
ele an wi h esponden s’ pe cep ions o socie al iews on old age
han hei sel -pe cep ions o aging.
3.3 Associa ion o di e en ac o s wi h
combina ions o socie al age s e eo ypes
and pe sonal eelings o own old age
Table3 shows he associa ion o esponden s’ sociodemog aphic
and con ex ual ac o s wi h combina ions o socie al age s e eo ypes
and pe sonal eelings o own old age. Re e ence ca ego y was hose
who ook a posi i e/neu al s ance owa ds socie al iews on old age
and el sa e o neu al abou hei own old age. The likelihood o being
placed in o he h ee ca ego ies (nega i e socie al iews– eeling ea
abou own old age, nega i e socie al iews– eeling sa e/neu al abou
own old age, posi i e/neu al socie al iews– eeling ea abou own old
age,) in compa ison o he e e ence ca ego y was speci ied in he able.
Mul i a ia e analysis con olled o all a iables e ealed ha
esponden s who we e emale, had he highes educa ional le el, a ed
quali y o li e as wo se, and conside ed poli icians, social se ices and
employe s disc imina ing agains olde people we e mo e likely o
hink ha old age was ega ded nega i ely in he socie y and we e
ea ul abou hei own old age. These ac o s appea ed o beassocia ed
wi h in e naliza ion o nega i e socie al age s e eo ypes in o sel -
pe cep ions o aging. Howe e , since o he ca ego ies o he ou come
a iable also had many signi ican p edic o s, indings a e desc ibed
acco ding o he explana o y a iables as ollows.
3.3.1 Gende
Being emale inc eased he isk o being placed in all he h ee
ca ego ies o he ou come a iables. Mul i a ia e analysis con olled
o all a iables e en s eng hened he gende e ec o all he
ca ego ies. The esul s sugges ha compa ed o males, emales we e
mo e likely o hink ha socially sha ed belie s abou aging we e
nega i e and hey we e also inclined o eel ea abou own old age.
3.3.2 Age
Responden s aged be ween 26 and 75 yea s compa ed wi h he
younges age g oup we e mo e likely o pe cei e aging nega i ely on
socie al le el as well as o eel ea abou own old age. Howe e ,
adjus ing o o he a iables a enua ed associa ion e ec o age. In
e ms o ha ing nega i e socie al iews on old age while eeling sa e/
neu al abou own old age, olde age inc eased he isk o being placed
in his ca ego y. In con as , bo h uni a ia e and mul i a ia e analyses
e ealed ha olde esponden s we e ob iously less likely o eel ea
abou own old age while ha ing posi i e/neu al socie al iews on old
age. The esul s o age e ec indica e ha olde age was ele an o
shaping nega i e pe cep ions o socie al age s e eo ypes and was much
unlikely o be ela ed o ea o aging.
3.3.3 Li ing a angemen s
Whe he he esponden s li ed wi h a pa ne o no was gene ally
no ela ed o di e en combina ions o socie al age s e eo ypes and
pe sonal eelings o own old age.
3.3.4 Educa ion
A aining e ia y educa ion was signi ican ly associa ed wi h
being placed in he wo ca ego ies: nega i e socie al iews on old age–
eeling ea abou own old age, and nega i e socie al iews on old age–
eeling sa e/neu al abou own old age. Adjus ing o o he a iables
e en inc eased he e ec o e ia y educa ion. Howe e , educa ional
le el was no ela ed o he ca ego y o ha ing posi i e/neu al socie al
iews on old age and simul aneously eeling ea abou own old age.
This indica es ha nega i e a i udes among people wi h e ia y
educa ion we e di ec ed mainly o socie al belie s abou aging a he
han o pe sonal eelings o own old age.
3.3.5 Income le el
Uni a ia e analysis e ealed ha li ing unde insu icien
economic si ua ion inc eased he isk o belonging o he wo
ca ego ies: ha ing nega i e socie al iews on old age– eeling ea
abou own old age, and ha ing posi i e/neu al socie al iews on old
age– eeling ea abou own old age. This may indica e ha
disad an age in economic si ua ion was subjec o ea o own old age
a he han nega i e pe cep ions o socie al iews on aging. Howe e ,
adjus ing o o he a iables a enua ed associa ion e ec o
income le el.
Dis ibu ion in he
sample (%), N=  1,981
Sha e o ha ing
nega i e socie al
iews on old age (%)
Sha e o ha ing
nega i e subjec i e
iews on old age (%)
Sha e o eeling
ea abou own old
age (%)
The Social Insu ance Ins i u ion o Finland
No 81.3 39.5 13.7 23.7
Yes 18.7 63.9 20.5 45.3
P i a e companies
No 92.6 42.0 14.0 26.7
Yes 7.4 70.1 27.2 40.8
T anspo
No 90.3 41.4 14.3 26.0
Yes 9.7 69.3 21.4 44.3
Employe s
No 77.7 37.0 12.7 23.6
Yes 22.3 68.8 22.6 42.3
TABLE1 (Con inued)
Ishikawa 10.3389/ soc.2023.1291325
F on ie s in Sociology 08 on ie sin.o g
TABLE2 Associa ion o indi idual cha ac e is ics and con ex ual ac o s wi h combina ions o socie al age s e eo ypes and subjec i e iews on old age.
Ca ego iesaNega i e socie al iews on
old age–Nega i e subjec i e
iews on old age
Nega i e socie al iews on
old age–Posi i e/neu al
subjec i e iews on old age
Posi i e/neu al socie al
iews on old age–Nega i e
subjec i e iews on old age
Model 1bModel 2cModel 1 Model 2 Model 1 Model 2
Va iables RRd (95% CI) RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI)
Gende
Male e . e . e . e . e . e .
Female 1.62 (1.20–2.19) 1.89 (1.36–2.63) 1.83 (1.50–2.24) 2.08 (1.66–2.60) 1.02 (0.65–1.59) 0.97 (0.61–1.55)
Age in en-yea g oups
16–25 e . e . e . e . e . e .
26–35 1.61 (0.83–3.14) 0.92 (0.45–1.88) 1.28 (0.83–1.99) 0.85 (0.53–1.36) 0.84 (0.42–1.68) 0.79 (0.38–1.67)
36–45 1.46 (0.75–2.84) 0.73 (0.35–1.50) 1.25 (0.82–1.92) 0.78 (0.49–1.25) 0.63 (0.31–1.30) 0.54 (0.25–1.18)
46–55 1.57 (0.82–3.01) 0.76 (0.37–1.53) 1.63 (1.08–2.46) 1.06 (0.67–1.65) 0.48 (0.22–1.03) 0.37 (0.16–0.84)
56–65 2.87 (1.58–5.21) 1.49 (0.78–2.86) 2.09 (1.41–3.10) 1.38 (0.90–2.13) 0.40 (0.18–0.87) 0.30 (0.13–0.70)
66–75 1.98 (1.05–3.74) 1.31 (0.66–2.60) 2.61 (1.76–3.88) 1.97 (1.28–3.04) 0.18 (0.06–0.54) 0.15 (0.05–0.47)
76–85 1.79 (0.85–3.78) 1.32 (0.59–2.92) 2.84 (1.81–4.45) 2.26 (1.39–3.68) 0.36 (0.12–1.09) 0.32 (0.10–1.02)
86 + 1.13 (0.42–3.07) 0.98 (0.34–2.81) 0.83 (0.42–1.64) 0.82 (0.40–1.70) 0.60 (0.20–1.86) 0.45 (0.14–1.47)
Li ing a angemen s
Li ing wi h a pa ne e . e . e . e . e . e .
No li ing wi h a pa ne 1.25 (0.93–1.69) 1.20 (0.86–1.68) 1.04 (0.85–1.26) 1.09 (0.87–1.36) 2.10 (1.30–3.39) 1.61 (0.97–2.68)
Educa ion
Basic educa ion e . e . e . e . e . e .
Seconda y educa ion 1.41 (0.93–2.16) 1.59 (1.00–2.51) 1.22 (0.94–1.59) 1.28 (0.96–1.71) 0.81 (0.48–1.38) 0.96 (0.54–1.68)
Te ia y educa ion 2.04 (1.28–3.23) 2.98 (1.77–5.00) 1.65 (1.23–2.21) 1.99 (1.42–2.80) 0.76 (0.39–1.47) 1.02 (0.49–2.10)
Income le el
Su icien e . e . e . e . e . e .
Insu icien / eluc ance o
answe
1.65 (1.22–2.23) 0.91 (0.62–1.33) 1.26 (1.04–1.53) 1.02 (0.79–1.31) 1.17 (0.75–1.82) 0.68 (0.40–1.16)
Quali y o li e
Ve y good e . e . e . e . e . e .
Good 2.83 (1.20–6.67) 3.16 (1.31–7.65) 1.56 (1.07–2.27) 1.52 (1.01–2.30) 1.86 (0.65–5.37) 2.09 (0.72–6.10)
Nei he good no bad 4.61 (1.95–10.88) 4.73 (1.90–11.77) 1.98 (1.35–2.91) 1.69 (1.09–2.64) 2.69 (0.92–7.82) 3.74 (1.23–11.43)
Bad 13.58 (5.55–33.26) 14.74 (5.59–38.85) 2.96 (1.85–4.74) 2.79 (1.63–4.79) 6.52 (2.10–20.26) 9.92 (2.97–33.12)
Ve y bad 19.83 (6.15–63.96) 21.14 (5.93–75.38) 3.69 (1.57–8.64) 3.22 (1.25–8.28) 10.82 (2.37–49.33) 16.26 (3.25–81.31)
Ins i u ions in ol ed in old age disc imina ion ( e .: no disc imina ing)
Poli icians 3.66 (2.66–5.04) 1.56 (1.02–2.38) 4.05 (3.24–5.08) 2.32 (1.73–3.12) 1.01 (0.56–1.84) 1.12 (0.52–2.41)
Media 2.80 (1.84–4.28) 0.81 (0.48–1.37) 3.54 (2.62–4.79) 1.29 (0.89–1.86) 0.64 (0.23–1.80) 0.58 (0.18–1.80)
Heal h ca e se ices 3.64 (2.62–5.06) 1.52 (0.96–2.40) 3.10 (2.45–3.94) 1.39 (1.00–1.93) 0.81 (0.41–1.61) 0.86 (0.36–2.05)
Social se ices 3.57 (2.59–4.92) 1.58 (1.01–2.47) 3.00 (2.39–3.78) 1.37 (1.00–1.87) 1.12 (0.62–2.00) 1.36 (0.62–2.98)
The Social Insu ance
Ins i u ion o Finland
3.32 (2.35–4.69) 0.99 (0.61–1.60) 2.49 (1.93–3.22) 0.88 (0.62–1.25) 0.86 (0.42–1.77) 0.67 (0.27–1.66)
P i a e companies 5.03 (3.12–8.12) 2.35 (1.31–4.23) 2.77 (1.85–4.16) 1.22 (0.75–1.99) 1.23 (0.43–3.52) 1.11 (0.34–3.62)
T anspo 3.72 (2.38–5.82) 0.94 (0.53–1.64) 2.99 (2.12–4.24) 1.03 (0.67–1.57) 0.88 (0.31–2.49) 0.77 (0.24–2.49)
Employe s 4.95 (3.55–6.91) 2.60 (1.73–3.92) 3.57 (2.79–4.58) 1.84 (1.36–2.49) 1.45 (0.79–2.65) 1.69 (0.82–3.51)
Values showing he signi icance le el o p < 0.05 a e bolded.
aThose who ha e posi i e/neu al a i udes owa ds bo h socie al iews on old age and subjec i e iews on old age as e e ence ca ego y (n = 1,024, 51.7%). Dis ibu ion o o he ca ego ies:
nega i e socie al iews on old age–nega i e subjec i e iews on old age (n = 212, 10.7%), nega i e socie al iews on old age–posi i e/neu al subjec i e iews on old age (n = 661, 33.4%),
posi i e/neu al socie al iews on old age–nega i e subjec i e iews on old age (N = 84, 4.2%).
bAll a iables included in o he model sepa a ely.
cAll a iables included in o he model simul aneously.
dRa io o p obabili y o ou come in espec i e ca ego ies s e e ence ca ego y.