Like dee caugh in he headligh s: Human e alua ion o he in ensi y o
emo ions in b achycephalic and no mocephalic dogs – A pilo s udy
Pe a E e o ´
a
a,*
, Quanxiao Liu
a
, Te eza Neko ´
aˇ
o ´
a
b,c
, Helena Chaloupko ´
a
a
,
Lucie Pˇ
ibylo ´
a
a
, ˇ
S´
a ka K iˇ
cko ´
a
a
, P´
e e Pong ´
acz
d
a
Depa men o E hology and Companion Animal Science, Facul y o Ag obiology, Food and Na u al Resou ces, Czech Uni e si y o Li e Sciences P ague, Kamýck´
a 129,
P ague 165 00, Czech Republic
b
Depa men o Zoology, Facul y o Science, Cha les Uni e si y, Viniˇ
cn´
a 7, P ague 128 00, Czech Republic
c
Na ional Ins i u e o Men al Heal h, Topolo ´
a 748, Klecany 250 67, Czech Republic
d
Depa men o E hology, ELTE E¨
o ¨
os Lo ´
and Uni e si y, P´
azm´
any P´
e e s´
e ´
any 1/C, Budapes 1117, Hunga y
ARTICLE INFO
Keywo ds:
Dog
B achycephaly
Emo ional cues
Baby schema
Dog-human ela ionship
ABSTRACT
While he acous ic communica ion o dogs ac oss a wide a ie y o con ex s has been s udied in ensi ely, isual
communica ion is, o da e, less unde s ood. Cu en esea ch has ocused on human assessmen o canine signals
o ac o s a ec ing he abili y o humans o unde s and hem. In his con inua ion o ou p e ious pionee ing
s udy (E e o ´
a e al., 2024), we ha e examined he human-pe cei ed in ensi y o basic emo ions (Happiness,
Ange , Fea , Sadness, Cu iosi y) in wo dog b eeds (b achycephalic - Bos on Te ie , no mocephalic – Jack
Russell Te ie ) ia sho , soundless ideos showing canine beha iou s in ou simula ed si ua ions. Pa icipan s
o he s udy submi ed hei assessmen ia online su ey, using a 7-poin Like scale and also p o ided in-
o ma ion abou hei demog aphics, a i ude owa ds dogs, and dog owne ship his o y. S a is ical analysis was
conduc ed in R 4.2.2. using one ull cumula i e link mixed model pe each inspec ed emo ion. Dis ibu ions o
posi i e and nega i e emo ions di e ed ac oss si ua ions in bo h b eeds, sugges ing good ecogni ion o he
p esen ed ma e ials o con ex s e en wi hou any audi o y s imuli p esen . Resul s showed ha owne s o no -
mocephalic dogs iewed Bos on Te ie s as ha ing lowe in ensi ies o posi i e emo ions (Happiness, Cu iosi y;
P <0.05) han b achycephalic dog owne s, while also assessing Jack Russell Te ie s wi h highe sco es o
posi i e and neu al emo ions (P <0.05). Olde pa icipan s also ended o assign lesse in ensi y o posi i e
emo ions han hei younge coun e pa s (P <0.001). Despi e p edic ing ha he Bos on Te ie s would be a ed
wi h an o e all highe in ensi y o posi i e emo ions, his was no he case. This pilo s udy poin s ou ha
con ex and he alence o he emo ion a e c i ical in i s pe cei ed in ensi y, and ha he expe ience and a i ude
o assesso s a e impo an ac o s. Mo e esea ch which employs he scalabili y o emo ions needs o be con-
duc ed on a wide a ie y o canine pheno ypes o p ope ly unde s and human-pe cei ed dog emo ions and he
ac o s a ec ing hem.
1. In oduc ion
Compa ed o i s closes wild-li ing ela i e, he g ey wol (Canis
lupus), domes ic dogs (Canis amilia is) (K o el e al., 2022) show a
ema kable a iabili y in skull size and leng h, body size, and limb and
ail leng h o shape. The sho ening and b oadening o he skull ha e
been p oposed as one o he paedomo phic changes ha eme ged in
ea ly dog domes ica ion (S´
anchez-Villag a e al., 2016). These a e
among he a ious pheno ypic ai s ha e been in ensi ied by humans
du ing he a i icial di ec ional selec ion o mo phological ex emes in
a ious dog b eeds (Wayne, 1986). The skull leng h and shape can be
e alua ed by he cephalic index (CI), which is calcula ed as skull wid h
di ided by skull leng h, hus a highe CI numbe indica es sho e heads.
Se e al small o mid-sized dog b eeds wi h an ex emely high CI (e.g.
F ench Bulldogs) ha e become ex emely popula wi h dog ancie s.
Acco ding o he “baby schema e ec ” (Lo enz, 1943), b achyce-
phalic cha ac e is ics such as la ge eyes, a p onounced o ehead, a small
nose, o a disp opo iona ely la ge, ound head elici nu u ing
* Co esponding au ho .
E-mail add ess: [email p o ec ed] (P. E e o ´
a).
Con en s lis s a ailable a ScienceDi ec
Applied Animal Beha iou Science
jou nal homepage: www.else ie .com/loca e/applanim
h ps://doi.o g/10.1016/j.applanim.2025.106767
Recei ed 22 Ap il 2025; Recei ed in e ised o m 23 July 2025; Accep ed 24 July 2025
Applied Animal Beha iou Science 292 (2025) 106767
A ailable online 24 July 2025
0168-1591/© 2025 Else ie B.V. All igh s a e ese ed, including hose o ex and da a mining, AI aining, and simila echnologies.
endencies in obse ing humans, as hey mimic he look o a human
child. Obse ing pho og aphs o in an aces causes physiological and
emo ional eac ions in humans, pa icula ly in women (Esposi o e al.,
2015; Hildeb and and Fi zge ald, 1978). Simila esul s in he pe cep-
ion o cu eness ha e been achie ed when e alua ing ju enile animal
aces, pa icula ly hose o imma u e dogs and ca s (Bo gi e al., 2014;
Lehmann e al., 2013). The aces o animals whose ea u es ha e been
al e ed o look mo e in an ile (such as enla ging he eyes and so ening
he ea u es) cap i a e human iewe s much mo e han he o iginal
aces (Bo gi e al., 2014; Hech and Ho owi z, 2015). I was ound ha
b eed- o age- ela ed b achycephaly in dogs (such as he sho and ound
heads o puppies) inc eased a pe cep ion o cu eness and/o ado abili y
in humans. In adul dogs, b achycephaly did no necessa ily elici
commen s like ‘cu e’ om obse e s (Alley, 1983; Paul e al., 2023a).
To unde s and he ac o s ha u he enhance he popula i y o
b achycephalic dogs, some esea che s ha e sugges ed ha hese dogs
may exp ess mo e dependen beha iou on humans (Uj alussy e al.,
2023). Dependency has been measu ed by he likelihood o hem look-
ing a humans du ing an ‘unsol able ask’. This sugges s ha b achy-
cephalic dogs can communica e di e en ly wi h humans han
no mocephalic b eeds do, and one could hypo hesise ha i hese b eeds
we e able o exp ess hemsel es in a mo e appealing o e ec i e way,
his could be o hei ad an age in he wo ld o humans.
Acous ic and isual modali ies in canine communica ion a e hose
mos a ec ed by b eed-speci ic changes in canine mo phology. I is
known ha canine ocalisa ion is a ec ed by body size (B´
alin e al.,
2013; Fa ag´
o e al., 2010a; Mo on, 1977) and he inne s a e o he
signalle (Fa ag´
o e al., 2017, 2010b; Sil a e al., 2021). Human lis ene s
a e ai ly accu a e a classi ying he con ex ual and a ec i e con en o
dog ba ks o g owls (B´
alin e al., 2013; E e o ´
a e al., 2020; Fa ag´
o
e al., 2017; Pong ´
acz e al., 2011, 2005; Sil a e al., 2021). On one hand,
while canine acous ic signalling has a ac ed a lo o scien i ic a en ion,
he in e speci ic isual communica ion o dogs had been, un il qui e
ecen ly, su p isingly unde ep esen ed in e hological esea ch, despi e
sigh being he dominan senso y inpu o humans (Spence e al., 2012).
The e is e idence o posi i e selec ion o exp essi e acial signalling in
dogs is he dog-speci ic de elopmen o he inne eyeb ow ele a o
muscle ( esponsible o he ‘sad puppy’ ace) – he co esponding muscle
is almos en i ely missing in wol es (Kaminski e al., 2019). Recen
s udies (Bloom e al., 2021; Bloom and F iedman, 2013) ha e showed
ha adul iewe s a e capable o iden i ying emo ions om pho og aphs
o no mocephalic canine aces.
A canine e sion o FACS (Facial Ac ion Coding Sys em), a coding
sys em designed o desc ibe acial exp ession in humans (Ekman and
F iesen, 1978), called DogFACS (B emho s e al., 2021; Walle e al.,
2013), has become a ailable o esea che s. Addi ionally, a i icial in-
elligence has been ecen ly used o de ec ing sho - e m mic o ex-
p essions in ca s (Feighels ein e al., 2023, 2022) and dogs
(Boneh-Shi i e al., 2022). Howe e , hese ools a e o en s ill limi ed
in hei success when e alua ing signals o dog aces wi h ex eme
mo phological modi ica ions, such as hose o b achycephalic b eeds. I
has been e ealed ha he al e ed acial ana omy o hese dogs (Scha z
e al., 2021) causes ce ain acial exp essions, pa icula ly hose o
a e si e inne s a es, o be ba ely, i a all ecognisable (Goodwin e al.,
1997). I has also been e ealed ha con ex ual isual cues o b achy-
cephalic dogs a e no e y well unde s ood by human ecei e s,
pa icula ly in nega i e (such as ea and agg ession) and s imula ing
si ua ions (such as play ul beha iou ) (Ma el e al., 2025). I was also
ound ha human iewe s a e, on some le el, awa e o he dec eased
signalling abili y (Goodwin e al., 1997; Ma el e al., 2025; Scha z
e al., 2021) o b achycephalic dog aces. Consequen ly, hey end o
igno e hese dogs’ acial exp essions when asked abou heinne s a e o
he dogs (E e o ´
a e al., 2024). Inc easing age has also been no ed a ec
human iewe s o epo exagge a edly high o low le els o s ess in
dogs (Ma i i e al., 2012) o o ha e a sligh nega i e e ec a ecog-
nising posi i e con ex s o human-o ien ed dog beha iou s (E e o ´
a
e al., 2024). Las , bu no leas , dog owne ship has been p o en o ha e
li le e ec on human abili y o co ec ly in e p e canine signals
(E e o ´
a e al., 2024; Lakes ani e al., 2014), including signals o
b achycephalic dogs by people who ha e owned such dogs (E e o ´
a
e al., 2024). Howe e , people who exp essed a ini y owa ds b achy-
cephalic b eed end o iew hem in a mo e posi i e ligh han people
wi h a neu al o nega i e app oach owa ds canine b achycephaly
(Bogn´
a and Kubinyi, 2023; E e o ´
a e al., 2024).
Human iewe s can ecognise mos con ex ual cues when e alua ing
pho os and soundless ideos o canine beha iou , wi h signals o no -
mocephalic dogs being usually be e ecognised han b achycephalic
dogs (E e o ´
a e al., 2024). Howe e , any di e ence a iden i ying
isually-only signalled emo ions by people in he wo g oups o dogs
wi h a ying skull leng h has no ye been examined. While i has been
shown ha humans a ied in success a e o iden i ying signals o dogs
om isual o mixed isual-audi o y s imuli (E e o ´
a e al., 2024;
Mein s e al., 2018; Tami and Gallaghe , 2009), he pe cep ion o in-
ensi y o indi idual emo ions is no well explo ed in he case o isual
exp essions. Mos esea ch ocusing on iden i ying emo ions in dogs by
people ope a es in absolu es (i.e. whe he one pa icula emo ion is
p esen o no in a sample s imulus), while, in ac , emo ions o en a y
in hei in ensi y.
While he e a e mul iple ways o de ine emo ions (Mulligan and
Sche e , 2012), he e we unde s and hem as de ined by Paul and Mendl
(2018): “an emo ion is a mul icomponen (subjec i e, physiological, beha -
iou al and cogni i e) esponse o a s imulus o e en ha is ypically o
impo ance o he indi idual, i is always alenced (pleasan o unpleasan )
and can a y in ac i a ion/a ousal and du a ion/pe sis ence.” Ande son
and Adolphs (2014) p oposed ou basic emo ional desc ip o s: alence
(posi i e o nega i e), scalabili y (in ensi y exp essed by a scale),
pe sis ence (linge ing p esence) and gene alisa ion (abili y o espond
based on p e ious expe ience). While hese desc ip o s a e key aspec s
o human emo ions, hey a e also mani es in a ious o ms ac oss a wide
ange o animal species, which p o ides a obus basis o compa a i e
s udies (i.e. Mu phy e al., 2014 o a e iew on measu ing emo ions in
pigs, Boissy e al., 2007 o measu ing emo ions in a m animals;
Lundblad, 2018 o inspec ing acial exp essions as s ess indica o s in
ho ses; o Paul e al., 2005 o measu ing emo ional alence in animals).
Emo ional s a es and hei al e a ion may be demons a ed by beha -
iou al and physiological eac ions, such as changes in locomo ion, acial
exp essions, hea and espi a o y a es, ho mone and neu o ansmi e
le els, body empe a u e, and inne eye co ne empe a u e
(Nee hi ajan e al., 2021).
In his pape we demons a e a no el app oach owa ds measu ing
he human pe cep ion o isually exp essed emo ions in dogs by
employing i e basic a ec i e s a es (emo ions) ha we e sco ed indi-
idually by hei pe cei ed in ensi y. We used his me hod o emo ional
scaling o compa ing he e icacy o isual signalling in wo dog b eeds
ha di e ed in hei cephalic index: a no mocephalic and a b achyce-
phalic b eed.
2. Objec i es and hypo heses
The p esen ed s udy is pa o a la ge p ojec ocused on he isual
cues o b achycephalic dogs. In he i s pa (E e o ´
a e al., 2024),
which sha es he same backg ound, ma e ial collec ion me hod and
expe imen o he p esen s udy, we examined he human abili y o
ecognise con ex ual signals o b achycephalic and no mocephalic dogs
and speci ic body and acial a eas o human in e es . Now, we ha e
in es iga ed di ec ly he human-pe cei ed in ensi ies o emo ions
a ibu ed o b achycephalic and no mocephalic dogs, ega dless o
hei ac ual co ec ness o alence.
In his s udy, i e ca ego ies o emo ions we e used o assess human-
pe cei ed emo ional in ensi y in b achycephalic and no mocephalic
dogs. Fou o hese ep esen basic emo ions ha a e conside ed uni-
e sal, while he i h — ‘Cu iosi y‘— was included as a neu al a ec i e
P. E e o ´
a e al.
Applied Animal Beha iou Science 292 (2025) 106767
2
s a e o en associa ed wi h explo a o y beha iou and ac i e in e es in
he en i onmen , due o i s wide p e alence in human-o ien ed canine
beha iou . These a ec i e s a es we e selec ed based on he assump ion
ha hey can be eliably elici ed and demons a ed in con olled
expe imen al si ua ions ( o inducing emo ions in dogs unde expe i-
men al se ings, see B emho s e al., 2022). We ocused on in es iga ing
(1) how human iewe s pe cei ed he in ensi ies o he i e emo ions
om sho , soundless ideos cap u ing he beha iou s o wo dog b eeds
in ou simula ed si ua ions. The wo b eeds (Bos on Te ie – b achy-
cephalic b eed, Jack Russell Te ie – no mocephalic b eed) we e cho-
sen based on hei o e all simila i ies (size, body p opo ions,
empe amen ), wi h he main pheno ypical di e ence be ween hem
being hei skull leng hs (and esul ing cephalic indexes). Addi ionally,
we examined ac o s on he side o human obse e s ega ding hei
abili y o assess dog emo ions. These ac o s included (2) hei age and
(3) hei ac ual expe ience wi h dogs in gene al and b achycephalic
dogs in pa icula , as well as (4) hei subjec i e sel -assessmen
ega ding expe ience wi h b achycephalic dogs.
We hypo hesised ha (i) he in ensi ies o posi i e and neu al
emo ions (Happiness, Cu iosi y) would be o e all signi ican ly highe in
b achycephalic dogs han in no mocephalic dogs, while he in ensi ies o
nega i e emo ions (Ange , Fea , Sandess) would no signi ican ly di e
be ween he wo b eeds. We hypo hesised ha (ii) owne s o b achy-
cephalic dogs and pa icipan s who did no own any dogs would a e
hese dogs wi h highe in ensi ies o posi i e and neu al emo ions han
hey would a e he no mocephalic dogs, while owne s o no moce-
phalic dogs would e alua e hem wi h lowe in ensi ies o posi i e o
neu al emo ions. We hypo hesised ha (iii) people who exp essed hei
willingness o owning a b achycephalic dog in he u u e would a e
hese dogs wi h highe in ensi y o posi i e emo ions and lowe in ensi y
o nega i e emo ions han people who would no conside ge ing a
b achycephalic dog. Las , bu no leas , we hypo hesised ha (i ) pa -
icipan s’ age would be in a nega i e associa ion wi h he in ensi y o
hei emo ional a ings.
3. Me hods
The ma e ials used in he s udy, along wi h he expe imen , is sha ed
wi h ou p e ious s udy (E e o ´
a e al., 2024). The p e ious s udy
analysed he human abili y o unde s and con ex ual cues o dogs and
a eas o he dogs’ bodies and aces ha we e mos in o ma i e o human
iewe s. The p esen ed s udy examined he human pe cep ion o
emo ional cues o he dogs.
3.1. Dogs ea u ed o he s udy
Al oge he , he beha iou s o 23 dogs we e eco ded. We eco ded
16 Bos on Te ie s as ep esen a i es o b achycephalic dogs (10 e-
males, 6 males; age ange 1.25–6 yea s, mean age 3.17 yea s,
σ
=1.64)
and 7 Jack Russell Te ie s o no mocephalic dogs (2 males, 5 emales;
age ange 2–4 yea s, mean age 2.75 yea s,
σ
=0.96). The owne s o he
dogs olun ee ed o he s udy on he basis o public calls ia Facebook.
The esea ch was conduc ed in acco dance wi h he guidelines o
he use o li e animals in esea ch, as ou lined by he Associa ion o he
S udy o Animal Beha iou (ASAB). The expe imen was ully non-
in asi e o he canine and human pa icipan s. The me hodology was
e iewed by he Animal Wel a e Commi ee o E¨
o ¨
os Lo ´
and Uni e si y
(Ce i ica e numbe PEI/001/1056–4/2015). W i en in o med consen
was ob ained om e e y dog owne pa icipa ing in he s udy. As he
ideo aping o he dogs did no in ol e human expe imen a ion o
sensi i e da a collec ion om he dog owne s, i was no necessa y o
ob ain human e hical app o al. The ideos ea u ed in he su ey we e
edi ed so ha pa icipa ing dog owne s we e no iden i iable. The dog
owne s we e in o med ha hey could wi hd aw om he expe imen a
any poin . Ma e ial collec ion was ca ied ou a he e hological labo-
a o y (a came a-equipped iled oom wi h a alse wall, dimensions
4.45 ×3.68 m) o he Depa men o E hology a E¨
o ¨
os Lo ´
and Uni-
e si y, Budapes , Hunga y, wi h ma e ial p ocessing and su ey c ea-
ion being ca ied ou mos ly a Depa men o E hology and Companion
Animal Science, Czech Uni e si y o Li e Sciences P ague, Czechia.
3.2. Pa icipan s in he online su ey
In o al, 350 pa icipan s ully comple ed and submi ed he ques-
ionnai e. In o med consen was ob ained om e e y pa icipan a he
e y beginning o he su ey. P oceeding wi h he su ey was impossible
wi hou comple ing he in o med consen . The in o med consen o m
was app o ed by he E hical Commi ee o he Czech Uni e si y o Li e
Sciences in P ague, no. 07/2023. The su ey included ques ions abou
he pa icipan s’ demog aphics. Pa icipan s we e asked o disclose hei
age (in ull yea s) and gende (male/ emale/do no wish o disclose).
Ano he ques ion aimed o in es iga e he numbe o child en he pa -
icipan s ha e aised a he ime o he su ey – answe ing his ques ion
was olun a y. Pa icipan s could lea e hei e-mail add esses o o he
con ac in o ma ion o ecei e in o ma ion abou he s udy o he al-
idi y o hei answe s bu we e no demanded o do so. All pe sonal
in o ma ion p o ided by he pa icipan s was anonymised and handled
in acco dance wi h he Gene al Da a P o ec ion Regula ion (Regula ion
(EU) 2016/679) (GDPR).
The su ey was dis ibu ed in h ee language e sions (Czech – 246
pa icipan s, Hunga ian – 93 pa icipan s, English – 14 pa icipan s).
Pa icipan s had o be 18 yea s old o olde , ega dless o hei dog
owne ship s a us o le el o dog- ela ed expe ience. The pa icipan s
anged be ween 18 and 73 yea s old (mean age 33.33 yea s,
σ
=11.74
yea s; median 30 yea s). O all pa icipan s, 90 % epo ed ha ing
owned dogs in he pas o p esen , and 25.5 % epo ed ha ing owned a
leas one b achycephalic dog in hei li es.
3.3. Expe imen design
3.3.1. Ob aining he s imuli
In o de o demons a e o human iewe s he body cues and acial
exp ession o he ea u ed dogs, we ideo- eco ded he dogs in ou
simula ed si ua ions (‘Called by name’, ‘Play’, ‘Sepa a ion om owne ’,
‘Th ea ened by a s ange ’). All dogs we e eco ded in indi idual ses-
sions by he same wo expe imen e s ( wo women in hei 20 s). The
expe imen used wo came as (Sony FDR-AX33). One was moun ed on a
ipod and eco ded con inuously, ope a ed by Expe imen e 1 du ing
he i s wo si ua ions (‘Called by name’, ‘Play’), while unning eely in
he las wo (‘Sepa a ion’, ‘S ange ’). The o he came a was hand-
ope a ed by Expe imen e 2 du ing he i s wo si ua ions and by
Expe imen e 1 du ing he las wo.
The eco ding session was p eceded by a 5-minu e pe iod o ee lab
explo a ion and habi ua ion o each dog. Du ing his pe iod, he owne
was in o med abou he expe imen and he /his ole in i . No dogs
p esen o ilming ailed o habi ua e wi hin he 5-minu e pe iod. Each
dog was eco ded only once and indi idually, e en in cases when one
owne a i ed wi h mul iple dogs. In such cases, he dogs we e habi u-
a ed oge he and only sepa a ed o he indi idual es ing. Du ing he
habi ua ion phase, Expe imen e 1 was ee o engage wi h he dogs
while Expe imen e 2 a oided con ac o main ain un amilia i y. I was
explained o he owne s no o engage wi h he dogs physically o
e bally du ing he expe imen .
A e he 5-minu e ee explo a ion phase, he expe imen began. I
consis ed o he ollowing eigh s eps, each las ing 1 min. The o de o
he si ua ions was ixed. Jus be o e he i s si ua ion occu ed, he dog
was leashed o a handle secu ed o he loo nea he owne . The leash
allowed he dog o si o s and in on o he owne bu p e en ed
physical con ac wi h he ipod came a. I he s ep design pe mi ed, he
owne sa on a old-ou chai a he back o he oom.
The s ep-by-s ep sequence o he expe imen and addi ional in o -
ma ion a e p esen ed in Table 1. The in e ludes se ed o allow he dogs
P. E e o ´
a e al.
Applied Animal Beha iou Science 292 (2025) 106767
3
o peace ully ansi ion be ween emo ions be o e he nex si ua ion
occu ed. We ha e no encoun e ed a case in which a dog would be
unable o pa icipa e in he expe imen .
3.3.2. Ma e ials
The ob ained ideo eco dings showing dogs’ beha iou s in he ou
si ua ions we e ca ego ised by he con ex s (si ua ions), and all ideos
we e mu ed. The ideos we e hen included in an online su ey
dis ibu ed o he gene al public. Fo his s udy, assessmen o emo ional
in ensi y was collec ed o each combina ion o b eed/si ua ion.
The ideo sequences ob ained h ough he expe imen we e u he
edi ed using A idemux 2.7 (coded by ‘Mean’, ‘G un s e ’, and ‘Fah ’,
2019), a ee ideo-edi ing so wa e. Soundless ideo clips las ing 5–10 s
we e c ea ed om he sou ce ma e ials. Only ma e ials showing he dogs
om a on al o semi- on al angle whe e bo h eyes and he en i e ace
we e isible we e included in p e-selec ion. The main c i e ion o he
ideos was su icien ly p esen beha iou s ypical o dogs in each
ea u ed si ua ion. Ou o he esul ing pool o p e-selec ed ideos, 16
ideos o beha iou s o Bos on Te ie s and 16 ideos o beha iou s o
Jack Russell Te ie s ( ou ideos pe each si ua ion in each b eed) we e
included in he inal su ey. The ideos we e selec ed based on a dis-
cussion wi h animal beha iou and communica ion expe s a ilia ed
wi h ei he o he pa icipa ing ins i u ions. The expe s pa icipa ing in
he ideo selec ion p ocess we e handed he en i e pool o ideos and
asked o selec hose hey pe cei ed as he mos accu a e o each si u-
a ion and b eed combina ion. Using his me hod, he inal ba ch o 4
ideos pe b eed and si ua ion was ob ained. These we e hen assigned
a andom o one o he ou e sions o he su ey o achie e necessa y
andomisa ion. Each e sion o he su ey con ained one ideo o each
combina ion o b eed/si ua ion.
3.3.3. Ques ionnai e
The s udy was conduc ed using an online su ey p o ided by Google
Docs. This pla o m was chosen o i s easy a ailabili y and uncompli-
ca ed p og amming, which allowed us o c ea e mul iple e sions o he
su ey wi hin one documen . Th ee language copies (Czech, Hunga ian,
and English) iden ical in hei con en s and o de o ma e ial p esen a-
ion we e c ea ed ( o he English e sion, see Supplemen a y Ma e ial
I). Da a collec ion was open om June 2020 o No embe 2021. Each
pa icipan could only ill ou he su ey once. The su ey included he
demog aphic pa ame e s o he pa icipan s as desc ibed abo e.
Fu he mo e, we enqui ed in o e e y pa icipan ’s dog-owne ship
his o y (‘ ealised owne ship’). When a pa icipan epo ed ha ing
owned a leas one dog in hei li e, hey p oceeded o speci y wha kind
o dog hey owned. We asked hem whe he hey had e e owned a leas
one b achycephalic dog (such as a Bos on Te ie , F ench Bulldog, En-
glish Bulldog, Pug o Shih-Tzu – ‘b achycephalic’) o had e e owned
dolichocephalic o mesocephalic (‘no mocephalic’) dogs (such as a Jack
Russell Te ie ).
Mo eo e , pa icipan s we e asked o disclose hei a i udes owa ds
dog owning. We asked hem whe he hey would e e conside owning a
b achycephalic dog, o only e e a no mocephalic dog, o whe he hey
would p e e ne e owning a dog a all (‘hypo he ical owne ship’).
Pa icipan s had o choose one o h ee possible answe s – hey would
ne e conside owning any dogs (‘ne e ’), hey would conside owning
only dolichocephalic o mesocephalic dogs (‘no mocephalic’), o ha
hey would conside owning b achycephalic dogs (‘b achycephalic’). All
pa icipan s answe ed ques ions ega ding bo h ypes o owne ships
( ealised and hypo he ical) and bo h alues we e inspec ed sepa a ely.
Finally, pa icipan s epo ed hei expe ience wi h b achycephalic
dogs on a 7-poin Like scale (‘subjec i e expe ience’). On his scale, 1
s ands o he mos inexpe ienced while 7 o mos expe ienced.
Following his, pa icipan s we e asked o selec one o he ou andom
e sions o he su ey and could p oceed.
In each e sion o he su ey, pa icipan s we e p esen ed a se o 8
ideos (4 pe each b eed and he 4 p esen ed si ua ions). The o de o
si ua ions was andomised in all su ey e sions. All ou si ua ions o
bo h b eeds we e ep esen ed in each e sion o he su ey by one ideo,
o alling eigh ideos seen by e e y s udy pa icipan . Pa icipan s also
de e mined he con ex s (si ua ions) o he p esen ed ma e ials as well as
e alua ing he emo ional in ensi y (see E e o ´
a e al., 2024), hus hei
emo ional in ensi y assessmen s could ha e been somewha in luenced
Table 1
S ep-by-s ep sequence o e iew o he expe imen al si ua ions and oles o ea u ed pe sonnel.
Phase Dog Expe imen e 1 Expe imen e 2 Owne Du a ion Expec ed beha iou s
P e-phase F ee explo a ion F ee in e ac ion wi h
he dog
No in e ac ion wi h he dog Filling ou
backg ound
in o ma ion o m
5 min
Si ua ion 1 – Dog
called by name
(‘Called’)
Being called by
name by
Expe imen e 1,
leashed
Calling he dog by
name while
ope a ing ipod
came a
Ope a ing hand-held came a Si ing on he chai
behind he dog
1 min Eye con ac wi h he
expe imen e , head il , ea
wi ch
In e lude 1 – F ee play Playing eely wi h
a ennis ball,
unleashed
Playing wi h he dog Filming on he hand-held came a Si ing on he chai
behind he dog.
1 min
Si ua ion 2 – Play
(‘Play’)
Playing and
eac ing o a ennis
ball, leashed
Playing wi h he dog
while ope a ing
ipod came a
Ope a ing hand-held came a Si ing on he chai
behind he dog
1 min Eye con ac a he ennis ball,
jumping, ba king, a emp ed
chasing o he ball
In e lude 2 – Owne
lea ing
Remained leashed
in place
Ope a ing he hand-
held came a, no
in e ac ion wi h he
dog
Le he oom Le he oom <1 min
Si ua ion 3 –
Sepa a ion om he
owne (‘Sepa a ion’)
Remained leashed
in place
Ope a ing he hand-
held came a, no
in e ac ion wi h he
dog
Absen om he oom Absen om he
oom
1 min Eye con ac in he di ec ion o
he owne , pacing, ba king o
whining
In e lude 3 – Owne
e u ning
Remained leashed
in place
Ope a ing he hand-
held came a, no
in e ac ion wi h he
dog
En e ed he
oom ollowing Si ua ion 3, wen o
s and behind he dog
En e ed he oom
a e he owne
and closed he
doo
<1 min
Si ua ion 4 –
Th ea ened by a
s ange (based on
Vas e al., 2005)
(‘S ange ’)
Remained leashed
in place
Ope a ing he hand-
held came a, no
in e ac ion wi h he
dog
App oached he dog wi h a
hunched-o e pos u e and wide,
sliding s eps while main aining eye
con ac wi h he dog (s a ing in o i s
eyes), blinking as li le as possible
S anding behind
he dog while no
ini ia ing con ac
wi h i
1 min Eye con ac wi h he
expe imen e ollowed by
looking away, lip licking, head
down, looking back a he
owne , backing away
P. E e o ´
a e al.
Applied Animal Beha iou Science 292 (2025) 106767
4
by which si ua ion hey ma ched o each ma e ial.
Fo analysis, we chose he pa icipan - epo ed in ensi ies o i e
emo ions (Happiness, Ange , Fea , Sadness, Cu iosi y) displayed by he
dogs. Howe e , hese emo ions may no ha e been expe ienced by he
dogs (o we e no e i ied) bu could ha e been a ibu ed o hem by
human obse e s. We did no inspec he ue a ec i e s a es in he dogs
ia DogFACS o o he means, ins ead we concen a ed me ely on he
human pe cep ion o hem. Fo each ideo, a se o i e emo ions
(Happiness, Ange , Fea , Sadness, Cu iosi y) was p o ided on a 7-
poin Like scale, whe e 1 was he lowes in ensi y and 7 he highes
in ensi y. The in ensi y o each epo ed emo ion was assessed sepa-
a ely and independen ly o he o he emo ions. Pa icipan s we e asked
o a e he pe cei ed in ensi y o all epo ed emo ions in each ea u ed
ideo (‘emo ional in ensi y’). These emo ions we e chosen as hey
we e, in gene al, easonable o expec in he dogs in he chosen si ua-
ions and because hey should be pe cei able by human iewe s. The
si ua ions ‘Called by name’ and ‘Play’ we e designed o elici mos ly
posi i e (Happiness) o neu al (Cu iosi y) emo ions in he dogs, while
‘Sepa a ion’ and ‘S ange ’ we e designed o elici mos ly nega i e
emo ions in he dogs (Ange , Fea , Sadness). Pa icipan s could no
p oceed wi h he su ey un il hey had ma ked he in ensi y o all e-
po ed emo ions o all p esen ed ideos. Once he pa icipan s had
a ed he en i e ideo se , hey we e allowed o submi he su ey.
3.4. S a is ical analysis
Analysis was pe o med in R 4.2.2 (R co e eam, 2022)
To in es iga e how b eed, ealised owne ship, hypo he ical owne -
ship and age a ec he e alua ion o emo ional in ensi ies om he
ideos, o each emo ion, we c ea ed one ull mixed model wi h cumu-
la i e link (‘clmm’ unc ion in he ‘o dinal’ package in R) o a oid high
compu a ional cos s. In o al, we c ea ed i e ull models. The sco e o
emo ion was he o de ed, ac o ial esponse a iable. The in e ac ion
be ween hypo he ical owne ship (would own a b achycephalic dog –
‘b achy’/would only conside owning a no mocephalic dog – ‘no mo’/
would ne e conside owning a dog – ‘ne e ’), b eed (Bos on e ie –
‘bos on’/Jack Russell Te ie – ‘jack’) and si ua ion (‘Play’/ ‘Called’/
‘Sepa a ion’/‘S ange ’), he in e ac ion be ween ealised owne ship
(owns o owned a leas one b achycephalic dog – ‘b achy’/only owned
no mocephalic dogs – ‘no mo’/ne e owned a dog – ‘ne e ’), b eed and
si ua ion, and he age o he pa icipan s we e included as ixed ac o s
and he co a ia e. The IDs o pa icipan s and dogs we e ea ed as
andom in e cep s. We hen compa ed his ull model o all o i s de-
i a i e models (‘d edge’ unc ion in he package ‘MuMIn’) and selec ed
he mos pa simonious model wi hin 2ΔAICc o he op model as he bes
model. I any in e ac ion was included in he bes model, we pe o med
pos hoc analysis (‘emmeans’ unc ion in package ‘emmeans’) o examine
he di e ences be ween in e ac ed ac o s. We also boo s apped he
mean and 95 % CIs o he emo ion sco e pe emo ion (‘boo ’ and ‘boo .ci’
unc ions in he ‘boo s’ package in R).
4. Resul s
De ailed da a cha ac e is ics, including in ensi y dis ibu ion pe
each emo ion and desc ip i e s a is ics in Bos on Te ie s, can be ound
in Supplemen a y Ma e ials 2 Table I-1 Desc ip ion o he Bos on Te ie
emo ional in ensi y alues. Fo in ensi y dis ibu ion pe each emo ion
and desc ip i e s a is ics in Jack Russell Te ie s, see Supplemen a y
Ma e ials 2 Table I-2 Desc ip ion o he Jack Russell Te ie emo ional
in ensi y alues.
The bes model o each emo ion had di e en combina ions o he
in e ac ions o ixed ac o s and co a ia e (See Table 2). Only e ec s
p o en as signi ican we e included in he inal models. The in ensi y o
each ea u ed emo ion was inspec ed independen ly. The in e ac ion o
b eed and si ua ion was included in all bes models excep o he model
o Cu iosi y. The in e ac ion be ween ealised owne ship and b eed was
included in he bes models o Happiness and Ange . The in e ac ion
be ween hypo he ical owne ship and b eed was included in he bes
model o Cu iosi y, and he in e ac ion be ween hypo he ical owne -
ship and si ua ion was included in he bes model o Sadness. O e all,
he age o pa icipan s had a weakly nega i e o no e ec on hei
assessmen o canine emo ions.
No inspec ed emo ion showed an e ec o subjec i e expe ience
( esul s no shown).
4.1. Happiness
The desc ip i e his og am showing he o al aw sco es (be o e s a-
is ical analysis was done) o Happiness a ings o all si ua ions com-
bined and di ided by b eed can be seen in Supplemen a y Ma e ials 2
Figu e II-1 - To al dis ibu ion o Happiness in ensi y in all si ua ions in
Bos on Te ie s and Jack Russell Te ie s. Fo analysis, we chose he bes
i ing model based on ΔAICc. Fo i e bes i ing models o Happiness,
see Supplemen a y Ma e ials 2 Table II-1 Fi e bes i ing models o
analysing Happiness based on ΔAICc.
The in e ac ion be ween si ua ions and b eed, he in e ac ion be-
ween ealised owne ship and b eed, and he age o he pa icipan s
signi ican ly a ec ed a ings o Happiness (see Supplemen a y Ma e ials
2 Table II-2 Full model analysis de ails o Happiness). Fo Bos on Te -
ie s, ‘Play’ (mean [boo s apped 95 % CIs] =3.31 [3.12, 3.51], he
same o all alues lis ed below) had a highe Happiness a ing han
o he si ua ions. ‘Called’ (2.58 [2.41, 2.79]) was a ed highe in
Happiness han ‘Sepa a ion’ (1.47 [1.38, 1.56]). The si ua ion o
‘S ange ’ (2.2 [2.04, 2.37]) did no di e om ei he ‘Sepa a ion’ o
‘Called’ in he pos hoc es .
In Jack Russell Te ie s, he in ensi y a ing o Happiness was highes
in he si ua ion ‘Called’ (5.30 [5.11, 5.49]); ‘Called’ and ‘Play’ (3.90
[3.63, 4.16]) we e a ed highe han ‘S ange ’ (2.06 [1.90, 2.23]) and
‘Sepa a ion’ (1.79 [1.67, 1.92]), and he la e wo did no di e om
each o he . Be ween he wo b eeds, Jack Russell Te ie s we e a ed
highe o Happiness han Bos on Te ie s o he ‘Called’ si ua ion. The
e ec o he in e ac ion o b eed and si ua ion in he wo b eeds,
including di e ences in he a ing o Happiness, can be seen in Fig. 1. A
de ailed pos -hoc analysis o he b eed and si ua ion in e ac ion on he
a ing o Happiness can be seen in Supplemen a y Ma e ials 2 Table II-3
Emmeans o b eed/si ua ion in e ac ion o Happiness.
The in e ac ion be ween ealised owne ship and b eed signi ican ly
a ec ed he a ing o Happiness as including his in e ac ion signi i-
can ly imp o ed he i ing o he model. No mocephalic dog owne s
ga e Bos on Te ie s lowe Happiness sco es (2.31 [2.21, 2.43]) han
b achycephalic dog owne s (2.46 [2.28, 2.65]). Addi ionally, when
a ing happiness o Bos on Te ie s and Jack Russell Te ie s, no mo-
cephalic dog owne s a ed Bos on Te ie s lowe han hey o Jack
Russell Te ie s (3.33 [3.18, 3.48]). The e ec o in e ac ions o b eed,
a) ealised owne ship and b) hypo he ical owne ship, including di e -
ences on he a ing o Happiness, can be seen in Fig. 2. A de ailed pos -
Table 2
Summa y o ixed ac o s and in e ac ions in he bes models. The √ ma ks a
signi ican complex e ec , while he ↓ ma ks a nega i e e ec .
Emo ion/E ec Happiness Sadness Ange Fea Cu iosi y
Age ↓ ↓
B eed * Si ua ion √ √ √ √
B eed * Realised owne ship √√
B eed * Hypo he ical
owne ship
√
Si ua ion * Hypo he ical
owne ship
√
Si ua ion √ √ √ √ √
B eed √ √ √ √ √
Realised owne ship √√
Hypo he ical owne ship √ √
P. E e o ´
a e al.
Applied Animal Beha iou Science 292 (2025) 106767
5
hoc analysis o he b eed and ealised owne ship in e ac ion on he
a ing o Happiness can be seen in Supplemen a y Ma e ials 2 Table II-4
Emmeans o b eed/ ealised owne ship in e ac ion o Happiness.
The e was a sligh nega i e e ec o pa icipan s’ age (OR =0.97; CI
=0.96 – 0.98; p <0.001; o de ails see Supplemen a y Ma e ials 2
Table II-1 Full model analysis de ails o Happiness) on he in ensi y
a ing o Happiness in dogs.
4.2. Ange
The desc ip i e his og am showing he o al aw sco es (be o e s a-
is ical analysis was done) o Ange a ings o all si ua ions combined
and di ided by b eed can be seen in Supplemen a y Ma e ials 2
Figu e III-1 – To al dis ibu ion o Ange in ensi y in all si ua ions in
Bos on Te ie s and Jack Russell Te ie s. Fo i e bes i ing models o
Fig. 1. cap ion: The e ec o b eed and si ua ion in e ac ion on he dis ibu ion o he in ensi y a ing o Happiness in Bos on Te ie s and Jack Russell Te ie s. The
do and whiske s deno e emmeans and 95 % CIs. The wid h o he iolin ep esen s he dis ibu ion o he sco es, and he dashed lines indica e qua iles. Same o
Figs. 2–9. †p<0.1; * p <0.05; *** p <0.001.
Fig. 2. The e ec o a) ealised owne ship and b) hypo he ical owne ship on pa icipan s’ dis ibu ion o he in ensi y a ing o Happiness in Bos on Te ie s and Jack
Russell Te ie s. Emmeans, 95 % CI, †p<0.1; * p <0.05.
P. E e o ´
a e al.
Applied Animal Beha iou Science 292 (2025) 106767
6
Ange , see Supplemen a y Ma e ials 2 Table III-1 Fi e bes i ing models
o analysing Ange based on ΔAICc.
The wo in e ac ions (in e ac ion be ween si ua ions and b eed,
in e ac ion be ween ealised owne ship and b eed) a ec ed a ings o
Ange (see Supplemen a y Ma e ials 2 Table III-2 Full model analysis
de ails o Ange ). Al hough he e we e no di e ences in he a ings o
Ange in all si ua ions in Bos on Te ie s (‘Called’ =1.22 [1.15, 1.31];
‘Play’ =1.28 [1.20, 1.37]; ‘Sepa a ion’ =1.33 [1.27, 1.41]; ‘S ange ’ =
1.22 [1.16, 1.29]), in Jack Russell Te ie s, he Ange a ing o he si -
ua ion ‘Play’ (2.37 [2.18, 2.57]) was highe han ‘Called’ (1.34 [1.09,
1.19]) and ‘S ange ’ (1.7 [1.57, 1.82]), bu was no s a is ically
di e en om ‘Sepa a ion’ (1.47 [1.38, 1.57]). Be ween he wo b eeds,
Bos on Te ie s we e a ed wi h less Ange in ‘Play’ and ‘S ange ’ si -
ua ions han Jack Russell Te ie s bu we e a ed wi h mo e Ange in he
‘Called’ si ua ion. The e ec o he in e ac ion o b eed and si ua ion in
he wo b eeds, including di e ences in he a ing o Ange , can be seen
in Fig. 3. A de ailed pos -hoc analysis o he b eed and si ua ion in e -
ac ion on he a ing o Ange can be seen in Supplemen a y Ma e ials 2
Table III-3 Emmeans o b eed/si ua ion in e ac ion o Ange .
No mocephalic dog owne s ga e Bos on Te ie s lowe Ange sco es
(1.26 [1.21, 1.30]) han hey ga e Jack Russell Te ie s (1.74 [1.65,
1.83]). In e es ingly, pa icipan s wi hou any dogs a ed Bos on Te -
ie s as ang ie (1.52 [1.37, 1.72]) han did pa icipan s owning a leas
one b achycephalic dog (1.18 [1.13, 1.24]). The e ec o in e ac ions o
b eed, a) ealised owne ship and b) hypo he ical owne ship, including
di e ences on he a ing o Ange , can be seen in Fig. 4. A de ailed pos -
hoc analysis o he b eed and ealised owne ship in e ac ion on he
a ing o Ange can be seen in Supplemen a y Ma e ials 2 Table IIII-4
Emmeans o b eed/ ealised owne ship in e ac ion o Ange .
4.3. Fea
The desc ip i e his og am showing he o al aw sco es (be o e s a-
is ical analysis was done) o Fea a ings o all si ua ions combined and
di ided by b eed can be seen in Supplemen a y Ma e ials 2 Figu e IV-1 –
To al dis ibu ion o Fea in ensi y in all si ua ions in Bos on Te ie s
and Jack Russell Te ie s. Fo i e bes i ing models o Fea , see
Supplemen a y Ma e ials 2 Table IV-1 Fi e bes i ing models o ana-
lysing Fea based on ΔAICc.
The a ings o Fea we e only a ec ed by he in e ac ion be ween
si ua ion and b eed (see Supplemen a y Ma e ials 2 Table IV-2 Full
model analysis de ails o Fea ). In Bos on Te ie s, ideos o he ‘Play’
si ua ion we e a ed wi h he leas Fea (1.67 [1.54, 1.81]) compa ed o
‘Called’ (2.24 [2.07, 2.42]), ‘S ange ’ (2.72 [2.55, 2.91]) and ‘Sepa a-
ion’ (3.54 [3.35, 3.75]). The ‘Called’ si ua ion also go a lowe Fea
sco e han ‘Sepa a ion.’ In Jack Russell Te ie s, he ‘Called’ si ua ion
was a ed wi h he lowes Fea (1.38 [1.29, 1.47]), ollowed by ‘Play’
(2.07 [1.92, 2.24]), ‘Sepa a ion’ (2.90 [2.72, 3.09]) and ‘S ange ’ (4.48
[4.27, 4.61]). Be ween he wo b eeds, Bos on Te ie s we e a ed wi h
mo e Fea han Jack Russell Te ie s in ‘Called’ and ‘Sepa a ion’ bu
we e a ed as ha ing less Fea in he ‘S ange ’ si ua ion. The e ec o
he in e ac ion o b eed and si ua ion in he wo b eeds, including di -
e ences in he a ing o Fea , can be seen in Fig. 5. A de ailed pos -hoc
analysis o he b eed and si ua ion in e ac ion on he a ing o Fea can
be seen in Supplemen a y Ma e ials 2 Table IV-3 Emmeans o b eed/
si ua ion in e ac ion o Fea .
4.4. Sadness
The desc ip i e his og am showing he o al aw sco es (be o e s a-
is ical analysis was done) o Sadness a ings o all si ua ions combined
and di ided by b eed can be seen in Supplemen a y Ma e ials 2 Figu e V-
1 – To al dis ibu ion o Sadness in ensi y in all si ua ions in Bos on
Te ie s and Jack Russell Te ie s. Fo i e bes i ing models o
Sadness, see Supplemen a y Ma e ials 2 Table V-1 Fi e bes i ing
models o analysing Sadness based on ΔAICc.
Sadness a ings we e a ec ed by he in e ac ion be ween si ua ion
and b eed, as well as he in e ac ion be ween he hypo he ical
Fig. 3. cap ion: The e ec o b eed and si ua ion in e ac ion on he dis ibu ion o he in ensi y a ing o Ange in Bos on Te ie s and Jack Russell Te ie s. †p<0.1;
* p <0.05; *** p <0.001.
P. E e o ´
a e al.
Applied Animal Beha iou Science 292 (2025) 106767
7
owne ship and si ua ion (see Supplemen a y Ma e ials 2 Table V-2 Full
model analysis de ails o Sadness). In Bos on Te ie s, ideos o he
‘Play’ si ua ion we e a ed wi h he lowes in ensi y o Sadness (1.59
[1.47, 1.72]) compa ed o ‘Called’ (2.01 [1.83, 2.18]), ‘S ange ’ (2.37
[2.18, 2.55]), and ‘Sepa a ion’ (3.02 [2.81, 3.23]). In Jack Russell
Te ie s, he ‘Play’ (1.49 [1.38, 1.63]) and ‘Called’ (1.19 [1.13, 1.26])
si ua ions we e a ed wi h less in ensi y o Sadness han ‘S ange ’ (2.37
[2.19, 2.57]) and ‘Sepa a ion’ (2.91 [2.72, 3.11]). Be ween he wo
b eeds, Bos on Te ie s we e a ed wi h a highe in ensi y o Sadness
han Jack Russell Te ie s in he ‘Called’ si ua ion. The e ec o he
in e ac ion o b eed and si ua ion in he wo b eeds, including di e -
ences in he a ing o Sadness, can be seen in Fig. 6. A de ailed pos -hoc
Fig. 4. cap ion: The e ec o a) ealised owne ship and b) hypo he ical owne ship on pa icipan s’ dis ibu ion o he in ensi y a ing o Ange in Bos on Te ie s and
Jack Russell Te ie s. Emmeans, 95 % CI, †p<0.1; * p <0.05.
Fig. 5. cap ion: The e ec o b eed and si ua ion in e ac ion on he dis ibu ion o in ensi y a ing o Fea in Bos on Te ie s and Jack Russell Te ie s. †p<0.1; *
p<0.05; *** p <0.001.
P. E e o ´
a e al.
Applied Animal Beha iou Science 292 (2025) 106767
8
Fig. 6. cap ion: The e ec o b eed and si ua ion in e ac ion on he dis ibu ion o in ensi y a ing o Sadness in Bos on Te ie s and Jack Russell Te ie s. †p<0.1; *
p<0.05; *** p <0.001.
Fig. 7. cap ion: The e ec o hypo he ical owne ship and si ua ion in e ac ion on he dis ibu ion o he in ensi y a ing o Sadness in bo h b eeds. Emmeans, 95 %
CI, †p<0.1; * p <0.05; ** p <0.01; *** p <0.001.
P. E e o ´
a e al.
Applied Animal Beha iou Science 292 (2025) 106767
9
