In e na ional Jou nal o
En i onmen al Resea ch
and Public Heal h
A icle
Academic Compe ence, Teache –S uden Rela ionship,
and Violence and Vic imisa ion in Adolescen s: The Class oom
Clima e as a Media o
Te esa I. Jiménez 1, Da id Mo eno-Ruiz 2,* , Es e anía Es é ez 3, Juan E a is o Callejas-Je ónimo 4,
Ginesa López-C espo 1and Sonsoles Valdi ia-Salas 1


Ci a ion: Jiménez, T.I; Mo eno-Ruiz,
D.; Es é ez, E.a; Callejas-Je ónimo,
J.E.; López-C espo, G.; Valdi ia-Salas,
S. Academic Compe ence,
Teache –S uden Rela ionship,
and Violence and Vic imisa ion in
Adolescen s: The Class oom Clima e
as a Media o . In . J. En i on. Res.
Public Heal h 2021,18, 1163.
h ps://doi.o g/10.3390/
ije ph18031163
Academic Edi o : Paul B. Tchounwou
Recei ed: 23 Decembe 2020
Accep ed: 25 Janua y 2021
Published: 28 Janua y 2021
Publishe ’s No e: MDPI s ays neu al
wi h ega d o ju isdic ional claims in
published maps and ins i u ional a il-
ia ions.
Copy igh : © 2021 by he au ho s.
Licensee MDPI, Basel, Swi ze land.
This a icle is an open access a icle
dis ibu ed unde he e ms and
condi ions o he C ea i e Commons
A ibu ion (CC BY) license (h ps://
c ea i ecommons.o g/licenses/by/
4.0/).
1Depa men o Psychology and Sociology, Uni e si y o Za agoza, 44003 Te uel, Spain;
ijimgu @uniza .es (T.I.J.); glopezc @uniza .es (G.L.-C.); sonso al@uniza .es (S.V.-S.)
2Depa men o Social Psychology, Uni e si y o Valencia, 46010 Valencia, Spain
3Depa men o Heal h Psychology, Miguel He nández Uni e si y, 03202 Alican e, Spain; [email p o ec ed]
4Depa men o Applied Economics I, Uni e si y o Se illa, 41018 Se illa, Spain; [email p o ec ed]
*Co espondence: da id.mo eno- uiz@u .es; Tel.: +34-96-382-81-21
Abs ac :
School iolence is a se ious social and public heal h p oblem p e alen wo ldwide. Al hough
he ele ance o eache and class oom ac o s is well es ablished in he li e a u e, ew s udies ha e
ocused on he ole o eache pe cep ions in school iolence and ic imisa ion and he po en ial
media ional ole o class oom clima e in his ela ionship. A o al o 2399 adolescen s (50% gi ls),
aged be ween 11 and 18 yea s (M = 14.65, SD = 1.78) and en olled in i e Spanish Seconda y Compul-
so y Educa ion schools comple ed measu es o class oom clima e, school iolence owa ds pee s and
pe cep ion o pee ic imisa ion, and hei eache s in o med abou hei academic compe ence and
he eache –s uden ela ionship. Co ela ional analyses e ealed ha whe eas academic compe ence
pe cei ed by he eache was nega i ely ela ed o o e iolence and ic imisa ion, i s ela ionship
wi h pu e ela ional iolence was posi i e. S uc u al equa ion modelling analyses showed ha
a iables o class oom clima e (in ol emen , a ilia ion, and eache suppo ) pe cei ed by he s u-
den s unc ioned as pa ial media o s be ween eache pe cep ions o academic compe ence and
o eache –s uden ela ionship and iolence and ic imisa ion. In he media ional model, eache
pe cep ion o academic compe ence ac ed as a di ec p o ec i e ac o agains iolence and ic im-
isa ion, and eache pe cep ion o eache –s uden ela ionship ac ed as a di ec isk o iolence,
as well as an indi ec p o ec i e ac o h ough class oom clima e o ic imisa ion. The in e p e a ion
o hese esul s poin s o he impo ance o he eache ’s subjec i e pe cep ions in he p e en ion
o iolence and ic imisa ion p oblems and hei p ac ical implica ions o he class oom clima e
pe cei ed by s uden s.
Keywo ds:
school iolence; school ic imisa ion; class oom clima e; academic compe ence; eache –
s uden ela ionship; adolescen s; media ion
1. In oduc ion
School iolence is a se ious social and public heal h p oblem p e alen wo ldwide [1–5].
This p oblem g ea ly wo ies pa en s [
6
], eache s [
7
], esea che s [
8
], and clinicians [
9
] due
o he signi ican and nega i e—some imes de as a ing—consequences o he physical
and psychological heal h o he s uden s in ol ed [
10
–
12
], wi h long- e m epe cussions
in he pe son’s in eg al de elopmen [
13
,
14
]. Among he di e en ac s comp ised in he
e m o school iolence (i.e., agg ession owa ds school s a , p ope y damage, andalism
on he campus, bullying), he cu en s udy ocused on iolence pe pe a ion owa ds
pee s, ha is, on nonacciden al beha iou s aimed a o he classma es o cause ha m
in en ionally; and on pee ic imisa ion, ha is, he pe cep ion o being exposed o nega i e
ac ions by one o mo e s uden s wi h he in en ion o hu ing he ic im [
15
]. These ac s
In . J. En i on. Res. Public Heal h 2021,18, 1163. h ps://doi.o g/10.3390/ije ph18031163 h ps://www.mdpi.com/jou nal/ije ph
In . J. En i on. Res. Public Heal h 2021,18, 1163 2 o 16
can mani es as physical (pushing, hi ing, kicking, sho ing, e c.), e bal (name-calling,
aun ing, h ea ening, e c.) o ela ional iolence (sp eading alse umou s o malicious
gossip, exclusion om ac i i ies, o wi hd awal o iendships). The i s wo o ms o
iolence– ic imisa ion ha e been labelled as “di ec o o e ” bullying, and he hi d
as “indi ec o ela ional” bullying. To unde s and he ele an nega i e consequences
o hese beha iou s ( o a e iew see [
16
]), i is impo an o no e ha he pe pe a o
usually ac s wi h he desi e and in en ion o domina e and o exe con ol o e he o he
pe son [
17
], c ea ing a dynamic o iolence om which he ic ims no mally canno de end
hemsel es o escape by hei own means, due o he imbalance o powe be ween he
people in ol ed [18,19].
Conside ing p e alence, in e na ional da a indica e g ea a ia ion among he s udies
ca ied ou in di e en coun ies. Fo example, in a s udy conduc ed wi h adolescen s om
43 coun ies (Eu ope, USA, and Canada) by he Wo ld Heal h O ganiza ion (WHO) [
20
],
he a es o bullying pe pe a ion a ied om 1 o 36%, and he a es o bullying ic-
imisa ion be ween 2 and 32%. Hymel and Swea e [
21
] a gued ha he a ia ion a e
depends on he coun y o o igin o he sample and he way ha school iolence was
assessed. In adolescen Spanish samples, Ga cía-Ga cía e al. [
22
] conduc ed a sys ema ic
e iew o he p e alence o school iolence in 32 s udies, which yielded a a e o 11.45% o
o e all bullying. Mo eo e , in c oss-cul u al su eys, a es o iolence pe pe a ion and
ic imisa ion a e consis en ly highe o boys han o gi ls [
23
]. Mo e speci ically, whe eas
ini ial esea ch ound ha boys pe pe a e and expe ience mo e o e iolence, and ha
indi ec o ms o iolence may be mo e ypical o gi ls [
24
,
25
], la e e idence sugges s ha
bo h sexes a e in ol ed in simila le els o ela ional iolence and ic imisa ion [26,27].
In he sea ch o causes o school iolence and ic imisa ion, many au ho s ha e
explo ed he pe sonal p o iles o agg esso s and ic ims and he cha ac e is ics o hei
amily, school, and communi y en i onmen s ( o a e iew om an ecological pe spec i e,
see [
28
]). Speci ically, wi hin he school con ex , mos s udies ha e ocused on he s uden ’s
poin o iew by analyzing he pe cep ion o low eache and pee suppo [
12
,
29
,
30
] o he
impo ance o g oup p ocesses such as he sea ch o a social epu a ion and pee accep-
ance [28,31]. Howe e , ew s udies ha e ocused on he ole ha eache s’ pe cep ions o
hei s uden s can play in he explana ion o school iolence and ic imisa ion. Speci ically,
in his s udy, we ocus on academic compe ence and he eache –s uden ela ionship
pe cei ed by he eache .
Conce ning academic compe ence, p e ious s udies in he ield ha e shown a small
bu signi ican nega i e co ela ion be ween pee ic imisa ion and academic achie emen
ac oss s udies ( o a me a-analy ic e iew, see [
32
]). In addi ion, a me a-analysis pe o med
by Sa age e al. [
33
] in o ms o he consis ency o a nega i e and mode a e ela ionship
be ween academic achie emen and iolen an isocial beha iou . Mo e speci ically, adi-
ional s udies on bullying ic imisa ion [
15
] ha e iewed agg ession owa ds pee s as a
eac ion o us a ions and ailu es a school. Howe e , la e s udies p o ided no e idence
o his hypo hesis. Fo example, Woods and Wolke [
34
] ound no ela ionship be ween
academic achie emen and di ec o ms o iolence pe pe a ion, and a posi i e and signi i-
can ela ionship wi h indi ec o ela ional iolence pe pe a ion. Howe e , mos o hese
s udies used academic achie emen (i.e., g ade poin a e ages) as an objec i e indica o o
academic compe ence and ew o hem ha e inco po a ed o he eache a ings o academic
unc ioning, such as subjec i e a ings o e o and in e es in class oom ac i i ies [
32
].
In his s udy, we a e in e es ed in ex ending p e ious li e a u e in his ield, conside ing a
b oade de ini ion o academic compe ence because i has been shown ha eache s’ subjec-
i e expec a ions abou he pe o mance o hei s uden s a ec hei in e ac ions wi h hem,
in luencing he class oom clima e as a whole [
35
]. Mo eo e , i is well es ablished ha he
quali y o school and class oom clima e is ela ed o gene al an isocial beha iou [
36
] and,
speci ically, o iolence and ic imisa ion p oblems [11,12].
Conce ning he eache s uden – ela ionship, Boucha d and Smi h [
37
] ha e ecen ly
a gued ha esea che s and educa o s should ecognize mo e ully he ole o eache –
In . J. En i on. Res. Public Heal h 2021,18, 1163 3 o 16
s uden ela ionship quali y in s uden s’ bullying– ic imisa ion expe iences. D awing om
he social–ecological sys em heo y, hese au ho s claimed he need o conside mesosys em
analyses—such us eache –s uden ela ionship quali y— o deeply unde s and school
iolence and ic imisa ion expe iences. Speci ically, some au ho s ha e a gued ha pee
in e ac ions in class ooms, including iolence and ic imisa ion, a e implici ly shaped
by he quali y o eache –s uden ela ionships because eache –s uden in e ac ions can
be seen as a ela ional model [
37
–
39
]. Indeed, p e ious empi ical e idence a es s o he
impo ance o eache –s uden ela ionships in psychosocial de elopmen ( o a e iew,
see [
40
]). Fo example, s uden s who epo be e ela ionships wi h eache s show highe
sel -es eem and p osocial beha iou , and ewe dep essi e and p oblem beha iou s such
as pee agg ession [
41
,
42
]. Con e sely, o he s udies ha e shown ha eache –s uden ela-
ionships ha a e dis an and con lic i e con ibu e o highe le els o school iolence and
ic imisa ion [
43
,
44
]. Resea ch has also e ealed ha low pe cei ed suppo om eache s
o pe cep ion o low eache in ol emen is ela ed o high school agg ession
[28,41,45].
Howe e , mos o he ci ed li e a u e abou he bene i s o eache –s uden ela ionships on
psychosocial adjus men has ocused on p ima y school and ew s udies ha e analysed
hese bene i s in seconda y schools, whe e he school s uc u e does no allow eache s o
ha e as many oppo uni ies o de elop signi ican and ca ing ela ionships wi h hei s u-
den s. Mo eo e , he exis ing li e a u e has almos exclusi ely conside ed eache –s uden
ela ionships om he s uden s’ poin o iew. This s udy aims o con ibu e o illing
his gap by analysing he link be ween he eache –s uden ela ionship pe cei ed by he
eache and iolence and ic imisa ion in seconda y school.
As no ed be o e, eache s a e one o he mos in luen ial agen s in es ablishing he
quali y o class oom clima e [
46
,
47
]. Tha is, eache s who engage in hei p ac ice and
show a ec i e a unemen and suppo owa d hei s uden s could con ibu e o gene -
a ing a class oom clima e whe e he s uden s pe cei e suppo om o he s, pa icipa e
in p oposed ac i i ies, and iew pee s as iends [
48
]. The e o e, we can expec ha he
eache ’s ac o s ha e an indi ec in luence on he le els o iolence and ic imisa ion
h ough hei con ibu ion o he class oom clima e in which pee s in e ac wi h each o he .
Indeed, some p e ious bu sca ce s udies suppo he idea ha class oom clima e could
se e as a media o be ween eache p ac ices and pee iolence [
49
]. Conside ing eache
pe cep ions, he o iginal esea ch conduc ed by Ha is and Rosen hal [
50
] showed ha
eache expec a ions abou hei s uden s we e ela ed o academic ou comes h ough
a class oom ac o such as a wa m and a ec i e clima e mo e han h ough he quali y
o dyadic in e ac ions. Subsequen s udies [
35
,
51
] ha e shown ha eache s wi h high
expec a ions o hei s uden s c ea ed a di e en ins uc ional and socioemo ional en i-
onmen , cha ac e ized by a be e s uc u ing o he eaching p ocess, mo e eedback,
a g ea e numbe o open ques ions and mo e posi i e and p e en i e managemen o
dis up i e beha iou in he class oom. Howe e , mos o he p e ious li e a u e ocused
on eache pe cep ions and class oom clima e ha e ocused on academic ou comes and no
on beha iou al ou comes such as school iolence and ic imisa ion.
In sum, conside ing ha (1) eache pe cep ions o academic compe ence and o
eache –s uden ela ionship a e ela ed o school iolence and ic imisa ion, (2) eache
pe cep ions abou hei s uden s a ec class oom clima e, and (3) class oom clima e is
ela ed o iolence and ic imisa ion among pee s, we migh hypo hesize ha class oom
clima e could be a mechanism linking eache pe cep ions and iolence and ic imisa ion.
We ha e no ound any p e ious s udies in he ield join ly analysing hese ac o s, s ill less
ocusing on he eache s’ poin o iew.
The P esen S udy
In acco dance wi h he p e ious li e a u e, in he p esen s udy, we seek o (1) explo e
he ela ionships be ween he s udy a iables (i.e., academic compe ence and eache –
s uden ela ionship pe cei ed by eache s, class oom clima e, school iolence, and school
ic imisa ion); (2) de e mine whe he class oom clima e media es he ela ionship be-
In . J. En i on. Res. Public Heal h 2021,18, 1163 4 o 16
ween eache pe cep ions o academic compe ence and o eache –s uden ela ionship,
and school iolence and ic imisa ion; and (3) examine he ole o sex as a po en ial mod-
e a o o hese ela ionships in a sample o Spanish adolescen s. We expec o ind he
media ing ole o class oom clima e and no sex di e ences in he pa e n o media ional
ela ionships among he s udy a iables.
2. Ma e ials and Me hods
2.1. P ocedu e
The da a we a e analysing in he p esen s udy we e collec ed as pa o a la ge
s udy on iolen beha iou in Spanish adolescen s. Ini ially, a le e wi h a summa y o
he esea ch p ojec was sen o he p incipal o each school o explain he pu pose o
he esea ch and o eques pe mission o ca y ou he s udy. The managemen eam o
each school eques ed he collabo a ion o hei eache s a . Then, we eques ed pa en s’
and gua dians’ consen o he s uden s o pa icipa e in his s udy (only 1% did no
gi e consen ). We explained he ele ance and he goals o he s udy o he s uden s and
in o med hem ha pa icipa ion was olun a y and anonymous. The adminis a ion o he
ins umen s was ca ied ou by a g oup o ained esea che s wi h a leas one quali ied
esea che (wi h a Ph.D.) p esen du ing he adminis a ion o he ins umen s o p o ide
s uden s wi h he necessa y suppo . Measu es we e collec ed in he class ooms du ing
a egula class pe iod wi hou he p esence o he eache s. In pa allel, he home oom
eache s ( he eache who is in cha ge o aking ca e o s uden s’ academic and pe sonal-
ela ed a ai s in a speci ic class oom du ing an academic cou se) illed in he PROF-A
scale. The esea ch was conduc ed in compliance wi h e hical s anda ds equi ed o
esea ch wi h human beings, espec ing he basic p inciples included in he Helsinki
Decla a ion and app o ed by he E hics Commi ee o he Vice-Chancello o Resea ch o
he hos uni e si y.
2.2. Pa icipan s
The s udy was an ex pos ac o c oss-sec ional and desc ip i e s udy. Pa icipan s
in he s udy we e 2399 adolescen s (50% gi ls), aged be ween 11 and 18 yea s (M= 14.65,
SD = 1.78), en olled in i e s a e schools o Seconda y Compulso y Educa ion and High
School loca ed in he Valencian Communi y (Spain). Sample dis ibu ion by academic
g ade, amily composi ion, and pa en s’ educa ional le el is p esen ed in Table 1.
2.3. Ins umen s
2.3.1. Academic Compe ence and Teache –S uden Rela ionship
To e alua e academic compe ence and eache –s uden ela ionship, we used wo o
he subscales o he School Adjus men Pe cei ed by he Teache Scale (Escala de Ajus e
Escola Pe cibido po el P o eso —PROF-A) de eloped by Ca a e al. [
52
]. This scale was
designed o assess he eache s’ pe cep ion o di e en ac o s ela ed o hei s uden s’
social adjus men and academic pe o mance by using 14 i ems. Academic compe ence
con ains ou i ems (e.g., “S uden ’s pa icipa ion in ac i i ies, discussions, deba es, e c.,
p oposed in class”) and he eache –s uden ela ionship subscale includes ou i ems
(e.g., “The ime I spend alking o his s uden ”). I ems a e a ed on a 10-poin scale anging
om 1 ( e y low) o 10 ( e y high). Adequa e psychome ic p ope ies we e ob ained in p e-
ious s udies pe o med wi h samples o seconda y s uden s [
52
,
53
], wi h C onbach’s alpha
alues anging om 0.85 o 0.92. In ou s udy, he C onbach’s alpha ob ained o academic
compe ence was 0.94 and 0.85 o he eache –s uden ela ionship. In he con i ma o y ac-
o ial analysis (CFA), he chi-squa e alues and i indexes we e
χ2(103) = 1137.31, p< 0.001,
RMSEA ( oo mean squa e e o o app oxima ion) = 0.06
[0.06, 0.07],
and CFI (compa a i e
i index) = 0.96.
In . J. En i on. Res. Public Heal h 2021,18, 1163 5 o 16
Table 1. Sociodemog aphic cha ac e is ics o he sample.
Sociodemog aphic Cha ac e is ics %
Academic G ade
1s g ade seconda y 21.4
2nd g ade seconda y 20.1
3 d g ade seconda y 18.2
4 h g ade seconda y 18.6
1s g ade high school 10.2
2nd g ade high school 11.5
Family Composi ion
In ac 83.5
Single pa en 15.7
Li ing wi h o he adul s 0.8
Mo he Educa ional Le el
Highe educa ion 35.9
High school 16.4
P ima y school 25.8
No s udies 8
Unknown 13.9
Fa he Educa ional Le el
Highe educa ion 37.8
High school 14.7
P ima y school 22
No s udies 8.4
Unknown 17.1
2.3.2. Class oom En i onmen
We used he ela ionship dimension o he class oom en i onmen scale [
54
] in he
Spanish e sion o Fe nández-Balles e os and Sie a [
55
]. This scale e alua es class oom
en i onmen om he poin o iew o he s uden s h ough 27 bina y-choice ( ue– alse)
i ems, o ming h ee subscales: (1) In ol emen (deg ee o s uden a en i eness, in e -
es , and pa icipa ion in class ac i i ies; 9 i ems, e.g., “S uden s pu a lo o ene gy in o
wha hey do he e”); (2) A ilia ion (deg ee o iendship and suppo among s uden s;
9 i ems, e.g., “S uden s in his class ge o know each o he eally well”); and (3) Teache
suppo (amoun o help, us , and iendship he eache o e s o s uden s; 9 i ems,
e.g., “The eache akes a pe sonal in e es in he s uden s”). The C onbach’s alpha al-
ues ob ained in p e ious s udies wi h Spanish adolescen samples anged be ween 0.77
and 0.89
(e.g., [41,56]).
In ou sample, C onbach’s alpha o he global scale was 0.77 and
0.66, and 0.64 and 0.63 o he h ee subscales, espec i ely. In he CFA analysis, he chi-
squa e alues and i indexes we e χ2(368) = 1075.66, p< 0.001, RMSEA = 0.03 [0.03, 0.04],
and CFI = 0.91.
2.3.3. School Violence
This scale by Li le e al. [
57
] (bidi ec ional ansla ion in o Spanish, using he pa allel
back- ansla ion p ocedu e o B islin [
58
]) includes 25 i ems ha assess pa icipa ion in
agg essi e beha iou owa ds pee s a school o e he las 12 mon hs. I is a ed on a 4-poin
Like - ype scale anging om 1 (ne e ) o 4 (always). The scale e alua es six dimensions o
iolence wi hin wo ypes o agg essi e beha iou : o e o di ec and ela ional o indi ec ;
and h ee unc ions o iolence: pu e, eac i e, and ins umen al. These dimensions a e
pu e o e (e.g., “I’m he kind o pe son who hi s, kicks, o punches o he s”); eac i e
o e (e.g., “I o he s make me upse o hu me, I o en pu hem down”); ins umen al
o e (e.g., “I o en h ea en o he s o ge wha I wan ”); pu e ela ional (e.g., “I’m he
kind o pe son who says mean hings abou o he s”); eac i e ela ional (e.g., “I o he s
ha e h ea ened me, I o en say mean hings abou hem”); and ins umen al ela ional
(e.g., “To ge wha I wan , I o en igno e o s op alking o o he s”). In he p esen s udy,

In . J. En i on. Res. Public Heal h 2021,18, 1163 6 o 16
all i ems we e summed o gene a e six composi e a iables ha ed he school iolence
cons uc in he measu emen model. The C onbach’s alpha alues ob ained in p e ious
s udies wi h Spanish adolescen samples anged be ween 0.64 and 0.87 [
12
,
56
,
59
]. In ou
sample, C onbach’s alpha o he global scale using hese composi es was 0.85 ( he eliabili y
ange o he six subscales a ied om 0.63 in ins umen al ela ional o 0.77 in eac i e
o e ). In he CFA analysis, he chi-squa e alues and i indexes we e
χ2
(178) = 641.02,
p< 0.001, RMSEA = 0.03 [0.03, 0.04], and CFI = 0.91.
2.3.4. School Vic imisa ion
The pee – ic imisa ion scale de eloped by Ca a e al. [
60
] is a measu e o sel - epo ed
ic imisa ion based on he mul idimensional pee – ic imisa ion scale o Myna d and
Joseph [
61
] and he social expe ience ques ionnai e sel - epo o C ick and G o pe e [
25
].
This scale consis s o 20 i ems ha assess he equency wi h which he s uden s had
expe ienced 20 ic imizing expe iences in he las school yea on a esponse ange o
1 (ne e ) o 4 (always). The scale p esen s a h ee- ac o s uc u e: ela ional ic imi-
sa ion (e.g., “a pee go ang y wi h me and sepa a ed me om my g oup o iends o
p e en me om playing o pa icipa ing in any ac i i y”); o e physical ic imisa ion
(e.g., “a pee hi me o eally ha m me”); and o e e bal ic imisa ion (e.g., “a pee
insul ed me”). The psychome ic p ope ies o he scale a e adequa e in p e ious s ud-
ies [
60
,
62
] and in he p esen s udy, wi h in e nal consis encies (C onbach
α
) o he h ee
subscales anging om 0.68 o 0.92 (0.92 in he p esen s udy o he o al scale and 0.89, 0.63,
and 0.84 o he h ee subscales, espec i ely). In he CFA analysis, he chi-squa e alue
and i indexes we e
χ2
(159) = 664.28, p< 0.001, RMSEA = 0.04 [0.03, 0.04],
and CFI = 0.93.
2.4. S a is ical Analysis
Fi s ly, desc ip i e s a is ics (means and s anda d de ia ions) and Pea son’s co e-
la ions we e compu ed using SPSS (Ve sion 25.0, IBM, A monk, NY, USA). Secondly,
mul i a ia e in e en ial analyses we e conduc ed using s uc u al equa ion modelling
(SEM) h ough Mplus 8.4 so wa e [
63
]. Complemen a y analyses we e also pe o med
o de e mine he signi icance and magni ude o he po en ial media ing e ec [
64
,
65
].
Finally, o u he check he obus ness o he p oposed model, we es ed i using s uc u al
in a iance ac oss sex h ough mul ig oup analyses.
The maximum likelihood es ima e wi h obus s anda d e o s (MLR) was used o
co ec he non-no mali y o he da a [
63
]. The ixed models we e e alua ed using a
combina ion o indices, including he compa a i e i index se and he Tucke –Lewis index
(CFI/TLI > 0.90 o a easonable i and CFI/TLI > 0.95 o a good i ), he mean squa e oo
o he esiduals (SRMR < 0.09 o a good i ), and he mean squa e oo o he app oxima ion
(RMSEA < 0.06 o a p ope i ) [
66
]. Al hough a chi-squa e es (S-B
χ2
) o model i was
also epo ed, i was no used o e alua e model i due o i s sensi i i y o la ge sample
sizes. Th oughou he whole s udy, he epo ed pa h coe icien s we e s anda dised alues.
To pe o m he es o he indi ec e ec , i s con idence in e als we e calcula ed using he
boo s ap me hod wi h 2000 samples.
3. Resul s
3.1. Desc ip i e Analysis
Rega ding he desc ip i e s a is ics and co ela ions o he obse ed a iables,
Table 2
shows he means, s anda d de ia ions, and bi a ia e co ela ions. In gene al, a iables o
school adjus men pe cei ed by he eache (academic compe ence and eache –s uden
ela ionship) we e posi i ely and signi ican ly ela ed o a iables o class oom clima e
(in ol emen , a ilia ion, and eache suppo pe cei ed by he s uden s), wi h anging
om 0.063 o 0.116, p< 0.01. Academic compe ence and a iables o school iolence and
ic imisa ion we e nega i ely and signi ican ly ela ed ( anging om
−
0.040, p< 0.05,
o ins umen al ela ional iolence o
−
0.239, p< 0.01, o eac i e o e iolence) excep
o pu e ela ional iolence ( his bi a ia e co ela ion was posi i e wi h = 0.045, p< 0.05).
In . J. En i on. Res. Public Heal h 2021,18, 1163 7 o 16
Teache –s uden ela ionship was nega i ely and signi ican ly ela ed o o e eac i e
iolence ( =
−
0.085, p< 0.01) and o e physical ic imisa ion ( =
−
0.047, p< 0.05), bu his
ela ionship was posi i e wi h pu e iolence (bo h o e and ela ional; = 0.043, p< 0.05
and = 0.077, p< 0.01, espec i ely). All bi a ia e co ela ions be ween class oom clima e
and iolence and ic imisa ion a iables we e nega i e and signi ican ( anging om
−0.051, p< 0.05, o −0.232, p< 0.01).
3.2. S uc u al Equa ion Modelling
To analyse he di ec ela ionship be ween school adjus men and school iolence and
ic imisa ion, and indi ec ly h ough class oom clima e, a s uc u al equa ion model (SEM)
was c ea ed. Table 3shows he la en a iables included in he models, hei espec i e
indica o s, he s anda d e o , and he associa ed p obabili y o each indica o in he
co esponding la en a iable.
The calcula ed model (see Figu e 1) showed an adequa e i o da a: S-B
χ2
= 461.304,
d = 45, p< 0.001, CFI = 0.952, RMSEA = 0.062 [0.057, 0.067]. A di ec and nega i e
ela ionship was obse ed be ween eache pe cep ion o academic compe ence and
school iolence (
β
=
−
0.157, p< 0.001), and be ween eache pe cep ion o academic
compe ence and school ic imisa ion (
β
=
−
0.114, p< 0.001). The di ec ela ionship
was posi i e, al hough smalle , be ween eache pe cep ion o eache –s uden ela ion-
ship and school iolence (
β
= 0.084, p< 0.01) and ic imisa ion (
β
= 0.052, p< 0.05).
Mo eo e , he h ee class oom clima e a iables (in ol emen , a ilia ion, and eache
suppo ) we e di ec ly and nega i ely ela ed o school iolence
(β=−0.069, p< 0.01;
β=−0.070,
p< 0.01;
β
=
−
0.094, p< 0.001), and wo o hem (in ol emen and a ilia ion)
o school ic imisa ion (
β
=
−
0.129 and
β
=
−
0.199, p< 0.001). The pe cen age o a iabili y
o school iolence and ic imisa ion explained by he s uc u al model was 15.1% (9.6% o
ic imisa ion and 5.5% o iolence), which can be conside ed o be a size o he e ec o
he s a is ical signi icance o he es ima ed model.
In . J. En i on. Res. Public Heal h 2021, 18, x 7 o 17
om 0.063 o 0.116, p < 0.01. Academic compe ence and a iables o school iolence and
ic imisa ion we e nega i ely and signi ican ly ela ed ( anging om –0.040, p < 0.05, o
ins umen al ela ional iolence o −0.239, p < 0.01, o eac i e o e iolence) excep o
pu e ela ional iolence ( his bi a ia e co ela ion was posi i e wi h = 0.045, p < 0.05).
Teache –s uden ela ionship was nega i ely and signi ican ly ela ed o o e eac i e
iolence ( = −0.085, p < 0.01) and o e physical ic imisa ion ( = −0.047, p < 0.05), bu his
ela ionship was posi i e wi h pu e iolence (bo h o e and ela ional; = 0.043, p < 0.05
and = 0.077, p < 0.01, espec i ely). All bi a ia e co ela ions be ween class oom clima e
and iolence and ic imisa ion a iables we e nega i e and signi ican ( anging om –
0.051, p < 0.05, o −0.232, p < 0.01).
3.2. S uc u al Equa ion Modelling
To analyse he di ec ela ionship be ween school adjus men and school iolence
and ic imisa ion, and indi ec ly h ough class oom clima e, a s uc u al equa ion model
(SEM) was c ea ed. Table 3 shows he la en a iables included in he models, hei e-
spec i e indica o s, he s anda d e o , and he associa ed p obabili y o each indica o
in he co esponding la en a iable.
The calcula ed model (see Figu e 1) showed an adequa e i o da a: S-B χ
2
= 461.304,
d = 45, p < 0.001, CFI = 0.952, RMSEA = 0.062 [0.057, 0.067]. A di ec and nega i e ela ion-
ship was obse ed be ween eache pe cep ion o academic compe ence and school io-
lence (β = −0.157, p < 0.001), and be ween eache pe cep ion o academic compe ence and
school ic imisa ion (β = −0.114, p < 0.001). The di ec ela ionship was posi i e, al hough
smalle , be ween eache pe cep ion o eache –s uden ela ionship and school iolence
(β = 0.084, p < 0.01) and ic imisa ion (β = 0.052, p < 0.05). Mo eo e , he h ee class oom
clima e a iables (in ol emen , a ilia ion, and eache suppo ) we e di ec ly and nega-
i ely ela ed o school iolence (β = −0.069, p < 0.01; β = −0.070, p < 0.01; β = −0.094, p <
0.001), and wo o hem (in ol emen and affilia ion) o school ic imisa ion (β = −0.129
and β = −0.199, p < 0.001). The pe cen age o a iabili y o school iolence and ic imisa ion
explained by he s uc u al model was 15.1% (9.6% o ic imisa ion and 5.5% o io-
lence), which can be conside ed o be a size o he e ec o he s a is ical signi icance o
he es ima ed model.
Figu e 1. Final s uc u al model wi h pa h coe icien s and s a is ical signi icance. *** p < 0.001; ** p < 0.01; * p < 0.05.
Figu e 1. Final s uc u al model wi h pa h coe icien s and s a is ical signi icance. *** p< 0.001; ** p< 0.01; * p< 0.05.
In . J. En i on. Res. Public Heal h 2021,18, 1163 8 o 16
Table 2. Pea son’s co ela ions and desc ip i e s a is ics o obse ed a iables.
1 2 3 4 5 6 7 8 9 10 11 12 13 14
1. Academic compe ence
-
2. Teache –s uden
ela ionship 0.468 ** -
3. In ol emen 0.028 0.116 ** -
4. A ilia ion 0.079 ** 0.086 ** 0.310 ** -
5. Teache suppo 0.078 ** 0.063 ** 0.334 ** 0.286 ** -
6. Pu e o e iolence −0.067 ** 0.043 * −0.097 ** −0.105 ** −0.149 ** -
7. Reac i e o e
iolence −0.239 ** −0.085 ** −0.125 ** −0.102 ** −0.118 ** 0.460 ** -
8. Ins umen al o e
iolence −0.090 ** 0.012 −0.067 ** −0.119 ** −0.127 ** 0.504 ** 0.398 ** -
9. Pu e ela ional
iolence 0.045 * 0.077 ** −0.051 * −0.086 ** −0.119 ** 0.417 ** 0.205 ** 0.412 ** -
10. Reac i e ela ional
iolence 0.021 0.030 −0.110 ** −0.083 ** −0.119 ** 0.288 ** 0.309 ** 0.302 ** 0.438 ** -
11. Ins umen al
ela ional iolence −0.040 * 0.004 −0.072 ** −0.106 ** −0.097 ** 0.327 ** 0.257 ** 0.603 ** 0.456 ** 0.401 ** -
12. Rela ional
ic imisa ion −0.066** −0.003 −0.173 ** −0.232 ** −0.125 ** 0.139 ** 0.039 0.160 ** 0.176 ** 0.164 ** 0.186 ** -
13. O e physical
ic imisa ion −0.123** −0.047 * −0.108 ** −0.160 ** −0.093 ** 0.144 ** 0.146 ** 0.139 ** 0.091 ** 0.017 0.092 ** 0.496 ** -
14. O e e bal
ic imisa ion −0.086** −0.039 −0.197 ** −0.222 ** −0.148 ** 0.203 ** 0.117 ** 0.174 ** 0.155 ** 0.117 ** 0.148 ** 0.752 ** 0.604 ** -
Mean 5.927 7.388 1.455 1.717 1.597 1.366 1.578 1.133 1.332 1.778 1.192 1.471 1.180 1.593
S anda d de ia ion 2.028 1.313 0.208 0.173 0.220 0.322 0.542 0.243 0.323 0.470 0.305 0.439 0.284 0.470
No e: ** p< 0.01; * p< 0.05.
In . J. En i on. Res. Public Heal h 2021,18, 1163 9 o 16
Table 3. Fac o ial loadings, s anda d e o , and associa ed p obabili y.
Fac o s and Va iables Fac o Loadings
(S anda d E o s)
School Violence
Pu e o e iolence 1 a
Reac i e o e iolence 1.834 ***
(0.432)
Ins umen al o e iolence 0.656 ***
(0.073)
Pu e ela ional iolence 0.750 ***
(0.081)
Reac i e ela ional iolence 0.861 ***
(0.101)
Ins umen al ela ional iolence 0.811 ***
(0.087)
School Vic imisa ion
Rela ional ic imisa ion 1 a
O e physical ic imisa ion 0.883 ***
(0.068)
O e e bal ic imisa ion 0.958 ***
(0.068)
No e: Robus s a is ics. S anda d e o s in b acke s. aFixed a 1 du ing es ima ion. *** p< 0.001.
Rega ding he indi ec ela ionships o media ional e ec s, we calcula ed all o hem
and he esul s showed mul iple signi ican indi ec e ec s o academic compe ence and
eache –s uden ela ionship on school iolence and ic imisa ion h ough he h ee a i-
ables o class oom clima e. Thei magni ude anged om
−
0.001 (e.g., Academic compe-
ence
→
In ol emen
→
Teache suppo
→
School iolence) o
−
0.014 (Teache –s uden
ela ionship
→
In ol emen
→
School ic imisa ion); ha is, all o hem we e e y small.
To al indi ec e ec s, di ec e ec s, and o al e ec s a e p esen ed in Table 4. O e all,
he o al e ec o eache pe cep ion o academic compe ence on school iolence and ic im-
iza ion was nega i e (i.e., p o ec i e, highe academic compe ence was associa ed o lowe
iolence and ic imiza ion) and highe han o al e ec o eache pe cep ion o eache –
s uden ela ionship (in his case o al e ec s we e posi i e, i.e., highe eache –s uden
ela ionship was associa ed o highe iolence and ic imiza ion). Conce ning he e ec o
eache pe cep ion o academic compe ence, bo h di ec and indi ec e ec s we e nega i e
and, in consequence, he o al e ec s we e highe and signi ican (β=−0.164 o iolence
and
β
=
−
0.122 o ic imiza ion, p< 0.001). Rega ding he e ec o eache pe cep ion o
eache –s uden ela ionship bo h, di ec and indi ec e ec s, we e smalle and opposi e
(a isk i di ec and a p o ec ion i indi ec ) and, in consequence, o al e ec s we e lowe
and only signi ican , as a isk, o school iolence (β= 0.067, p< 0.05).
Finally, o check he obus ness o his model, we es ed i s s uc u al in a iance
ac oss sex g oups (boys and gi ls) h ough a mul ig oup analysis. Two models we e es ed:
in he un es ic ed model, pa ame e es ima es ( ac o loadings and s uc u al pa hs) we e
eely es ima ed ac oss g oups; in he es ic ed model, we cons ained each o he ac o
loadings and he s uc u al pa hs o be in a ian ac oss g oups. I he chi-squa e o he
es ic ed model was signi ican ly la ge han he chi-squa e o he un es ic ed model,
he assump ion o in a iance would no be enable. The es ic ed model showed a i o
S-B
χ2
= 939.430, d = 139, p< 0.001, CFI = 0.907, RMSEA = 0.69 [0.065, 0.074]. The esul s
showed signi ican di e ences be ween boys and gi ls in he model (
∆χ2(35, 2399)
= 103.432,
p< 0.001). Obse ing he analyses and using index modi ica ion, h ee pa hs p esen ed
signi ican di e ences be ween he sex g oups: i s , he ela ionship be ween eache
suppo and school iolence was nega i e and signi ican in gi ls (
β
=
−
0.148, p< 0.001) bu
highe in boys (
β
=
−
0.165, p< 0.001); second, he loading o ins umen al o e iolence
in school iolence was posi i e and signi ican in gi ls (
β
=
−
0.448, p< 0.001) bu highe
in boys (
β
=
−
0.585, p< 0.001); and inally, he ela ionship be ween a ilia ion and school
In . J. En i on. Res. Public Heal h 2021,18, 1163 16 o 16
47.
Runions, K.C.; Shaw, T. Teache –child ela ionship, child wi hd awal and agg ession in he de elopmen o pee ic imiza ion.
J. Appl. De . Psychol. 2013,34, 319–327. [C ossRe ]
48.
Ju onen, J. Sense o belonging, social bonds, and school unc ioning. In Handbook o Educa ional Psychology; Alexande , P.A.,
Winne, P.H., Eds.; Law ence E lbaum Associa es Publishe s: Mahwah, NJ, USA, 2006; pp. 655–674.
49.
Valdés-Cue o, Á.A.; Ma ínez-Fe e , B.; Ca los-Ma ínez, E.A. The ole o eaching p ac ices in he p e en ion o school iolence
among pee s. J. Psychodidac ics 2018,23, 33–38. [C ossRe ]
50.
Ha is, M.J.; Rosen hal, R. Media ion o in e pe sonal expec ancy e ec s: 31 me a-analyses. Psychol. Bull.
1985
,97, 363–386.
[C ossRe ]
51.
Rubie-Da ies, C. Expec ing he Bes : Ins uc ional P ac ices, Teache Belie s and S uden Ou comes. Ph.D. Thesis, Uni e si y o
Auckland, Auckland, New Zealand, 2004.
52.
Ca a, M.J.; Po edano, A.; Buelga, S.; Musi u, G. Psychome ic analysis o he Scale o Teache ’s Pe cep ion o School Adjus men
(PROF-A). In e . Psicosoc. 2015,24, 63–69. [C ossRe ]
53.
Cañas, E.; Es é ez, E.; León-Mo eno, C.; Musi u, G. Loneliness, amily communica ion, and school adjus men in a sample o
cybe ic imized adolescen s. In . J. En i on. Res. Public Heal h 2020,17, 335. [C ossRe ]
54. Moos, R.H.; T icke , E.J. Class oom En i onmen Scale Manual; Consul ing Psychologis P ess: Palo Al o, CA, USA, 1973.
55. Fe nández-Balles e os, R.; Sie a, B. Escalas de Clima Social FES, WES, CIES y CES; TEA: Mad id, Spain, 1989.
56.
Jiménez, T.I.; Es é ez, E. School agg ession in adolescence: Examining he ole o indi idual, amily and school a iables. In . J.
Clin. Heal h Psychol. 2017,17, 251–260. [C ossRe ]
57.
Li le, T.D.; Hen ich, C.C.; Jones, S.M.; Hawley, P.H. Disen angling he “whys” om he “wha s” o agg essi e beha iou . In . J.
Beha . De . 2003,27, 122–131. [C ossRe ]
58.
B islin, R.W. The wo ding and ansla ion o esea ch ins umen . In Field Me hods in C oss-Cul u al Resea ch; Lonne , W.J., Be y,
J.W., Eds.; Sage: Newbu y Pa k, CA, USA, 1986; pp. 137–164.
59.
Mo eno-Ruiz, D.; Es é ez, E.; Jiménez, T.I.; Mu gui, S. Pa en ing s yle and eac i e and p oac i e adolescen iolence: E idence
om Spain. In . J. En i on. Res. Public Heal h 2018,15, 2634. [C ossRe ]
60.
Ca a, M.J.; Musi u, G.; Mu gui, S. Indi idual and social isk ac o s ela ed o o e ic imiza ion in a sample o Spanish
adolescen s. Psychol. Rep. 2007,101, 275–290. [C ossRe ]
61.
Myna d, H.; Joseph, S. De elopmen o he Mul idimensional Pee -Vic imiza ion Scale. Agg ess. Beha .
2000
,26, 169–178.
[C ossRe ]
62.
Es é ez, E.; Es é ez, J.F.; Segu a, L.; Suá ez, C. The in luence o bullying and cybe bullying in he psychological adjus men o
ic ims and agg esso s in adolescence. In . J. En i on. Res. Public Heal h 2019,16, 2080. [C ossRe ] [PubMed]
63. Mu hén, L.K.; Mu hén, B.O. Mplus Use ’s Guide, 8 h ed.; Mu hén & Mu hén: Los Angeles, CA, USA, 2017.
64. Mackinnon, D.P.; Dwye , J.H. Es ima ing media ed e ec s in p e en ion s udies. E al. Re . 1993,17, 144–158. [C ossRe ]
65. Mackinnon, D.; Fai child, A.; F i z, M. Media ion Analysis. Annu. Re . Psychol. 2007,58, 593–614. [C ossRe ] [PubMed]
66. Kline, R.B. P inciples and P ac ice o S uc u al Equa ion Modeling; The Guil o d P ess: New Yo k, NY, USA, 2011.
67.
Rubie-Da ies, C.M.; Pe e son, E.R.; Sibley, C.G.; Rosen hal, R. A eache expec a ion in e en ion: Modelling he p ac ices o high
expec a ion eache s. Con emp. Educ. Psychol. 2015,40, 72–85. [C ossRe ]
68.
Wolke, D.; Woods, S.; S an o d, K.; Schulz, H. Bullying and ic imiza ion o p ima y school child en in England and Ge many:
P e alence and school ac o s. B . J. Psychol. 2001,92, 673–696. [C ossRe ]
69.
McMillan, J.H.; My an, S.; Wo kman, D. Elemen a y eache s’ class oom assessmen and g ading p ac ices. J. Educ. Res.
2002
,95,
203–213. [C ossRe ]
70.
Pe e son, E.R.; Rubie-Da ies, C.; Osbo ne, D.; Sibley, C. Teache s’ explici expec a ions and implici p ejudiced a i udes o
educa ional achie emen : Rela ions wi h s uden achie emen and he e hnic achie emen gap. Lea n. Ins uc .
2016
,42, 123–140.
[C ossRe ]
71.
Rubie-Da ies, C.M.; Rosen hal, R. In e ening in eache s’ expec a ions: A andom e ec s me a-analy ic app oach o examining
he e ec i eness o an in e en ion. Lea n. Indi id. Di e . 2016,50, 83–92. [C ossRe ]
72.
Albe o, P.; T ou man, A.C. Applied Beha io Analysis o Teache s, 9 h ed.; Pea son Educa ion: Uppe Saddle Ri e , NJ, USA, 2012.
73.
Cañas, E.; Es é ez, E.; Ma inez-Mon eagudo, M.C.; Delgado, B. Emo ional adjus men in ic ims and pe pe a o s o cybe bully-
ing and adi ional bullying. Soc. Psychol. Educ. 2020,20, E76. [C ossRe ]
74.
T oop-Go don, W.; Kopp, J. Teache –child ela ionship quali y and child en’s pee ic imiza ion and agg essi e beha io in la e
childhood. Soc. De . 2011,20, 536–561. [C ossRe ]
75.
C ick, N.R.; Nelson, D.A. Rela ional and physical ic imiza ion wi hin iendships: Nobody old me he e’d be iends like hese.
J. Abno m. Child Psychol. 2002,30, 599–607. [C ossRe ] [PubMed]
76.
Bouman, T.; an de Meulen, M.; Goossens, F.A.; Ol ho , T.; Ve mande, M.M.; Ale a, E.A. Pee and sel - epo s o ic imiza ion
and bullying: Thei di e en ial associa ion wi h in e nalizing p oblems and social adjus men . J. Sch. Psychol.
2012
,50, 759–774.
[C ossRe ] [PubMed]
77.
Maxwell, S.E.; Cole, D.A.; Mi chell, M.A. Bias in c oss-sec ional analyses o longi udinal media ion: Pa ial and comple e
media ion unde an au o eg essi e model. Mul i a . Beha . Res. 2011,46, 816–841. [C ossRe ] [PubMed]