Reciprocal associations among teacher-child interactions, teachers' work engagement, and children's social competence
Full text
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/
Recip ocal associa ions among eache -child in e ac ions, eache s' wo k engagemen ,
and child en's social compe ence
© 2022 The Au ho s. Published by Else ie Inc.
Published e sion
Soininen, Viola; Paka inen, Eija; Le kkanen, Ma ja-K is iina
Soininen, V., Paka inen, E., & Le kkanen, M.-K. (2023). Recip ocal associa ions among eache -
child in e ac ions, eache s' wo k engagemen , and child en's social compe ence. Jou nal o
Applied De elopmen al Psychology, 85, A icle 101508.
h ps://doi.o g/10.1016/j.appde .2022.101508
2023
Jou nal o Applied De elopmen al Psychology 85 (2023) 101508
A ailable online 4 Janua y 2023
0193-3973/© 2022 The Au ho s. Published by Else ie Inc. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/).
Recip ocal associa ions among eache –child in e ac ions, eache s’ wo k
engagemen , and child en’s social compe ence
Viola Soininen
*
, Eija Paka inen , Ma ja-K is iina Le kkanen
Uni e si y o Jy ¨
askyl¨
a, 40014 Uni e si y o Jy ¨
askyl¨
a, P. O. Box 35, Finland
ARTICLE INFO
Keywo ds:
Teache –child in e ac ion
Wo k engagemen
Social compe ence
Fi s g ade
ABSTRACT
This s udy aimed a examining ecip ocal associa ions among eache –child in e ac ions, eache s’ wo k
engagemen , and child en’s social compe ence ac oss g ade 1. The pa icipan s we e 51 Finnish eache s and 815
child en. The quali y o eache –child in e ac ions (emo ional suppo , class oom o ganiza ion, and ins uc ional
suppo ) was assessed wi h he Class oom Assessmen Sco ing Sys em. In addi ion, eache s epo ed hei wo k
engagemen and a ed child en’s social compe ence (p osocial and an isocial beha io s). The esul s o he
mul ile el modeling indica ed ha high-quali y ins uc ional suppo was associa ed wi h mo e p osocial and less
an isocial beha io . P osocial beha io was associa ed wi h highe -quali y ins uc ional suppo . Wo k engage-
men was associa ed wi h less an isocial beha io , and p osocial beha io wi h highe wo k engagemen . Finally,
wo k engagemen was associa ed wi h highe -quali y eache –child in e ac ions. The esul s sugges ha eache s
can enhance child en’s social compe ence wi h high-quali y ins uc ional suppo . Mo eo e , child en’s p osocial
beha io can be e lec ed in eache –child in e ac ions and in eache s’ wo k engagemen .
In hei p osocial class oom model, Jennings and G eenbe g (2009)
p oposed ha eache s’ social and emo ional compe ence and well-
being enables hem o c ea e heal hy eache –child ela ionships,
conduc e ec i e class oom managemen , and implemen social-
emo ional lea ning in he class oom. This way, hey no ed, a eache
c ea es a posi i e class oom clima e ha enhances child en’s social,
emo ional, and academic ou comes. Thus, acco ding o hei p osocial
class oom model, eache s’ occupa ional well-being is e lec ed in hei
in e ac ions wi h child en, which again is e lec ed in child en’s ou -
comes, including hei social compe ence. Based on his model, i is
possible ha eache s’ wo k engagemen as pa o hei occupa ional
well-being is associa ed wi h he quali y o eache –child in e ac ions
and he a e age social compe ence o he child en in he class oom.
Howe e , ea lie esea ch on he associa ions be ween eache s’ wo k
engagemen and he quali y o eache –child in e ac ions has been
sca ce and c oss-sec ional, and hus has no been able o examine di-
ec ions o associa ions (e.g., Pen inen, Paka inen, on Suchodole z, &
Le kkanen, 2020). Di ec ions o he associa ions be ween he quali y o
eache –child in e ac ions and child en’s social compe ence also emain
somewha unclea because mos o he ea lie s udies ha e only exam-
ined i he quali y o eache -child in e ac ions enhances child en’s so-
cial compe ence (e.g., Siekkinen e al., 2013). Recen esea ch indica es
ha he associa ions migh be ecip ocal – a leas be ween emo ional
suppo and child en’s empa hy (Paka inen, Le kkanen, & on Sucho-
dole z, 2020). Fu he mo e, al hough ea lie esea ch has shown ha
eache s’ s ess is associa ed wi h child en’s lowe social compe ence
(He man, Hickmon-Rosa, & Reinke, 2018; Siekkinen e al., 2013), o
da e, i is no ye clea whe he wo k engagemen as a posi i e aspec o
occupa ional well-being plays a ole in child en’s social compe ence.
Al oge he , he e is a limi ed amoun o knowledge o he possible as-
socia ions among he h ee domains o eache –child in e ac ions
(emo ional suppo , class oom o ganiza ion, and ins uc ional suppo ),
eache s’ wo k engagemen , and a e age social compe ence (p osocial
and an isocial beha io ) o child en in he class ooms. To be e un-
de s and how o suppo he de elopmen o child en’s social compe-
ence, his s udy aimed o explo e ecip ocal associa ions among he
quali y o eache –child in e ac ions, eache s’ wo k engagemen , and
child en’s social compe ence o e he cou se o i s school yea .
The quali y o eache –child in e ac ions
In he p esen s udy, he quali y o eache –child in e ac ions was
concep ualized wi h he Teaching h ough In e ac ions (TTI) amewo k
de eloped by Ham e e al. (2013). The ocus o he TTI amewo k and,
* Co esponding au ho a : Depa men o Teache educa ion, Uni e si y o Jy ¨
askyl¨
a, P.O. Box 35, 40014 Uni e si y o Jy ¨
askyl¨
a, Finland.
E-mail add ess: [email p o ec ed] (V. Soininen).
Con en s lis s a ailable a ScienceDi ec
Jou nal o Applied De elopmen al Psychology
jou nal homepage: www.else ie .com/loca e/jappdp
h ps://doi.o g/10.1016/j.appde .2022.101508
Recei ed 19 Janua y 2022; Recei ed in e ised o m 12 Decembe 2022; Accep ed 28 Decembe 2022
Jou nal o Applied De elopmen al Psychology 85 (2023) 101508
2
hence, he p esen s udy, is a eache -child in e ac ions a he
class oom-le el, no a dyadic eache -child ela ionships (i.e., e-
la ionships be ween eache and indi idual child en). The TTI ame-
wo k is based on B on enb enne & Mo is’s (2006) bioecological
model, which sugges s ha egula in e ac ions ha child en ha e, o
example, wi h pa en s and eache s, a e a cen al d i e o hei
de elopmen (Ham e e al., 2013). Indeed, esea ch has indica ed ha
he quali y o eache -child in e ac ions in he class oom is a c i ical
indica o o he quali y o educa ion (Pian a, Downe , & Ham e, 2016,
see also Bu chinal, Vande g i , Pian a, & Mashbu n, 2010; Hu, Fan, Wu,
LoCasale-C ouch, & Song, 2019). In TTI, in e ac ions in he class oom
a e concep ualized unde h ee domains: emo ional suppo , class oom
o ganiza ion, and ins uc ional suppo . The quali y o hese h ee do-
mains can be assessed wi h an obse a ional ool, he Class oom
Assessmen Sco ing Sys em (CLASS; Pian a, La Pa o, & Ham e, 2008).
The domain o emo ional suppo has i s oo s in a achmen heo y
(Ainswo h, Beleha , Wa e s, & Wall, 1978; Bowlby, 1969) and sel -
de e mina ion heo y (Ryan & Deci, 2000; Skinne & Belmon , 1993),
which unde line he impo ance o sensi i e and esponsi e in e ac ions
in suppo ing lea ning and mo i a ion. Acco ding o he TTI amewo k,
emo ionally suppo i e eache s can c ea e and main ain a posi i e
clima e in hei class oom, hey a e sensi i e o child en’s indi idual
needs, and hey conside child en’s pe spec i es. In hese class ooms,
child en enjoy being wi h each o he and wi h he eache s; hey
app oach he eache eely and exp ess hei ideas. The second domain,
class oom o ganiza ion, is g ounded on ea lie class oom managemen
(see Emme & S ough, 2001) and sel - egula ion skill esea ch (e.g.,
Pa is & Pa is, 2001; Ra e , 2004; Rimm-Kau man e al., 2002). Teache s
wi h high-quali y class oom o ganiza ion manage child en’s beha io
and ime e icien ly and make he lessons in e es ing o he child en. In
well-o ganized class ooms, child en beha e well and a e in e es ed and
in ol ed in he ac i i ies and lea ning. The las domain, ins uc ional
suppo , is based on s udies in cogni i e and language de elopmen and
in how adul s can suppo his de elopmen wi h, o example, sca -
olding (e.g., Ca s, Fey, Zhang, & Tomblin, 1999; Skibbe, Behnke, &
Jus ice, 2004). In ins uc ionally suppo i e class ooms, eache s sup-
po child en’s lea ning in se e al ways: hey gi e child en speci ic
eedback ha is a ge ed on enhancing he lea ning, hey use e sa ile
language o suppo language de elopmen , and hey ask open-ended
ques ions such as “why” and “how” o encou age con e sa ion and o
suppo child en’s deepe unde s anding o concep s a he han
memo izing ac s. In hese class ooms, child en ac i ely ake pa in
con e sa ions, and ac i i ies a e ela ed o hei own expe iences and
li es.
Teache s’ wo k engagemen and he quali y o eache –child
in e ac ions
Teache s’ occupa ional well-being has ecei ed g owing in e es in
he ield o educa ional esea ch du ing he las decade (see Cumming,
2017). Resea ch in occupa ional well-being has i s ocused on chal-
lenges in well-being such as s ess and bu nou whe eas ecen esea ch
has also no iced he posi i e aspec s o well-being such as wo k
engagemen and job sa is ac ion (see e.g., Sepp¨
al¨
a e al., 2009). How-
e e , he ela ion be ween eache s’ occupa ional well-being and he
quali y o eache –child in e ac ions (e.g., F iedman-K auss, Ra e ,
Mo is, & Jones, 2014; Hoglund, Klingle, & Hosan, 2015) and child en’s
social compe ence (e.g., He man e al., 2018; Siekkinen e al., 2013) has
mainly been examined in e ms o s ess and bu nou . Teache s’ s ess
e e s o nega i e emo ions such as anxie y, us a ion, and ension
(Ky iacou, 2001) and i s associa ions wi h he quali y o eache -child
in e ac ions (e.g., Jennings, 2015; Pen inen e al., 2020) and child
ou comes (e.g., Siekkinen e al., 2013) ha e been nega i e. Since wo k
engagemen e e s o posi i e emo ions (see below), we expec ha i
has posi i e associa ions wi h he quali y o eache -child in e ac ions
and he a e age social compe ence o he child en in he class oom.
Wo k engagemen is cha ac e ized by h ee componen s: igo ,
dedica ion, and abso p ion (Schau eli, Salano a, Gonz´
alez- om´
a, &
Bakke , 2002). Teache s wi h a high le el o igo ha e a lo o ene gy
and a e eady o in es in hei wo k. Highly dedica ed eache s a e
p oud o hei wo k, conside i meaning ul, and eel en husiasm and
inspi a ion while wo king. Finally, eache s wi h a high abso p ion a e
highly concen a ed while wo king and migh eel ha “ ime lies” while
wo king. Compa ed wi h hose in o he p o essions, eache s usually
expe ience ela i ely high wo k engagemen (Hakanen, Ropponen,
Schau eli, & De Wi e, 2019). High wo k engagemen can be conside ed
a o able, as i is ela ed o highe job sa is ac ion (Høigaa d, Giske, &
Sundsli, 2012; Klassen e al., 2012), lowe le els o bu nou (Høigaa d
e al., 2012), and lowe in en ion o lea e he p o ession (Høigaa d e al.,
2012; Skaal ik & Skaal ik, 2016).
Bakke and Deme ou i (2008) in oduced a model o wo k engage-
men , in which hey p oposed ha people’s pe o mance a wo k is
p edic ed by hei wo k engagemen , which is p edic ed by hei e-
sou ces. The i s s udies ha examined he ela ion be ween wo k
engagemen and eaching pe o mance epo ed eache s’ wo k
engagemen as being posi i ely associa ed wi h child- a ed quali y o
ins uc ion in seconda y schools (Klusmann, Kun e , T au wein, Lüd ke,
& Baume , 2008) and eache s’ sel - a ed job pe o mance among
beginning eache s (Bakke & Bal, 2010). Howe e , o da e, he e has
been a limi ed amoun o esea ch conduc ed conce ning he associa-
ions be ween eache s’ wo k engagemen and he obse ed quali y o
eache –child in e ac ions. In kinde ga en class ooms, Pen inen e al.
(2020) epo ed ha eache s expe iencing a high le el o wo k
engagemen also p o ided high-quali y ins uc ional suppo . Howe e ,
eache s’ wo k engagemen was no associa ed wi h he quali y o
emo ional suppo o class oom o ganiza ion. To he bes o ou
knowledge, no esea ch examining he ela ion be ween wo k engage-
men and he obse ed quali y o eache –child in e ac ions in elemen-
a y school class ooms has been conduc ed. Mo eo e , he p e ious
esea ch conduc ed in kinde ga en class ooms has been c oss-sec ional;
hus, i is no clea i high-quali y in e ac ions wi h child en enhance
eache s’ occupa ional well-being o ice e sa. The e o e, in he p esen
s udy, we examined also i he quali y o eache –child in e ac ions is
associa ed wi h eache s’ le el o wo k engagemen .
Child en’s social compe ence and he quali y o eache –child
in e ac ions
Social compe ence can be de ined as an abili y o use a a ie y o
skills in in e ac ions wi h o he s o achie e meaning ul goals in social
g oups (Wa e s & S ou e, 1983). She idan and Walke (1999), and
Me ell and Gimpel (1998) ha e sugges ed ha hese skills can be
concep ualized unde wo sepa a e aspec s. The i s aspec e lec s he
abili y o lea n con ex -app op ia e social skills needed in social in e -
ac ion whe eas he second aspec e lec s he abili y o beha e in a so-
cially accep able manne . When de eloping he Mul isou ce Assessmen
o Social Compe ence Scale, Jun ila, Voe en, Kaukiainen, and Vau as
(2006) named he wo aspec s as p osocial beha io and an isocial
beha io . The i s aspec , p osocial beha io , e e s o beha io s ha
people usually see as desi able and he e o e encou age child en o
display (Jun ila e al., 2006). These beha io s include coope a i e be-
ha io s (e.g., o e ing help, s a ing con e sa ions, pa icipa ing in
g oup ac i i ies) and beha io s ha show empa hy (e.g., sensi i e and
accep ing beha io s, showing, and communica ing posi i e emo ions).
Thus, socially compe en child en ha e social skills ha can be used o
success ully in e ac wi h o he people (Jun ila e al., 2006; Me ell &
Gimpel, 1998; She idan & Walke , 1999).
In con as wi h p osocial beha io , an isocial beha io e e s o
beha io s ha people usually see as undesi able (Jun ila e al., 2006).
These beha io s include impulsi e beha io s (e.g., an ums, ge ing
easily i i a ed) and dis up i e beha io s (e.g., easing, qua eling,
ac ing wi hou hinking). Impulsi e child en ha e challenges in
V. Soininen e al.
Jou nal o Applied De elopmen al Psychology 85 (2023) 101508
3
es aining immedia e esponses, wai ing, and planning (Bae & Nie zel,
1991) whe eas dis up i e child en di ec hei dis up i e beha io s a
o he people, o en in en ionally o annoy hem (Kaplan, Gheen, &
Midgley, 2002). Thus, socially compe en child en can inhibi impul-
si i y and dis up i eness (i.e., an isocial beha io ) and beha e in a way
ha is accep able o o he people (Jun ila e al., 2006; Me ell &
Gimpel, 1998; She idan & Walke , 1999).
When de ining an isocial beha io , i is impo an o conside also
ex e nalizing beha io which sha es some simila i ies wi h an isocial
beha io . Ex e nalizing beha io e e s o a ie y o beha io s ha a e
agains social no ms and/o a e ha m ul o o he people (Kau en &
Ba y, 2020). These beha io s include bo h hype ac i e and agg essi e
beha io s (Hinshaw, 1987; Liu, 2004). Ex e nalizing beha io is some-
imes used as a synonym o an isocial beha io (Hinshaw, 1987; Liu,
2004) bu as Liu (2004) s a es, he e a e dis inc ions be ween he wo
concep s. Fo example, ex e nalizing beha io includes hype ac i e
beha io s bu no all hype ac i e child en a e an isocial (see Liu, 2004).
Mo eo e , ex e nalizing beha io can include beha io s such as sub-
s ance use (Kau en & Ba y, 2020) o delinquency (Liu, 2004) which a e
no pa o an isocial beha io as de ined abo e.
To conclude, child en conside ed as highly socially compe en
display high le els o p osocial beha io in combina ion wi h low le els
o an isocial beha io . A child wi h lowe social compe ence can display
ei he a mix u e o p osocial and an isocial beha io s o a lack o p o-
social beha io s and a p esence o an isocial beha io s. Thus, i is also
possible, ha child en display bo h high p osocial beha io and high
an isocial beha io o low le els o bo h ypes o beha io (Jun ila
e al., 2006). S udies ha e highligh ed he impo ance o child en’s so-
cial compe ence by showing i s ela ionship wi h se e al posi i e ou -
comes. Fo example, in kinde ga en, a link has been ound be ween
child en’s social compe ence and p eli e acy skills (Paka inen, Salmi-
nen, Le kkanen, & on Suchodole z, 2018), he numbe o iends, and
he le el o pee accep ance (Ladd, Bi ch, & Buhs, 1999). In elemen a y
school, longi udinal s udies ha e epo ed ha child en’s social
compe ence p edic s hei la e academic skills (Cap a a, Ba ba anelli,
Pas o elli, Bandu a, & Zimba do, 2000; Legkauskas & Magelinskai ˙
e-
Legkauskien˙
e, 2021), le el o anxie y (Jun ila, Vau as, Niemi, &
Laakkonen, 2012; Legkauskas & Magelinskai ˙
e-Legkauskien˙
e, 2021),
loneliness (Jun ila e al., 2012), and pee p e e ence (Cap a a e al.,
2000). Fu he mo e, child en wi h high social compe ence in kinde -
ga en a e mo e likely o ob ain a college deg ee and a ull- ime job in
ea ly adul hood and less likely o be engaged in c iminal ac i i y (Jones,
G eenbe g, & C owley, 2015).
The ole o eache -child in e ac ions in child en’s social compe ence
has been examined a he no ably in di e en educa ional se ings. A
me a-analysis by Pe lman e al. (2016) showed ha in p eschool class-
ooms whe e he quali y o ins uc ional suppo was high, child en’s
social compe ence was also high. Howe e , he quali y o emo ional
suppo o class oom o ganiza ion was no ela ed wi h child en’s social
compe ence (Pe lman e al., 2016). Simila ly, in Finnish kinde ga en
class ooms whe e he quali y o ins uc ional suppo was high in all,
child en displayed mo e empa hy and less dis up i eness in sp ing
(Siekkinen e al., 2013). In con as , in a s udy by B oekhuizen e al.
(2016), high-quali y emo ional suppo and class oom o ganiza ion –
bu no ins uc ional suppo – in p ekinde ga en and kinde ga en
class ooms p edic ed child en’s high social compe ence in i s g ade.
Mo eo e , i has been shown in middle school da a ha high emo ional
suppo was associa ed wi h mo e p osocial beha io and high class-
oom o ganiza ion wi h less agg ession (Luckne & Pian a, 2011).
Howe e , he e has been less esea ch examining ecip ocal associa ions
o i child en’s social compe ence p edic s he quali y o eache -child
in e ac ions. As an excep ion, one ecen s udy epo ed a ecip ocal
link be ween kinde ga ene s’ empa hy and he quali y o emo ional
suppo indica ing ha he quali y o eache -child in e ac ions and
child en’s social compe ence migh be ecip ocally associa ed. I he e
was high-quali y emo ional suppo in he class oom in all, child en
displayed mo e empa hy in sp ing and ice e sa: i child en displayed
mo e empa hy in he class oom in all, he quali y o emo ional suppo
was highe in sp ing (Paka inen e al., 2020). Fu he mo e, i he e was
mo e impulsi i y in he class oom in all, he quali y o emo ional sup-
po was lowe in sp ing (Paka inen e al., 2020).
To conclude, many s udies ha e shown ha high-quali y emo ional
suppo (B oekhuizen e al., 2016; Paka inen e al., 2020), class oom
o ganiza ion (B oekhuizen e al., 2016; Luckne & Pian a, 2011), and
ins uc ional suppo (Pe lman e al., 2016; Siekkinen e al., 2013) a e
ela ed wi h highe a e age social compe ence o he child en in he
class oom. Howe e , only ew s udies ha e examined i child en’s social
compe ence plays a ole in eache -child in e ac ions and no all do-
mains o eache -child in e ac ions ha e ye been examined ecip ocally
wi h social compe ence (see Paka inen e al., 2020). Fo his eason,
mo e esea ch is needed o ully unde s and he dynamics be ween he
quali y o eache -child in e ac ions and a e age social compe ence o
he child en in he class oom.
Child en’s social compe ence and eache s’ wo k engagemen
Ea lie esea ch on he associa ion be ween eache s’ occupa ional
well-being and child en’s social compe ence has ocused on he ole o
eache s’ s ess in child en’s social compe ence. In Finnish kinde ga en
class ooms, child en displayed mo e impulsi i y and dis up i eness, less
empa hy, and lowe coope a ion skills i he eache expe ienced highe
s ess (Siekkinen e al., 2013). Simila ly, in US kinde ga en o 4 h g ade
class ooms, he e was mo e dis up i e beha io and less p osocial
beha io in he class oom i he eache expe ienced high le els o s ess
and bu nou oge he wi h low le els o coping (He man e al., 2018). In
u n, esea ch on he associa ions be ween eache s’ wo k engagemen
and child en’s social compe ence emains limi ed. Howe e , Jennings
and G eenbe g’s (2009) p osocial class oom model sugges s ha
eache s’ social and emo ional compe ence and well-being enhance hei
capabili y o p o ide high-quali y in e ac ions which suppo child en’s
social, emo ional, and academic skill de elopmen . A he same ime, he
model indica es ha i eache s ha e challenges o hei well-being, i
may be di icul o hem o espond o child en’s indi idual needs, be
p oac i e, and manage child beha io , which is e lec ed in child ou -
comes such as child beha io in he class oom. Thus, based on his
model, i is possible, ha eache s’ wo k engagemen as one aspec o
eache s’ occupa ional well-being is associa ed wi h bo h he quali y o
eache –child in e ac ions and he a e age social compe ence o he
child en in he class oom.
Educa ion sys em in Finland
The cu en s udy was conduc ed in Finland, whe e nine yea s o
comp ehensi e school begin in he yea when he child u ns se en yea s
old. P ima y school is p eceded by a one-yea , ee o cha ge, and
manda o y kinde ga en yea , and ollowed by h ee yea s o seconda y
educa ion in ei he uppe seconda y school o oca ional school. P i-
ma y and seconda y educa ion in Finland is also ee o cha ge.
Elemen a y school eache s (g ades 1–6) a e equi ed o ha e a leas a
mas e ’s deg ee in educa ion, and an a e age class size in elemen a y
school is 20 child en (O ganisa ion o Economic Co-ope a ion and
De elopmen , 2018). Class sizes a e no na ionally egula ed. In i s
g ade, one school week consis s o a minimum o 20 h o eaching. Fi s
g ade eaching in Finland is ocused on he de elopmen o child en’s
academic, social, and mo o skills (Finnish Na ional Agency o Educa-
ion, 2016). In e ms o social compe ence, de elopmen o coope a ing
skills is impo an as child en lea n o wo k oge he in g oups.
The aims o he p esen s udy
The e has been a limi ed amoun o esea ch on he ole o eache s’
wo k engagemen in he obse ed quali y o eache –child in e ac ions
V. Soininen e al.
Jou nal o Applied De elopmen al Psychology 85 (2023) 101508
4
(wi h he excep ion o a c oss-sec ional s udy by Pen inen e al., 2020).
Mo eo e , p e ious s udies examining he associa ions be ween each-
e –child in e ac ions and child en’s social compe ence ha e mainly
examined i he quali y o eache –child in e ac ions enhance child en’s
a e age social compe ence in he class ooms (e.g., Siekkinen e al.,
2013) and less a en ion has been gi en o ecip ocal associa ions (see
Paka inen e al., 2020). Finally, he associa ions be ween eache s’ wo k
engagemen and child en’s social compe ence emain uns udied. Since
li le is known abou he associa ions and di ec ion o e ec among he
a iables, his s udy examined ecip ocal associa ions o he quali y o
eache –child in e ac ions, eache s’ wo k engagemen , and child en’s
social compe ence be ween class ooms. Mo e speci ically, he ollowing
esea ch ques ion was in es iga ed: To wha ex en a e he quali y o
eache –child in e ac ions (emo ional suppo , class oom o ganiza ion, and
ins uc ional suppo ), eache s’ le el o wo k engagemen , and a e age so-
cial compe ence (p osocial and an isocial beha io ) o he child en in he
class ooms associa ed ac oss g ade 1?
Mo eo e , he ollowing hypo heses we e se : Fi s , based on he
model o wo k engagemen by Bakke and Deme ou i (2008), we ex-
pec ed ha eache s’ le el o wo k engagemen is posi i ely associa ed
wi h subsequen quali y o eache -child in e ac ions in he class ooms
(Hypo hesis 1). Second, because ecen esea ch has sugges ed ha he
quali y o eache -child in e ac ions and child en’s social compe ence
migh be ecip ocally associa ed (see Paka inen e al., 2020), we
assumed ha he quali y o eache -child in e ac ions is posi i ely
associa ed wi h subsequen social compe ence o he child en in he
class ooms (Hypo hesis 2a) and ha a e age social compe ence o he
child en in he class ooms is posi i ely associa ed wi h subsequen
quali y o eache -child in e ac ions (Hypo hesis 2b). Thi d, al hough
ea lie esea ch has no examined he associa ions be ween eache s’
wo k engagemen and a e age social compe ence o he child en in he
class oom, based on he p osocial class oom model by Jennings and
G eenbe g (2009) and ea lie esea ch on he associa ions be ween
eache s’ s ess and child en’s social compe ence (Siekkinen e al.,
2013), we hypo hesized ha eache s’ le el o wo k engagemen is
posi i ely associa ed wi h subsequen social compe ence o he child en
in he class ooms (Hypo hesis 3).
Me hod
Pa icipan s and p ocedu es
Pa icipan s o he p esen s udy we e 51 i s g ade eache s and 815
child en om hei class ooms who we e pa icipa ing in a ollow-up
s udy (Le kkanen & Paka inen, 2016-2022) in he 2017–2018 school
yea . Pa icipa ion in he s udy was olun a y, and eache s and chil-
d en’s gua dians ga e w i en consen o hei o hei child’s pa ici-
pa ion. The e hical s a emen om he e hical commi ee o he
uni e si y was ecei ed p io o commencing he s udy.
Be o e da a collec ion, all 70 elemen a y schools in eigh munici-
pali ies in Cen al Finland we e con ac ed o ask pe mission om he
p incipal o conduc he s udy in he school. All schools p o ided
pe mission. Nex , all 136 i s g ade eache s in he schools we e con-
ac ed and in i ed o pa icipa e in he s udy. Finally, 54 eache s om
36 schools ag eed o pa icipa e in he s udy. Numbe o pa icipa ing
class ooms pe school a ied be ween one and ou and 88.5% o he
child en om he class ooms pa icipa ed in he s udy based on he
gua dians’ consen s. Pa icipa ing schools we e loca ed in bo h u al and
subu ban a eas. Because he e was missing da a om h ee o he
pa icipa ing 54 class ooms, only 51 class ooms we e included in he
analyses o he p esen s udy. The mean age o he pa icipa ing eache s
was 45 yea s (SD =8.85, min. 27 yea s, max. 62 yea s), and hei
a e age wo k expe ience a schools was 16.54 yea s (SD =9.31, min.
0.5 yea s, max. 39 yea s). All eache s we e Finnish-speaking Caucasians
and mos o hem (94.1%) we e emale.
In he pa icipa ing class ooms, class sizes a ied om se en o 25
(M =19.25, SD =4.40). Child en’s (50.1% gi ls) mean age a he
beginning o he school yea was 7.19 yea s (SD =0.34, min. 4.83 yea s,
max. 9.28 yea s). The highes educa ion o pa en s a ied as ollows: no
oca ional educa ion (2.8%), oca ional cou ses (min. 4 mon hs; 1.2%),
oca ional school deg ee (33.1%), college-le el aining (7%), poly-
echnic deg ee (23.2%), uni e si y/college deg ee (22.8%), and licen-
ia e o doc o al deg ee (4.3%). Howe e , pa en al educa ion
in o ma ion was a ailable only o 67% o he pa icipa ing child en.
Da a o he p esen s udy we e collec ed wice du ing he i s school
yea : in all (2017; Time 1 [T1]) and sp ing (2018; Time 2 [T2]). Da a
consis ed o ideo eco dings om he class ooms, eache ques ion-
nai es on hei wo k engagemen , and eache a ings on child en’s so-
cial compe ence. Video eco dings o he classes we e collec ed o assess
he quali y o eache –child in e ac ions in he class oom. Video e-
co dings we e conduc ed du ing one egula school day, and h ee o
ou lessons we e eco ded in each class oom in all (T1) and again in
sp ing (T2). Simila ly, he eache ques ionnai es and hei a ings on
child en’s social compe ence we e collec ed i s in all (T1) and again in
sp ing (T2). Reliabili y in o ma ion, ha is, C onbach’s alphas o all
measu es, is p esen ed in Table 2.
Measu es
The quali y o eache –child in e ac ions
The quali y o eache –child in e ac ions was assessed wi h he
Class oom Assessmen Sco ing Sys em (CLASS K-3: Pian a e al., 2008).
Validi y o he CLASS in he Finnish da a has been epo ed by Paka inen
e al. (2010). The measu e has been widely used o measu e he quali y
o eache -child in e ac ions in Finland (e.g., Paka inen e al., 2020;
Pen inen e al., 2020; Salminen, Paka inen, Poikkeus, Laakso, & Le k-
kanen, 2022; Siekkinen e al., 2013) and a ound he wo ld (e.g., Canada:
Hoglund e al., 2015; Chile: Ley a e al., 2015; China: Hu e al., 2019).
CLASS concep ualizes eache –child in e ac ions unde h ee domains:
emo ional suppo , class oom o ganiza ion, and ins uc ional suppo .
Each o he domains consis s o h ee o ou dimensions, which again
a e measu ed wi h mo e speci ic beha io indica o s. Indica o s, di-
mensions, and domains a e all desc ibed in he coding manual (Pian a
e al., 2008), which guides he assessmen o he quali y o in e ac ions.
In he p esen s udy, app oxima ely se en (M =6.82, SD =1.18, min. 5,
max. 9) cycles pe class oom in all and app oxima ely six (M =5.90, SD
=0.73, min. 5, max. 8) cycles pe class oom in sp ing we e a ed by six
ce i ied code s. The a e age leng h o one cycle was 18.5 min (SD =3.0,
min. 10.5., max. 26.5). The mean sco e o cycles o each domain a one
ime poin was used in he analyses. To assess he in e - a e eliabili y
o he CLASS a ings, 26% o he cycles we e double coded, and he
adjacen ag eemen s and in aclass co ela ion coe icien s (ICCs) o he
a ings we e calcula ed (see Table 1). Adjacen ag eemen e e s o he
pe cen age o a ings ha a e wi hin one poin o each o he (Pian a
e al., 2008). De elope s o he CLASS measu e (Pian a e al., 2008) use
his measu e o indica e su icien in e - a e eliabili y o he a ings. In
he p esen s udy, adjacen ag eemen s a ied om 77.2% o 100% (see
Table 1). Thus, he le el o adjacen ag eemen was accep able o all en
dimensions a bo h imepoin s (Pian a e al., 2008).
ICCs we e calcula ed using wo-way andom e ec s model wi h ab-
solu e ag eemen (Lande s, 2015). Fo mos o he dimensions, ICCs
indica ed good (0.60–0.74) o excellen (0.75–1.00) eliabili y (Cic-
che i & Spa ow, 1981; see Table 1). As excep ions, he le el o
ag eemen was ai o he a ings o Rega d o s uden pe spec i es in
he sp ing da a and poo o sp ing a ings o Nega i e clima e in he
sp ing da a. As ICC as a measu e is subjec o a a ie y o s a is ical
assump ions (e.g., no mali y, s able a iance), low ICC o nega i e
clima e migh be caused by he nonno mali y o he dis ibu ion and low
a iance in he a ings o his dimension. As shown in Table 1, adjacen
ag eemen o he a ings o nega i e clima e in sp ing was e y and o e
90% o he a ings we e wi hin one poin o each o he .
V. Soininen e al.
Jou nal o Applied De elopmen al Psychology 85 (2023) 101508
5
Teache s’ wo k engagemen
Teache s’ wo k engagemen was measu ed wi h he U ech Wo k
Engagemen Scale (UWES; Schau eli e al., 2002). Validi y o he UWES
measu e in he Finnish da a has been epo ed by Sepp¨
al¨
a e al. (2009).
The measu e has been widely used in Finland (e.g., Hakanen, Bakke , &
Scha eli, 2006; Pen inen e al., 2020; Vuo io, Suominen, Kau iainen, &
Ko honen, 2019) and a ound he wo ld (e.g., Eu opean coun ies:
Hakanen e al., 2019; India: Ka a ia, Ga g, & Ras ogi, 2013; Se bia:
Pe o i´
c, Vukeli´
c, &
ˇ
Cizmi´
c, 2017). UWES consis s o nine i ems ha he
eache s answe ed on a scale om 1 (ne e ) o 7 (daily). These i ems all
unde h ee domains: igo (e.g., “A my wo k, I eel bu s ing wi h en-
e gy”), dedica ion (e.g., “I am en husias ic abou my job”), and ab-
so p ion (e.g., “I ge ca ied away when I am wo king”). The mean sco e
o he nine i ems was used in he analysis.
Child en’s social compe ence
Teache s a ed child en’s social compe ence using he Mul isou ce
Assessmen o Child en’s Social Compe ence Scale (MASCS; Jun ila
e al., 2006). Validi y o he MASCS measu e in he Finnish da a has been
epo ed by Jun ila e al. (2006). The measu e has been ecen ly used o
measu e child en’s social compe ence o example in s udies by Panula,
Jun ila, A omaa, Rau a a, and R¨
aih¨
a (2020), Paka inen e al. (2020),
and Salminen e al. (2022). MASCS consis s o 15 i ems and ou sub-
scales: coope a ing skills (5 i ems, e.g., “o e s help o o he child en”),
empa hy (3 i ems, e.g., “is sensi i e o he eelings o o he s”), impul-
si i y (3 i ems, e.g., “is easily i i a ed”), and dis up i eness (4 i ems, e.
g., “bo he s and annoys o he child en”). All 15 i ems we e a ed by
eache s on a scale om 1 (ne e ) o 4 ( e y equen ly). O he sub-
scales, coope a ing skills and empa hy e lec p osocial beha io ,
whe eas impulsi i y and dis up i eness e lec an isocial beha io .
Mean sco es o p osocial beha io and an isocial beha io we e used in
he analysis.
Con ol a iables
Teache s epo ed hei wo k expe ience and class size on he ques-
ionnai e. Wo k expe ience and class size we e con olled o in he
analyses because ea lie esea ch has indica ed ha hese eache and
class oom cha ac e is ics a e ela ed wi h he quali y o eache -child
in e ac ions (e.g., F iedman-K auss e al., 2014; Slo , Le kkanen, &
Leseman, 2015). Child en’s gende was con olled o as p e ious
s udies ha e shown ha gi ls’ social compe ence is usually highe han
boys’ (e.g., Jun ila e al., 2006; Paka inen e al., 2018; Siekkinen e al.,
2013).
Da a analyses
Fi s , as p elimina y analysis, co ela ions among he a iables we e
conduc ed wi h Mplus e sion 8 (Mu h´
en & Mu h´
en, 1998-2017) (see
Table 2). Mo eo e , o in es iga e i he e we e di e ences among
class ooms in child en’s social compe ence o gende , in aclass co e-
la ion coe icien s (ICCs; see Table 2) we e calcula ed. Thi d, o examine
o wha ex en a e quali y o eache -child in e ac ions, eache s’ le el o
wo k engagemen , and child en’s a e age social compe ence in class-
ooms associa ed ac oss g ade 1 all (T1) and g ade 1 sp ing (T2),
mul ile el models we e conduc ed wi h Mplus. In he models, he social
compe ence T1 sco es we e g oup-mean cen e ed a he wi hin-le el. A
he be ween-le el, social compe ence a T1 was used as a class oom-
mean a e aged a iable (clus e -mean). Mul ile el modeling is an
ideal analysis s a egy o nes ed da a because i enables he a iance in
he obse ed a iables o be di ided in o he a ia ion caused by he
membe ship in a ce ain class oom (be ween-le el a ia ion) and a i-
a ion due o indi idual child en’s di e ences, a e aking in o accoun
class oom membe ship (wi hin-le el a ia ion) (Hox, 2010).
A he be ween-le el o he models (see schema ic ep esen a ion o
he mul ile el models, Fig. 1), s abili y and c oss-lagged pa hs we e
es ima ed be ween he quali y o eache –child in e ac ions, eache ’s
le el o wo k engagemen , and child en’s a e age social compe ence in
he class oom a T1 and T2. All s udy a iables we e allowed o co ela e
wi h each o he . Mo eo e , a he be ween-le el o he models, eache s’
wo k expe ience and class size we e con olled o . Howe e , hese wo
be ween-le el con ol a iables we e excluded om he inal models due
o model pa simony. Excluding he be ween-le el con ol a iables om
he inal models did no in luence he esul s.
A he indi idual le el o he models (wi hin-le el), g oup-mean
cen e ed social compe ence a T1 p edic ed social compe ence a T2.
Social compe ence a T1 was p edic ed by gende . Mul ile el-models
we e conduc ed sepa a ely o he h ee domains o eache –child in-
e ac ions (i.e., emo ional suppo , class oom o ganiza ion, and
ins uc ional suppo ) and wo domains o child en’s social compe ence
(i.e., p osocial and an isocial beha io ) because bo h he h ee CLASS
domains and wo MASCS domains we e highly co ela ed. Conse-
quen ly, six models we e conduc ed (Figs. 2–7). Goodness-o - i o he
models was e alua ed wi h a
χ
2 es , compa a i e i index (CFI), Tucke
Lewis index (TLI), oo mean squa e e o o app oxima ion (RMSEA),
and s anda dized oo mean squa e esidual (SRMR). Model i s o he
in es iga ed models a e p esen ed in Table 3.
Resul s
Desc ip i e esul s, co ela ions among he s udy a iables, and
in aclass co ela ion coe icien s (ICCs) a e shown in Table 2. ICCs we e
es ima ed using he class oom as a clus e a iable. ICCs we e s a is i-
cally signi ican o bo h dimensions o child en’s social compe ence (i.
e., p osocial, and an isocial beha io ) a bo h ime poin s (i.e., i s g ade
all and i s g ade sp ing), indica ing ha he e was s a is ically sig-
ni ican a ia ion in child en’s social compe ence among class ooms.
ICC o child en’s gende was no s a is ically signi ican , indica ing ha
he e was no s a is ically signi ican a ia ion in child en’s gende
among he class ooms. Consequen ly, child en’s gende was ea ed as a
wi hin-le el a iable ( a ia ion only wi hin class ooms) in u he
analysis, whe eas child en’s social compe ence was ea ed as bo h a
wi hin- and be ween-le el a iable.
Table 1
In e - a e eliabili ies o he class oom assessmen sco ing sys em a ings.
PC NC TS RSP BM PR ILF CD QF LM
Fi s g ade all
Adjacen ag eemen 97.4% 100% 94.7% 93.4% 98.7% 97.4% 96.1% 96.1% 96.1% 97.4%
ICC 0.82 0.88 0.84 0.86 0.86 0.81 0.78 0.84 0.85 0.89
Fi s g ade sp ing .
Adjacen ag eemen 88.3% 93.2% 88.3% 77.7% 93.2% 94.2% 93.2% 87.4% 92.2% 91.3%
ICC 0.70 0.24 0.67 0.54 0.71 0.65 0.73 0.72 0.69 0.70
No e: PC =Posi i e clima e; NC =Nega i e clima e; TS =Teache sensi i i y; RSP =Rega d o s uden pe spec i es; BM =Beha io managemen ; PR =P oduc i i y;
ILF =Ins uc ional lea ning o ma s; CD =Concep de elopmen ; QF =Quali y o eedback; LM =Language modeling; ICC =In aclass co ela ion coe icien .
V. Soininen e al.
Jou nal o Applied De elopmen al Psychology 85 (2023) 101508
6
The quali y o emo ional suppo
The esul s i s showed ha a he be ween-le el o he model 1,
quali y o emo ional suppo , eache s’ wo k engagemen , and chil-
d en’s p osocial beha io we e associa ed wi h hei p eceding alues
(Fig. 2: Model 1). Mo eo e , a e age p osocial beha io o he child en
in he class oom in all was ma ginally signi ican ly associa ed wi h he
quali y o emo ional suppo , and s a is ically signi ican ly associa ed
wi h he le el o eache s’ wo k engagemen in sp ing. Thus, i child en
displayed mo e p osocial beha io in he class oom a he beginning o
g ade 1, he e was a highe quali y o emo ional suppo in he class-
oom a he end o g ade 1, al hough he associa ion was a he small.
Fu he mo e, i he e was mo e p osocial beha io in he class oom in
all, eache s epo ed highe le els o wo k engagemen in sp ing. A he
wi hin-le el o he model, child en’s p osocial beha io a he beginning
o g ade 1 was posi i ely associa ed wi h hei p osocial beha io a he
end o g ade 1, and gi ls displayed mo e p osocial beha io han boys.
Resul s u he showed ha a he be ween-le el o model 2, a e age
an isocial beha io o he child en in he class oom was s able ac oss
g ade 1 (Fig. 3: Model 2). Mo eo e , eache s’ wo k engagemen was
associa ed wi h he subsequen quali y o emo ional suppo ; i eache s
epo ed highe wo k engagemen in all, he obse ed quali y o
emo ional suppo in he class oom was highe in sp ing. A he wi hin-
le el o he model, child en’s an isocial beha io in all was associa ed
wi h hei an isocial beha io in sp ing. Mo eo e , boys displayed mo e
an isocial beha io han gi ls.
The quali y o class oom o ganiza ion
Resul s showed ha a he be ween-le el o model 3, class oom o -
ganiza ion was s able ac oss g ade 1 (Fig. 4: Model 3). Mo eo e ,
eache s’ wo k engagemen was posi i ely associa ed wi h subsequen
quali y o class oom o ganiza ion. Thus, i eache s epo ed highe
wo k engagemen in he all, he quali y o class oom o ganiza ion in he
class oom was highe in he sp ing. Fu he mo e, class oom o ganiza-
ion was posi i ely ela ed o subsequen p osocial beha io o he
child en in he class oom, albei ma ginally signi ican ly. Thus, i he
quali y o class oom o ganiza ion in he class oom was highe in all,
he e was on a e age mo e p osocial beha io in he class oom in sp ing,
al hough he associa ion was a he small.
The esul s u he showed ha eache s’ wo k engagemen was
nega i ely associa ed wi h subsequen an isocial beha io o he chil-
d en in he class oom (Fig. 5: Model 4). Thus, i eache s epo ed highe
wo k engagemen in he all, he e was on a e age less an isocial
beha io in he class oom in sp ing.
The quali y o ins uc ional suppo
Ins uc ional suppo showed s abili y ac oss he school yea (Fig. 6:
Model 5). Mo eo e , he e was a ecip ocal associa ion be ween he
quali y o ins uc ional suppo and child en’s p osocial beha io ; Fi s ,
ins uc ional suppo was posi i ely and s a is ically signi ican ly asso-
cia ed wi h subsequen p osocial beha io o he child en in he class-
oom. Thus, in class ooms whe e he quali y o ins uc ional suppo was
highe in all, child en displayed on a e age mo e p osocial beha io in
sp ing. Second, a e age p osocial beha io o he child en in he class-
oom in all was posi i ely associa ed wi h subsequen quali y o
ins uc ional suppo in sp ing. Thus, in class ooms whe e child en
displayed on a e age mo e p osocial beha io in he all, he quali y o
ins uc ional suppo was highe in sp ing.
Finally, he esul s showed ha he quali y o ins uc ional suppo
was nega i ely associa ed wi h subsequen an isocial beha io o he
child en in he class oom (Fig. 7: Model 6). Thus, i he quali y o
ins uc ional suppo was highe in all, he e was on a e age less
an isocial beha io in he class oom in sp ing. Mo eo e , eache s’ wo k
engagemen was posi i ely associa ed wi h subsequen ins uc ional
Table 2
Desc ip i e s a is ics and bi a ia e co ela ions be ween s udy a iables (wi hin-le el co ela ions below he diagonal and be ween-le el co ela ions abo e he diagonal).
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.
Fi s g ade all
1. Emo ional sup. T1 1 0.72*** 0.54*** 0.41*** 0.51*** −0.08 0.70*** 0.43*** 0.43*** 0.35** 0.46** −0.17 – 0.08 −0.28*
2. Class oom o g. T1 – 1 0.46*** 0.43*** 0.46*** 0.06 0.56*** 0.57*** 0.40*** 0.35** 0.49** −0.02 – 0.03 −0.06
3. Ins uc ional sup. T1 – – 1 0.18 0.29
†
−0.25
†
0.26* 0.16 0.45*** 0.10 0.40** −0.41** – 0.20 −0.21
4. Wo k engag. T1 – – – 1 0.40** −0.22 0.46*** 0.47*** 0.34** 0.64*** 0.29
†
−0.30 – 0.19 −0.05
5. P osocial beh. T1 – – – – 1 −0.10 0.58*** 0.48*** 0.37** 0.51*** 0.65*** −0.17 – 0.23 −0.23
6. An isocial beh. T1 – – – – −0.32*** 1 −0.13 −0.20 −0.23
†
−0.19 0.03 0.82*** – −0.35* −0.06
Fi s g ade sp ing
7. Emo ional sup. T2 – – – – – – 1 0.73*** 0.51*** 0.45*** 0.39** −0.27
†
– 0.08 −0.25*
8. Class oom o g. T2 – – – – – – – 1 0.52*** 0.42*** 0.38** −0.18 – 0.05 −0.08
9. Ins uc ional sup. T2 – – – – – – – – 1 0.23** 0.23
†
−0.28* – 0.10 0.14
10. Wo k engag. T2 – – – – – – – – – 1 0.11 −0.09 – 0.01 −0.24*
11. P osocial beh. T2 – – – – 0.75*** −0.36*** – – – – 1 −0.11 – 0.08 −0.31
†
12. An isocial beh. T2 – – – – −0.26*** 0.84*** – – – – −0.42*** 1 – −0.42* 0.03
Con ol a iables
13. Child gende
1
– – – – −0.16** 0.29*** – – – – −0.19*** 0.28*** 1 – –
14.Class size – – – – – – – – – – – – – 1 −0.06
15. Wo k expe ience – – – – – – – – – – – – – – 1
Desc ip i e s a is ics
Mean 5.42 5.62 2.55 6.08 3.04 1.64 5.36 5.68 2.60 6.16 3.02 1.70 1.50 19.25 16.54
S d. de ia ion 0.46 0.45 0.44 0.75 0.30 0.33 0.40 0.43 0.46 0.83 0.26 0.34 0.50 4.39 9.31
Minimum 4.14 4.15 1.78 4.13 2.13 1.07 4.38 4.44 1.87 2.67 2.28 1.09 1 7 0.5
Maximum 6.40 6.47 3.53 7.00 3.65 2.58 6.20 6.50 3.73 7.00 3.53 2.83 2 25 39
α
0.79 0.74 0.82 0.87 0.89 0.92 0.74 0.75 0.79 0.93 0.88 0.92 – – –
ICC
2
– – – – 0.24*** 0.18*** – – – – 0.18*** 0.13*** 0.00 – –
No e: T1 = i s g ade all, T2 = i s g ade sp ing; N =51; ***p <.001, **p <.01, *p <.05;
†
p <.10;
1
Child gende : 1 = emale, 2 =male;
2
In aclass co ela ion coe icien s.
V. Soininen e al.
Jou nal o Applied De elopmen al Psychology 85 (2023) 101508
7
suppo ; i he eache epo ed highe wo k engagemen in all, he
quali y o ins uc ional suppo in he class oom was highe in sp ing.
Discussion
The aim o his s udy was o examine ecip ocal associa ions among
he quali y o eache –child in e ac ions, eache s’ le el o wo k
engagemen , and a e age social compe ence o he child en in he
class oom du ing i s g ade. The esul s showed ecip ocal associa ions
be ween he quali y o ins uc ional suppo and child en’s a e age
p osocial beha io in he class ooms. O he associa ions we e in one
di ec ion: Teache s’ wo k engagemen was posi i ely associa ed wi h
subsequen quali y o eache –child in e ac ions (emo ional suppo ,
class oom o ganiza ion, and ins uc ional suppo ). Child en’s a e age
p osocial beha io in he class oom was posi i ely associa ed wi h
subsequen quali y o emo ional suppo whe eas he quali y o
class oom o ganiza ion was posi i ely associa ed wi h subsequen p o-
social beha io in he class oom. Howe e , hese wo associa ions we e
a he small and he e o e only ma ginally signi ican . Finally, eache s’
high le els o wo k engagemen in all we e associa ed wi h less an i-
social beha io in he class ooms in sp ing whe eas child en’s a e age
p osocial beha io in he class oom in all was associa ed wi h highe
wo k engagemen o eache s in sp ing. Toge he hese esul s e eal
new in o ma ion on he di e se associa ions among eache -child in-
e ac ions, eache s’ occupa ional well-being, and child en’s p osocial
and an isocial beha io s in i s g ade class ooms.
The ole o wo k engagemen in he quali y o eache –child in e ac ions
As hypo hesized (Hypo hesis 1) based on he model o wo k
engagemen (Bakke & Deme ou i, 2008), eache s’ wo k engagemen
was posi i ely associa ed wi h subsequen quali y o eache –child
Fig. 1. Schema ic Rep esen a ion o he Models. No e: T1 = i s g ade all; T2 = i s g ade sp ing
Fig. 2. Model 1: Emo ional Suppo , Wo k Engagemen , and P osocial Beha io . No e: T1 = i s g ade all; T2 = i s g ade sp ing;
†
p<.10, *p<.05, **p<.01,
***p<.001. Non-signi ican associa ions we e es ima ed bu no displayed due o easons o model cla i y. T1 social compe ence a he wi hin-le el was g oup-mean
cen e ed, and a he be ween-le el, he mean o child en’s social compe ence in he class oom was used (clus e -mean).
V. Soininen e al.
Jou nal o Applied De elopmen al Psychology 85 (2023) 101508
8
in e ac ions. Associa ions be ween obse ed quali y o eache –child
in e ac ions and eache s’ wo k engagemen ha e been s udied ea lie
only c oss-sec ionally (see Pen inen e al., 2020). Thus, he esul s o he
p esen s udy expand ou unde s anding o he ole o eache s’ wo k
engagemen in he quali y o eache –child in e ac ions by showing ha
eache s’ wo k engagemen a he beginning o g ade 1 is ela ed wi h
he quali y o emo ional suppo , class oom o ganiza ion, and ins uc-
ional suppo a he end o g ade 1.
When eache s expe ience high wo k engagemen , hey ha e ene gy
o wo k, hey ind hei job as being meaning ul, and i is easy o hem
o concen a e on hei wo k (Schau eli e al., 2002). This enjoymen and
dedica ion migh be e lec ed in eache s’ mo i a ion o pe o m in hei
job as well as possible (see Bakke & Deme ou i, 2008). The p esen
s udy u he sugges s ha he ene gy, dedica ion, and concen a ion
ha eache s eel when expe iencing wo k engagemen can be seen in all
h ee domains o eache -child in e ac ions: in he wa m and suppo i e
in e ac ions (emo ional suppo ), in p oac i e beha io managemen
and e icien ime managemen (class oom o ganiza ion), and in high-
quali y ins uc ions, con e sa ions, and eedback (ins uc ional sup-
po ). Al oge he , he esul s o his s udy, oge he wi h he ea lie e-
sul s (Pen inen e al., 2020) and heo e ical models (Bakke &
Deme ou i, 2008; Jennings & G eenbe g, 2009), highligh he impo -
ance o eache s’ occupa ional well-being in he quali y o eache –-
child in e ac ions.
The in e play be ween social compe ence and he quali y o eache –child
in e ac ions
The esul s o he p esen s udy suppo ed he Hypo hesis 2a and 2b
by showing ha he e we e ecip ocal associa ions be ween he quali y
o ins uc ional suppo and a e age p osocial beha io o he child en
in he class oom. Mo e p ecisely, he esul s indica ed ha in line wi h
Hypo hesis 2a, he quali y o ins uc ional suppo was associa ed wi h
subsequen p osocial and an isocial beha io s in he class oom; i he e
was high-quali y ins uc ional suppo e iden in he class oom in all,
child en showed on a e age mo e p osocial beha io and less an isocial
beha io in he class oom in sp ing. Mo eo e , in line wi h he Hy-
po hesis 2b, in class ooms whe e child en displayed mo e p osocial
beha io in all, he quali y o ins uc ional suppo was highe in sp ing.
Associa ions be ween he emaining wo domains o eache -child in-
e ac ions and child en’s social compe ence we e no ecip ocal bu he
di ec ion o he associa ion was di e en o emo ional suppo and
Fig. 3. Model 2: Emo ional Suppo , Wo k Engagemen and An isocial Beha io . No e: T1 = i s g ade all; T2 = i s g ade sp ing;
†
p<.10, *p<.05, **p<.01,
***p<.001. Non-signi ican associa ions we e es ima ed bu no displayed due o easons o model cla i y. T1 social compe ence a he wi hin-le el was g oup-mean
cen e ed, and a he be ween-le el, he mean o child en’s social compe ence in he class oom was used (clus e -mean).
Fig. 4. Model 3: Class oom O ganiza ion, Wo k Engagemen , and P osocial Beha io . No e: T1 = i s g ade all; T2 = i s g ade sp ing;
†
p<.10, *p<.05, **p<.01,
***p<.001. Non-signi ican associa ions we e es ima ed bu no displayed due o easons o model cla i y. T1 social compe ence a he wi hin-le el was g oup-mean
cen e ed, and a he be ween-le el, he mean o child en’s social compe ence in he class oom was used (clus e -mean).
V. Soininen e al.