1 s2.0 S0742051X21001876 Main

Teaching and Teacher Education 106 (2021) 103463
Contents lists available at ScienceDirect
Teaching and Teacher Education

journal homepage: www.elsevier.com/locate/tate
Collective teacher innovativeness in 48 countries: Effects of teacher

autonomy, collaborative culture, and professional learning
Dong Nguyen a, *, Marcus Pietsch b, Sedat Gümüş c
a
School of Education, University of Glasgow, United Kingdom
b
School of Education, Leuphana University of Lueneburg, Germany & University of Zurich, Switzerland
c
School of Education, Aarhus University, Denmark
h i g h l i g h t s
Teacher autonomy and school culture influence collective teacher innovativeness.

Teacher autonomy and collaboration are positively interrelated.
A collaborative culture promotes teachers' participation in professional learning.
Collective participation in professional learning enhances innovativeness.
a r t i c l e i n f o a b s t r a c t
Article history: The current article examines the effects of teacher classroom autonomy and school culture on collective
Received 17 August 2020 teacher innovativeness, through integrated professional learning. This examination draws upon an
Received in revised form analysis of TALIS dataset 2018 gathered from 241 426 teachers of 15 672 schools in 48 OECD countries,
14 July 2021
using a multilevel structural equation modelling approach. The article underscores the importance of
Accepted 15 July 2021
enhancing teachers' sense of classroom autonomy and collaborative school culture to promote collective
Available online xxx
teacher innovativeness. It suggests that a collaborative culture would encourage teachers to participate in
integrated professional learning activities, and that collective participation in professional learning
Keywords:
Collaborative culture
would enhance teachers’ collective innovativeness.
Innovativeness © 2021 Published by Elsevier Ltd.
Multilevel structural equation modelling
Professional learning
Teacher autonomy
Teaching and learning international survey
(TALIS)
1. Introduction functions as a critical factor influencing teacher innovative behav-

iours and a prerequisite to innovation and change (Kern & Graber,
Innovation in education is critical in promoting positive changes 2018). Research has explored innovativeness at the individual
for improvement and sustainable development in schools and teacher level and implementation of specific innovations and re-
beyond (Serdyukov, 2017). Teachers’ innovative behaviours are forms (e.g., Akar, 2019). This line of literature has evidenced and
central to implementation of innovations in schools (McGeown, conceptualised important issues around teacher change. However,
1980; Thurlings, Evers, & Vermeulen, 2015). Innovative behaviour it is argued that the field would benefit from more studies that
is defined as a self-initiated process of generating, promoting and centrally discuss collective innovativeness in schools (Buske, 2018;
realising change (Thurlings et al., 2015). Teacher innovativeness, Moolenaar et al., 2014; see also the review of; Newman, Round,
which is conceptualised as teacher receptivity, openness, and Wang, & Mount, 2020). This first argument for the current article
willingness to adopt change (Fullan, 2007; Goldsmith, 1986), is highly credible since collective innovativeness is of paramount
importance in implementation, sustainment, and spreading of in-
novations in schools (Buske, 2018; Schwabsky, Erdogan, &
Tschannen-Moran, 2019; Moolenaar et al., 2014).
* Corresponding author.
A limited albeit growing number of empirical studies have
E-mail address: dong.nguyen@glasgow.ac.uk (D. Nguyen).
https://doi.org/10.1016/j.tate.2021.103463
0742-051X/© 2021 Published by Elsevier Ltd.
D. Nguyen, M. Pietsch and S. Gümüş Teaching and Teacher Education 106 (2021) 103463
identified important factors potentially influencing teacher inno- culture and teacher professional learning. This consideration of
vativeness and innovative behaviours, including teacher autonomy relevant literature informs development of the conceptual frame-
defined as teachers’ degree of given professional discretion (Buske, work for this study.
2018) and school culture (Gilad-Hai & Somech, 2016). These studies
have been conducted in a few single-nation contexts such as Ger- 2.1. Collective teacher innovativeness
many (Buske, 2018) and Israeli (Gilad-Hai & Somech, 2016). To
further advance the current literature, secondly it is argued to Innovation has been viewed as a new thing such as a product, an
require sophisticated analyses of robust datasets collected across a idea, a method and the process of introducing and implementing
number of countries. Such analyses would forward a holistic un- that new thing (Gopalakrishnan & Damanpour, 1997). This process
derstanding of the conditions for collective teacher innovativeness. involves three stages: ideation, implementation, and a change as a
Teachers' behaviours and attitudes towards innovative changes result of implementing that innovation (Serdyukov, 2017). In school
for improving professional practices develop along the process of settings, an innovation may be a new method or strategy in
their participation in professional learning, as discussed further in pedagogy or assessment that aims to enhance teaching and
the subsequent sections. Professional learning may occur in formal learning quality. Innovation is a crucial factor that contributes to
learning activities (e.g., attending a university-based professional enhanced teaching quality and student learning and school im-
development course) and job-embedded professional learning ac- provements (Cohen-Vogel, Cannata, Rutledge, & Socol, 2016;
tivities in the workplace (e.g., in-depth professional conversations). Serdyukov, 2017; Zimmer et al., 2017).
A synthesis of literature suggests that teacher autonomy (Kwak- Innovativeness in schools is conceptualised as teachers' recep-
man, 2003) and school culture (Jurasaite-Harbison & Rex, 2010; tivity, adoption, and internalisation of innovative changes and their
Opfer & Pedder, 2011) influence teacher's participation in profes- continual participation in change-related professional activities
sional learning activities, and that teacher professional learning is (McGeown, 1980). Innovativeness can be defined and measured at
likely to promote teacher innovativeness (Clarke & Hollingsworth, the individual level (Agarwal & Prasad, 1998; Akar, 2019) and col-
2002; Guskey, 2002; Timperley, Wilson, Barrar, & Fung, 2007). The lective level (e.g., team or organisation) (Buske, 2018; Moolenaar
findings and implications from these separated studies highlight et al., 2014; Schwabsky et al., 2019). Innovativeness at the indi-
professional learning as a potential mediator in the relationships vidual level refers to an individual teacher's openness to new ideas
between teacher autonomy and school culture and teacher inno- while collective innovativeness can be understood as a team's or an
vativeness. While a great deal of research has focused exclusively organisation's receptivity and readiness for change (Buske, 2018).
on either formal or job-embedded professional learning (e.g., The current article centres on collective innovativeness of teachers
Jurasaite-Harbison & Rex, 2010; Nguyen & Ng, 2020), researching in schools.
integrated professional learning within a single study is argued to Teacher innovativeness is subject to a compound of individual
provide added value to the evidence base (Darling-Hammond, and organisational factors. Firstly, the theoretical models of teacher
Hyler, & Gardner, 2017; Garet, Porter, Desimone, Birman, & Yoon, change (Clarke & Hollingsworth, 2002; Guskey, 1986; Timperley
2001; Parise & Spillane, 2010; Shirrell, Hopkins, & Spillane, 2019). et al., 2007) suggest that teachers' innovativeness might result
Integrated professional learning in this article involves a combined from the process of their professional learning and experimenta-
form of formal and job-embedded professional learning of tion of new practices. Clarke and Hollingsworth (2002) highlight
teachers. that teachers’ change, of varying degrees, inevitably occurs through
Built on the three aforementioned arguments, this article ex- their participation in professional learning activities. Secondly and
amines the effects of teacher classroom autonomy and collabora- theoretically, there are some arguments for the benefits of sup-
tive school culture on collective teacher innovativeness and looks at porting teacher autonomy in enhancing innovativeness and prox-
the potential role of integrated professional learning in mediating imal factors such as professional motivation and satisfaction (e.g.,
these effects. This examination was based on an analysis of the Cribb & Gewirtz, 2007; Deci & Ryan, 2008). Thirdly and empiri-
database of the OECD Teaching and Learning International Survey cally, the factors pertaining to school climate and culture such as an
2018 (TALIS 2018) gathered in 48 countries. Three underlying emphasis on academic excellence (Schwabsky et al., 2019), open-
research questions which guided this analysis are presented as ness and support for change (McGeown, 1980; Moolenaar et al.,
follows. 2014), and social support among teachers and school leaders
(Gilad-Hai & Somech, 2016) have been evidenced as enablers of
(1) What are the relationships among teacher autonomy, teacher innovativeness.
collaborative culture, and collective teacher innovativeness? In summary, the literature has identified teacher autonomy,
(2) How does integrated professional learning mediate the re- collaborative school culture and teacher professional learning as
lationships between teacher autonomy and collaborative potential factors influencing teacher innovativeness. The subse-
school culture and collective teacher innovativeness? quent parts elaborate on these major factors.
(3) How much variance of findings from Research Questions 1
and 2 above is explained at the teacher and school level? 2.2. Teacher autonomy
The article advances an understanding, of the international Professional autonomy is a complex concept that refers to
scope, on factors influencing collective teacher innovativeness and multiple levels: individual autonomy, collective autonomy, and
the mediational role of integrated professional learning in schools. institutional autonomy (Cribb & Gewirtz, 2007). At the individual
This understanding would provide evidence-based insights into level, autonomy conceptually concerns two aspects of degree of
promoting collective teacher innovativeness that encourages discretion (Dikilitaş & Mumford, 2019; Evans & Fischer, 1992; Lamb
change for improvement in schools. & Reinders, 2008) and self-governance and capacity to exercise that
professional discretion (Benson, 2010; Cribb & Gewirtz, 2007).
2. Literature review From the first perspective, Evans and Fischer (1992) define auton-
omy as “the amount of freedom a worker has to schedule their work
The current section presents the literature relevant to collective and to determine the procedures to be used in carrying it out” (p.
teacher innovativeness, teacher autonomy, collaborative school 1171). This view of centrally linking autonomy to the freedom of
2
choice to exercise professional practices has been widely recog- 2.3. Collaborative school culture
nised in the literature (e.g., Dikilitaş & Mumford, 2019; Lamb &
Reinders, 2008). It implies that individuals have a high degree of In the literature, there are two overlapping terminologies per-
control over specific domains of professional tasks, processes, and taining to teaching and learning environments: school culture and
outcomes that is exercised based on their moral and ethical school climate. School culture has its origins in anthropology while
judgements (Evans & Fischer, 1992). An evolving perspective views school climate is derived from the discipline of psychology
individual autonomy as both freedom and internal capacity to make (Aldridge & Fraser, 2016). Van Houtte (2005) views school culture
decisions and to act independently (Benson, 2010; Cribb & Gewirtz, as a factor of school climate while Schoen and Teddlie (2008)
2007; Mausethagen & Mølstad, 2015; Worth & Van den Brande, conceptualise school climate as a level of school culture.
2020). Research has classified domains or dimensions to measure each
In terms of dimensionality, Breaugh (1985) delineates autonomy construct. For example, Johnson, Stevens, and Zvoch (2007) pro-
into three dimensions related to work method, scheduling, and pose four dimensions of school climate: collaboration, interper-
evaluation of outcomes. Breaugh’s (1985) first two dimensions sonal relationships of staff and students, decision-making, and
pertain to professionals' freedom towards their work method and instructional innovation. Wang and Degol (2016) characterise
processes, while the third dimension refers to their capacity to school climate as a multidimensional construct that incorporates
decide on evaluation criteria of work performance. Specifically four dimensions, namely safety, community, academic, and insti-
considering teacher classroom autonomy, Vangrieken, Grosemans, tutional environment. The dimension of community includes fac-
Dochy, and Kyndt (2017) distinguish two aspects: didactical- tors of interpersonal relationships of staff and students and
pedagogical autonomy (e.g., in preparing lessons, teaching opportunities for participative decision-making in schools. Schoen
methods, student assignment, and managing student behaviour) and Teddlie (2008) incorporate four dimensions to measure this
and curricular autonomy (e.g., decisions about curriculum, setting construct: professional orientation, organisational structure, qual-
goals for students). Teacher autonomy in the current article refers ity of the learning environment, and student-centred focus. The
to the degree of teachers' discretion or freedom in making decisions dimension of professional orientation involves a degree of teachers’
on their course content, selection of teaching methods, assessment collaborative professionalism centred on student learning. The
of student learning, student discipline issues, and the amount of dimension of organisational structure includes factors of leadership
homework for their students, as defined by OECD (2019; 2020). styles (e.g., shared leadership), communication, and processes (e.g.,
There are theoretical perspectives or normative arguments for decision-making).
the positive effects of individual autonomy (e.g., Deci & Ryan, 2008; These typologies suggest school climate and school culture are
Little, 1995). The self-determination theory of Deci and Ryan (2008) multifaceted constructs and inextricably interrelated. These two
outlines the prominent factors influencing two types of employees' constructs clearly share some mutual dimensions regarding social
motivation. The factor of autonomy (i.e., discretion in professional systems (e.g., interpersonal relationships and interactions in a
decision-making and actions) positively influences an employee's school), teacher collaboration, and school processes (e.g., decision-
intrinsic motivation, which in turn influences job satisfaction, making). The current article adopts the term of collaborative cul-
retention and performance (Deci & Ryan, 2008). Little (1995) ar- ture (Carpenter, 2015; Lieberman, 1990) to refer to a school culture
gues that teachers with a high sense of professional autonomy tend of shared decision-making and responsibility, teacher collabora-
to be active and innovative in promoting student autonomy. Cribb tion, and social support in schools.
and Gewirtz (2007) highlight teacher autonomy as a source of job Research has argued for some effects of collaborative culture on
satisfaction, well-being, experimentation and innovation, and professional learning and school innovation in schools (e.g., Geijsel,
effectiveness. Granting teachers professional autonomy would Sleegers, Stoel, & Krüger, 2009; Goddard, Goddard, & Tschannen-
promote “creativity and innovation” in the countries that have “a Moran, 2007; Somech, 2010). Geijsel et al. (2009) suggest that
well-prepared and independent teaching workforce” (OECD, 2020, participative decision-making (as a characteristic of collaborative
p. 32). culture) has a positive, indirect effect on teacher professional
Some of these positive perspectives on teacher autonomy have learning activities through enhancing teacher self-efficacy and
been empirically verified in a few contexts (e.g., Skaalvik & internalising school goals into teacher personal goals. Goddard
Skaalvik, 2014; Worth & Van den Brande, 2020). Teachers' per- et al. (2007) suggest that teacher collaboration in a positive cul-
ceptions of professional autonomy are positively linked with their ture might foster professional learning for improved classroom
job satisfaction (Dou, Devos, & Valcke, 2017; Skaalvik & Skaalvik, instruction. Nguyen and Ng (2020) assert that teacher collaboration
2014; Worth & Van den Brande, 2020) and their professional is a crucial form or setting for ongoing professional learning and
commitment and engagement (Skaalvik & Skaalvik, 2014). Buske that sustained professional learning requires leadership advocacy
(2018), based on an analysis of survey data from 896 teachers of and a supportive school culture with positive teacher
15 vocational schools in Germany, stated that principal support and interrelationships.
teacher autonomy in making decisions on classroom practices At the organisational level, Somech (2010) proposes an analyt-
positively influence teachers’ innovativeness. ical framework suggesting the impacts of teacher participation in
The literature has outlined a variety of factors influencing school-level decision making on school productivity and innova-
teachers' perceptions of their professional autonomy such as policy tion. Somech’s (2010) analytical framework was constructed from a
context (Benson, 2000; Lundstro €m, 2015), school structure review of comprehensive literature, yet, of inconclusive or incon-
(Benson, 2010), and teacher attitudes (Pearson & Hall, 1993). sistent findings, and therefore remains open for empirical verifi-
Research has also suggested ways to promote teachers' effective cation across contexts.
enactment of their autonomy, within prevailing constraints, such as
strengthening development of teachers’ informed professional 2.4. Teacher professional learning
decision-making in teacher education programmes (e.g., Manzano
Vazquez, 2018) and supporting teacher research (e.g., Dikilitaş & Teacher professional learning is a critical contributor to
Mumford, 2019; Wang & Zhang, 2014). enhancing teaching quality, securing student learning, and pro-
moting educational reforms and improvements (Bakkenes,
Vermunt, & Wubbels, 2010; Darling-Hammond et al., 2017;
3
Lieberman & Pointer Mace, 2008; Parise & Spillane, 2010). Teacher example, Leithwood et al. (1998) accentuate a collaborative
learning is an individual and social process that occurs through organisational culture characterised by mutual teacher support and
ongoing participation and practice (Avalos, 2011; Lieberman & receptivity of teacher-led innovations as a positive factor influ-
Pointer Mace, 2008). encing teacher professional learning. Jurasaite-Harbison and Rex
Teacher professional learning includes activities of formal (2010) identify key characteristics of a school conducive to
learning such as attending professional development workshops teacher job-embedded professional learning that has a positive
and courses (Garet et al., 2001) and those of job-embedded learning culture promoting professional interactions and collaboration be-
such as peer observation and mentoring (Parise & Spillane, 2010). tween and among teachers and school leaders.
The formal, traditional model provides teachers with relatively Leadership at different levels has been evidenced to influence
structured professional development (e.g., courses and pro- professional learning. The roles of principal leadership (e.g.,
grammes) delivered by university teachers and researchers outside Nguyen, Ng, Luo, & Mansor, 2020; Printy, 2008; Robinson, Lloyd, &
their schools. This model of traditional teacher learning is grounded Rowe, 2008), middle leadership (e.g., Bryant, Wong, & Adames,
in the ideology of supporting teachers with “knowledge for prac- 2020; Harris, Michelle, Nashwa, & Nguyen, 2019), and teacher
tice” (Cochran-Smith & Lytle, 1999) from the formal knowledge leadership (e.g., Nguyen, Harris, & Ng, 2019; Nguyen & Ng, 2020;
base (i.e., knowledge generated from research). Teachers are ex- Stoll, Bolam, McMahon, Wallace, & Thomas, 2006) in promoting
pected to acquire this formal knowledge and implement it in their teacher professional learning in schools have been documented in
schools (Vescio, Ross, & Adams, 2008). the extant literature.
Job-embedded professional learning tends to be conceptualised At the individual level, the intensity and quality of professional
as “a socially constructed process of learning and development that learning is dependent on the factors such as perceived teacher job
is largely embedded in activities that occur inside of schools” satisfaction and autonomy, teacher efficacy, professional commit-
(Hallinger & Kulophas, 2019, p. 4). This type of professional learning ments, prior experiences and knowledge, and beliefs of teaching
aligns with the theoretical perspectives of situated learning and learning (Kwakman, 2003; Liu & Hallinger, 2018; Opfer &
(Camndy, 1991; Darling-Hammond, 1998; McLaughlin, 1997; Pedder, 2011).
Putnam & Borko, 2000). These perspectives imply that job-
embedded teacher learning occurs as a result of teachers' partici- 3. Conceptual framework
pation in professional activities and interactions in their schools
and occasionally beyond (Kwakman, 2003). Research has suggested The literature discussed above informs development of the
that both formal and job-embedded professional learning impacts conceptual framework or model for the current study. Fig. 1 pro-
teachers’ instructional practices (e.g. Buckler, Cordingley, & poses relationships among the variables of focus in the article,
Timperley, 2009; Cordingley, 2015; Garet et al., 2001; Parise & namely teacher autonomy, collaborative culture, integrated teacher
Spillane, 2010; Shirrell et al., 2019). professional learning, and teacher team innovativeness. This model
The literature has documented noteworthy features of effective examines the relationships among these variables at both the in-
teacher professional learning (Garet et al., 2011; Hiebert, 1999; King dividual teacher and school levels.
& Newmann, 2001; Lieberman & Pointer Mace, 2008). These The constructs of teacher autonomy and collaborative culture
include explicit focus on content knowledge, curriculum, and are proposed as two independent variables in this model. This
classroom instruction (Garet et al., 2011; King & Newmann, 2001); proposal is based on the extant literature regarding the possible
ongoing collaboration within and between schools (Darling- effects of teacher autonomy and school culture on teacher learning
Hammond et al., 2017; Lieberman & Pointer Mace, 2008); sus- and teacher innovativeness and change (Cribb & Gewirtz, 2007;
tained opportunities for active learning (Darling-Hammond et al., Geijsel et al., 2009; Goddard et al., 2007; Somech, 2010). The cur-
2017; Garet et al., 2011; King & Newmann, 2001); access to the rent article speculates that teacher autonomy and collaborative
expertise of external researchers and coaches (King & Newmann, culture are not mutually exclusive; these two variables are corre-
2001); coherence between learning activities (Garet et al., 2011); lated. This speculation is grounded upon the conceptual framework
and appropriate duration of learning (Darling-Hammond et al., of Vangrieken et al. (2017) that suggests interrelationships between
2017; Garet et al., 2011). These features underscore the impor- teacher autonomy and collaborative school culture. In addition, one
tance of ongoing, sustained, collaborative learning activities and aspect of collaborative culture conceptualised in this article refers
value of learning from teacher peers as well as from external to participative school decision-making that was found to be
coaches and scholars through professional development events positively related to teachers’ feeling of autonomy (see Lu, Jiang, Yu,
(e.g., courses). & Li, 2015).
Opportunities for teacher professional learning need to be This article takes an integrated approach to conceptualise
deliberately orchestrated, nurtured and sustained. Teacher profes- teacher professional learning (Garet et al., 2001; Parise & Spillane,
sional learning is susceptible to a multiplicity of factors at the 2010; Shirrell et al., 2019). More specifically, integrated professional
system, organisational, and individual levels. Formal continuing learning refers to teacher participation in both formal learning
professional development is contractually specified in many na- opportunities such as courses and workshops outside their schools
tional systems such as Russia and Singapore (see Harris, Jones, and job-embedded workplace learning through, for example, peer
Adams, Perera, & Sharma, 2014). In these systems, teachers are observation and peer coaching. Integrated professional learning is
normally supported financially to participate in formal continuing posited as a mediating variable for three reasons. Firstly, there are a
professional development events such as courses and workshops. number of studies (e.g., Jurasaite-Harbison & Rex, 2010; Kwakman,
Some systems also aim to systematically promote job-embedded 2003; Leithwood et al., 1998; Opfer & Pedder, 2011) that accentuate
professional learning through establishing formal professional school culture and climate as a major factor influencing teacher
learning communities in all schools (see Nguyen & Ng, 2020). professional learning, as discussed in the previous section. Sec-
At the organisational level, research has evidenced the effects of ondly, Kwakman (2003), in a survey design study of 10 secondary
school culture, structure, and leadership on teacher professional schools in the Netherlands, showed some positive effects of teacher
learning (Bektaş, Kılınç, & Gümüş, 2020; Gümüş, 2013; Jurasaite- autonomy in decisions around the pace, method, and order of work
Harbison & Rex, 2010; Kwakman, 2003; Leithwood, Leonard, & on their participation in professional learning activities though
Sharratt, 1998; Opfer & Pedder, 2011; Parise & Spillane, 2010). For these effects are small and thus need to be verified. Thirdly, there
4
Fig. 1. Conceptual model of relationships between teacher autonomy, collaborative culture, integrated professional learning, and collective teacher innovativeness.
has been evidence on the effects of teachers’ professional learning information concerning teachers’ professional development activ-
on their innovativeness and change (Clarke & Hollingsworth, 2002; ities within 12 months prior to the time of data collection (TT3G22A
Guskey, 2002; Timperley et al., 2007). to TT3G22I) (see Appendix A). The core question was “During the
In brief, the conceptual model of this article proposes teacher last 12 months, did you participate in any of the following develop-
autonomy and collaborative culture as two independent variables, ment activities?“. Three items (TT3G22A, TT3G22B, and TT3G22D)
collective teacher innovativeness as a dependent variable, and in- measured formal learning activities. The remaining items
tegrated professional learning as a mediating variable. This is a (TT3G22C, TT3G22E, TT3G22F, TT3G22G, TT3G22H, TT3G22I)
partial mediation model because it proposes that the effects of measured job-embedded learning activities. For each of the nine
teacher autonomy and collaborative school culture on collective items, the participants were required to choose “yes” or “no” that
teacher innovativeness are both direct and indirect (see Baron & reflected their experience. In this article, integrated professional
Kenny, 1986; Liu & Hallinger, 2020). The next section presents learning refers to teacher participation in both formal and job-
measurement of these variables. embedded learning opportunities. This variable was coded as 0/1
to indicate if a teacher took part in at least one (out of the six) job-
embedded and one (out of the three) formal professional learning
4. Design and method
activity during the last 12 months.
The two independent variables were labelled as Teacher Au-
4.1. Data and sample
tonomy (u ¼ :96) and Collaborative Culture (u ¼ :82). The latent
variable of Teacher Autonomy was based on five items (TT3G40A to
The data are derived from the OECD Teaching and Learning In-
TT3G40D) from the original scale of “Satisfaction with Target Class
ternational Survey 2018 (TALIS 2018; OECD, 2019). We used the
Autonomy” (T3SATAT) (see Appendix A). The core question was
teacher public use file (TTGINT3) and the principal public use file
“How strongly do you agree or disagree that you have control over the
(TCGINTT3). The data were gathered cross-sectionally between
following areas of your planning and teaching in this < target
September 2017 and November 2018 across n ¼ 49 countries, using
class > ?“.
a standardised teacher questionnaire. The public use file contains
The latent variable of Collaborative Culture was based on three
data from 48 countries as Iceland has no public available data. The
items from the original scale of “Participation among Stakeholders”
study, therefore, analysed the dataset collected from n ¼ 241 426
(see Appendix A). These items (TT3G48A, TT3G48D, TT3G48E)
teachers, nested in n ¼ 15 672 schools in these 48 countries.
measured the extent to which teachers participate in school-level
decision-making processes and collaborate for school issues. All
4.2. Measures items were answered on a four-point Likert-type scale, ranging
from “strongly disagree” to “strongly agree”. The core question was
The dependent variable in the conceptual model of this article “How strongly do you agree or disagree with these statements, as
was labelled as Collective Teacher Innovativeness (u ¼ :90). This is a applied to this school?“.
latent variable based on four indicators (TT3G32A, TT3G32B, To account for potentially confounding effects, we added com-
TT3G32C, TT3G32D) from the original TALIS scale of “Team Inno- mon demographic variables provided in the dataset of TALIS 2018
vativeness” (T3TEAM) (see Appendix A). It measured teachers’ to the model of this article. The control variables at the teacher level
perceptions of the extent to which teachers in their schools search are:
for, develop and apply new ideas in their practices (OECD, 2019). All
items were answered on a four-point Likert-type scale, ranging a) gender (TT3G01, coded: 1: female, 2: male),
from “strongly disagree” to “strongly agree”. b) age (TINNOVATIVENESS, coded: 1: under 25, 2: 25e29, 3:
The mediating variable is a binary variable, labelled as Integrated 30e39, 4: 40e49, 5: 50e59, 6: 60 and above), and
Professional Learning. This is a dummy variable based on the
5
c) teaching experience (TT3G11B, metric). information provided by teachers while they represent deviations
from the schools’ mean (the average school ratings) at level one due
At the school level, the control variables are: to implicit group-mean centring of constructs inherent in MPLUS
(Marsh et al., 2012). The dummy variable of Integrated Professional
a) school size, as indicated by the total number of students Learning, at level two, represents the percentage of teachers within
enrolled within a school (NENRSTUD, metric); a school that took part in professional learning activities while it
b) publicly- or privately-managed school (TC3G12, coded: 1: represents, at level one, the individual learning experience of the
public, 2: private); surveyed teachers. Both the level one and level two components
o c) school location e collapsed (SCHLOC, coded: 1: Rural (up to remain meaningful in and of themselves (Morin et al., 2014).
3000 people), 2: Town (3001e100 000 people), 3: City (more Comparing the group-mean centred estimates to grand-mean
than 100 000 people); and centred estimates enabled us to calculate the difference between
d) the target populations' ISCED level (IDPOP, coded: 1: ISCED 1 e level two effects and level one effects.
primary education, 2: ISCED 2 e secondary education 3: ISCED 3 The two latent constructs of collaborative culture and collective
e tertiary education, 8: PISA-link), which we recoded into a teacher innovativeness refer to the school level. These constructs are
binary variable (primary education ¼ 1, secondary reflective measures and more meaningful at level two than those at
education ¼ 2). level one since they go behind the individual experiences and
represent teachers' shared perceptions of their schools (see Bureau,
Table 1 presents Mean values, SD values, and a correlation ma- Gagne , Morin, & Mageau, 2017). At level one, these constructs tend
trix of all variables at both levels of analysis. Correlations among to be less meaningful but still interpretable since they support
these variables do not reveal any possible issues of multi- analyses of the effects of each teacher's own individual perception.
collinearity, as shown in Table 1. Appendix A presents a full list of The other latent construct of Teacher Autonomy refers to the indi-
scales and items used in the current study. vidual teacher and thus is a formative measure at the school level
(see Marsh et al., 2009).
4.3. Analytical approach To evaluate model fit, we used three classic fit indices:
comparative fit index (CFI), root mean square error of approxima-
All analyses were conducted by using Maximum Likelihood tion (RMSEA) and standardised root mean square residual (SRMR),
(ML) in MPLUS 8.3 (Muthe n & Muthe n, 2019). Missing data were provided by MPLUS 8.3, to estimate the fit of the models. CFI over
processed by applying the implemented Full Information 0.90, a RMSEA less than 0.08 and a SRMRw/SRMRb less than 0.08
Maximum Likelihood Estimators (FIML). The amount of missing are generally interpreted as acceptable fit (e.g., Hu & Bentler, 1999;
data in the model variables ranged from zero to 18.7 per cent. We Marsh, Hau, & Wen, 2004).
applied a multilevel structural equation model (MSEM) to the data We reported the standardised latent regression coefficients
for two reasons: firstly, teachers are nested in schools, and secondly together with their corresponding standard errors, as shown in
we were interested testing a partial mediation model as noted Table 3. The standardised latent regression coefficients can be
above. We fitted a doubly-latent contextual model that combines a interpreted as level-specific effect sizes.
structural equation model and a multilevel model. This approach
allowed us to control for measurement and sampling errors in the 4.4. Preliminary verification of multilevel assumptions
aggregation of individual ratings into school-level constructs to the
data (full correction approach; see Lüdtke, Marsh, Robitzsch, & When applying multilevel models, it is a good practice to check
Trautwein, 2011). To facilitate interpretation and limit multi- if there is enough variance at all levels of the proposed model to
collinearity, all indicator variables were standardised (M ¼ 0, conduct analyses and to test the interrater reliability of aggregates
SD ¼ 1) prior to the estimation (Marsh et al., 2012; Morin, Marsh, measured at lower levels. Our analysis explicitly, however, controls
Nagengast, & Scalas, 2014). All (non-essential) model covariates for the level of (dis)agreement among teachers in their ratings of
were defined either as within or between variables and within school constructs due to the estimation process of the doubly-
covariates were grand-mean centred. Appendix B presents an latent models (see Arens & Morin, 2016). Thus, we computed the
example of MPLUS inputs. intraclass correlations in terms of ICC1 (i.e., the proportion of
We simultaneously estimated all latent variables and model between-group variance to the total variance) and ICC2 (i.e., the
paths at both the teacher level (level one) and school level (level reliability of a schools mean) of all model variables, as shown in
two). The latent variables at level two represent aggregates of Table 2 below. Following the recommendations by Bliese (2000),
Table 1
Correlation matrix of all analysed variables.
Mean SD Correlations
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
1. Teacher Collective Innovativeness 2.99 .64 1/1

2. Teacher Autonomy 3.74 .77 .06/.32 1/1
3. Collaborative Culture 2.91 .60 .44/.71 .01/.20 1/1
4. Integrated Professional Learning 0.78 .39 .01/.34 .01/.04 .05/.32 1/1
5. Teacher Age (within only) 3.89 1.15 .09/- -.02/- .03/- -.03/- 1/-
6. Teacher Gender (within only) 1.31 .46 -.03/- -.02/- .00/- -.03/- .02/- 1/-
7. Experience as Teacher (within only) 17.20 10.52 .08/- -.01/- .03/- -.01/- .79/- -.02/- 1/-
8. School Type (between only) 1.94 .62 ¡/¡.19 ¡/¡.04 ¡/¡.15 -/.09 / / / -/1
9. Rural-Urban Profile (between only) 2.25 .73 ¡/¡.04 ¡/¡.07 ¡/¡.14 /.01 / / / -/.03 -/1
10. Public or Private School (between only) 1.22 .42 -/.08 -/.01 ¡/¡.09 -/.05 / / / ¡/¡.03 -/.24 -/1
11. School Size (between only) 2.78 1.41 ¡/¡.09 /.02 ¡/¡.20 -/.03 / / / -/.13 -/.39 -/.14 -/1
Note: All bold correlations e left: within, right: between e are statistically significant at p < .05.
6
Table 2 (Collaborative Culture: 0.82; Teacher Autonomy: 0.87; Collective

ICCs, reliability and interrater agreement of main variables. Teacher Innovativeness: 0.87). Table 2 presents ICC and rwg(j) values
ICC1 ICC2 rWG(J) Omega for all main variables, along with omega reliability values.
Collective Teacher Innovativeness .156 .756 .871 .901
To examine if all latent variables were measured in the same
Collaborative Culture .204 .811 .819 .820 way at both levels of analysis, we finally tested for multilevel
Teacher Autonomy .120 .695 .873 .964 invariance by constraining all factor loadings equally across levels
Integrated Professional Learning .134 .721 NIL NIL and fixing the between level residual variances to zero (Guenole,
Note. ICC ¼ Intraclass correlation coefficients. 2018; Jak, Oort, & Dolan, 2013). With c2 ¼ 62018.165(df ¼ 147),
p < .001, CFI ¼ 0.97, RMSEA ¼ 0.04, SRMRw ¼ 0.02 and
SRMRb ¼ 0.11. This model fitted the data well. Freeing the residual
ICC1-values greater than 0.05 and ICC2-values greater than 0.70 variance at the school level resulted in a slightly better model fit
indicate that the given construct is a group-level construct, and (c2 ¼ 21939.876(df ¼ 135), p < .001, CFI ¼ 0.99, RMSEA ¼ 0.03,
thus, a multilevel modelling approach should be used (see Lüdtke SRMRw ¼ 0.02 and SRMRb ¼ 0.08). Thus, all factor loadings can be
et al., 2008). Further, for all measures, we estimated rwg(j)- considered equal, relatively invariant across levels (Jak et al., 2013).
indexes (i.e., a measure of the interrater agreement among teach-
ers within schools; see James, Demaree, & Wolf, 1984), where:
5. Results
(i) values ranging from 0 to 0.59 suggest an unacceptable level of

In the first step, we evaluated the fit of the proposed model to
interrater agreement,
the data of TALIS 2018. The fit indexes showed that the model and
(ii) values between 0.60 and 0.69 indicate a weak level of
the data coincided well (c2 ¼ 20677:383(df ¼ 190), p < .001,
interrater agreement,
CFI ¼ 0.99, RMSEA ¼ 0.02, SRMRw ¼ 0.01, SRMRb ¼ 0.04). In the
(iii) values between 0.70 and 0.79 indicate a moderate level of
second step, we inspected the path coefficients among the model
interrater agreement, and
variables at both levels. This section presents results from analyses
(iv) values equal or greater than 0.80 indicate as a strong level of
at the teacher (level one) and school levels (level two) (see Table 3).
interrater agreement (Brown & Hauenstein, 2005).
5.1. Results from analyses at teacher level (level one)

In the current study, all of the ICC1-values were rather high with
Collaborative Culture (ICC1 ¼ 0.204) and Collective Teacher Innova-
Analyses at the teacher level found a strong positive relationship
tiveness (ICC1 ¼ 0.156). Both of these latent variables, referring to
between collaborative culture and collective teacher innovative-
the school level, show the highest amount of variance at the school
ness (b ¼ .433, SE ¼ 0.005, p < .001). It meant that teachers' per-
level among the main variables. Teacher Autonomy, a latent variable
ceptions of collaborative culture in their school were significantly
referring to the individual teacher (ICC1 ¼ 0.120), demonstrates the
associated with those of their school's collective innovativeness.
lowest amount of variance at the school level. The ICC1 for the
Teacher autonomy was identified to be a positive predictor of
mediating variable of Integrated Professional Learning was 0.134,
collective teacher innovativeness (b ¼ 0.016, SE ¼ 0.003, p < .001).
which suggested that about 87% of the variance is located within
This suggested that teachers’ individual perceptions of own au-
schools. Considering the reliability of the variable means at the
tonomy influenced their perceptions of collective innovativeness in
school level, all ICC2-values were satisfactory (Collaborative Culture:
their school.
ICC2 ¼ 0.811; Collective Teacher Innovativeness: ICC2 ¼ 0.756; Teacher
Teacher autonomy and collaborative culture were significantly
Autonomy: ICC2 ¼ 0.695; Integrated Professional Learning:
related to each other at level one (r ¼ 0.099, SE ¼ 0.003, p < .001).
ICC2 ¼ 0.721). In addition, the rwg(j)-indexes suggested a strong
While teachers’ individual perceptions of collaborative culture
level of agreement among teacher ratings within schools
increased, the perceptions of their own autonomy tended to
Table 3
Results of the multilevel structural equation model for collective teacher innovativeness.
b (S.E) p
Within (teacher level)

Collaborative Culture - > Collective Teacher Innovativeness .433 .005 .000
Teacher Autonomy - > Collective Teacher Innovativeness .016 .003 .000
Integrated Professional Learning - > Collective Teacher Innovativeness -.011 .002 .000
Collaborative Culture - > Integrated Professional Learning - > Collective Teacher Innovativeness -.001 .000 .000
Teacher Autonomy - > Integrated Professional Learning - > Collective Teacher Innovativeness .000 .000 .124
R2 .199
Collaborative Culture - > Integrated Professional Learning .049 .003 .000
Teacher Autonomy - > Integrated Professional Learning .004 .003 .101
R2 .004
Between (school level)
Collaborative Culture - > Collective Teacher Innovativeness .641 .011 .000
Teacher Autonomy - > Collective Teacher Innovativeness .179 .011 .000
Integrated Professional Learning - > Collective Teacher Innovativeness .133 0.11 .000
Collaborative Culture - > Integrated Professional Learning - > Collective Teacher Innovativeness .046 .004 .000
Teacher Autonomy - > Integrated Professional Learning - > Collective Teacher Innovativeness -.004 .002 .059
R2 .584
Collaborative Culture - > Integrated Professional Learning .348 .013 .000
Teacher Autonomy - > Integrated Professional Learning -.027 .014 .053
R2 .115
Notes. 1. The model has been controlled for by the following variables: Age, experience, and gender at individual teacher level as well as school size, type of school, ISCED level, and
school location at school level two. Results from the ML-SEM, direct and indirect effects (standardised).
7
increase, and vice-versa. teacher innovativeness. Firstly, school size was positively associated
Teacher autonomy was not significantly related to integrated with collective teacher innovativeness (b ¼ 0.026, SE ¼ 0.009,
professional learning (b ¼ 0.004, SE ¼ 0.003, p > .05). Teacher au- p < .001). Teachers’ perceptions of collective innovativeness in their
tonomy, therefore, had no indirect effects on collective teacher school increased with the size of the school.
innovativeness at level one. Secondly, school type was significantly related to collective
Collaborative school culture had a significant and positive teacher innovativeness. This indicated that teachers working in
relationship with integrated professional learning (b ¼ 0.049, private schools had significantly higher perceptions of their
SE ¼ 0.003, p < .001). Integrated professional learning was found to schools’ collective innovativeness (b ¼ 0.135, SE ¼ 0.009, p < .001)
be related negatively to teachers' perceptions of innovativeness in than those working in public schools.
their school (b ¼ 0.011, SE ¼ 0.002, p < .001). As noted earlier, Thirdly, school location was positively associated with collective
collective teacher innovativeness in this study was measured by teacher innovativeness (b ¼ 0.018, SE ¼ 0.009, p < .05). Teachers
teachers’ perceptions about the innovativeness of their school as a working at schools located in larger cities had higher perceptions of
whole. This result implied that teachers who got involved in inte- their schools’ collective innovativeness.
grated professional learning, including both formal and job- Fourthly, there was a negative relationship between schools’
embedded professional learning activities, tended to perceive ISCED level and collective teacher innovativeness (b ¼ 0.094,
their school as less innovative. These results showed a statistically SE ¼ 0.012, p < .001). Secondary schools had a lower degree of
significant indirect effect of collaborative school culture on collec- collective innovativeness as compared with primary schools, ac-
tive teacher innovativeness (b ¼ 0.001, SE ¼ 0.000, p < .001) cording to the perceptions of teachers.
though this effect was negative and very small. In the final step, we checked for appropriate contextual effects
Controlling variables on the dependent variable at level one, we of integrated professional learning and teacher autonomy on col-
found that both age (b ¼ 0.066, SE ¼ 0.004, p < .001) and experience lective teacher innovativeness by subtracting the level one from the
(b ¼ 0.018, SE ¼ 0.004, p < .001) were positively associated with level two effect and standardising these differences (see Morin
teachers’ perceptions of collective innovativeness. Male teachers et al., 2014). This additional parameter is identical to what would
tended to rate the degree of collective innovativeness in their have been obtained under a grand-mean centring procedure. It
schools (b ¼ 0.033, SE ¼ 0.002, p < .001) lower than female showed that a unique and statistically significant effect at the
teachers. school level was observable regarding professional learning
(b ¼ 0.083, p¼<.001) and teacher autonomy (b ¼ 0.047, p¼<.001).
5.2. Results from analyses at school level (level two)
6. Discussion and conclusions
The analyses at the school level found that collaborative culture
was significantly associated with collective teacher innovativeness The Results section highlights four prominent findings of this
(b ¼ 0.641, SE ¼ 0.011, p < .001). Those teachers who worked in a article. Firstly, teachers' perceptions of their classroom autonomy
highly collaborative school environment tended to be more inno- and those of collaborative culture in their school were significantly
vative as perceived by their colleagues. related to each other. Secondly, both teacher autonomy and
Similarly, the overall perception of teacher autonomy was collaborative culture were significantly and directly associated with
significantly associated with that of collective teacher innovative- teachers’ perceptions of collective innovativeness in their schools at
ness (b ¼ 0.179, SE ¼ 0.011, p < .001). This could be interpreted that both the teacher (level one) and school (level two) levels. Thirdly, no
teachers' rating of the degree of their own classroom autonomy in a significant indirect effects of teacher autonomy on collective
school (as average of all teachers’ individual perceptions) was teacher innovativeness, mediated by integrated teacher profes-
slightly related to their collective innovativeness in that school. sional learning, were found in this study at any level. Fourthly, the
At level two, teacher autonomy was not significantly associated results showed a statistically significant indirect effect of collabo-
with integrated teacher professional learning (b ¼ 0.027, rative culture on collective teacher innovativeness through inte-
SE ¼ 0.014, p > .05). There was no indirect relationship between grated professional learning at both the teacher and school levels.
teacher autonomy and collective teacher innovativeness at level While integrated professional learning mediated the indirect effect
two, which was similar to the results at level one. of collaborative culture on collective teacher innovativeness posi-
On the contrary, collaborative culture was found to be signifi- tively at the school level, this mediation effect tended to be negative
cantly related to integrated professional learning (b ¼ .348, at the individual teacher level. Before discussing major implications
SE ¼ 0.013, p < .001). The current study also found a significant of these findings, it is important to highlight some of the limitations
positive relationship between integrated teacher professional and to suggest further research to complement this study.
learning at the school level and collective teacher innovativeness
(b ¼ .133, SE ¼ 0.011, p < .001). In other words, collective teacher 6.1. Limitations and suggestions for future research
innovativeness tended to be considerably higher when the overall
participation in professional learning increased as a whole in that The dataset of TALIS 2018 used in this article was based on
school, i.e., more teachers in the same school participated in both teachers’ perceptions as a single source of data and collected at one
formal and job-embedded professional learning activities. These time. The retrospective and self-perceptual nature of the data
results highlighted a statistically significant positive indirect effect might limit the validity of some measures used in this study, given
of collaborative school culture on collective innovativeness potential recall biases (see Schwarz, 2007). A part of the variance in
(b ¼ 0.046, SE ¼ 0.004, p < .001). the model might be attributable to the measurement method
In addition, there was a significant relationship between teacher rather than to the constructs measured because a potential single
autonomy and collaborative school culture (r ¼ .194, SE ¼ 0.012, source bias is likely to threaten “the validity of the conclusions
p < .001). While the degree of perceived collaborative school cul- about the relationship between measures” (Podsakoff, MacKenzie,
ture increased, teachers’ overall perception on their autonomy Lee, & Podsakoff, 2003, p. 879). While this cross-sectional study
increased, and vice versa. does not allow strong claims on temporary causality, it supports
All four controlling variables (i.e., school size, school type, school testing a mediational effect from a statistical perspective that
location, and ISCED level) were significantly related to collective highlights possible mechanisms of effects of teacher autonomy and
8
collaborative culture on collective teacher innovativeness (see between and among teachers and other actors in schools do not
Baron & Kenny, 1986; David & Sava, 2015). A good combination of preclude teachers’ classroom autonomy. Furthermore, two of the
future time-series, longitudinal, and in-depth case study research three items in the scale of collaborative culture in this study aimed
would complement this study in modelling and evaluating the to investigate the extent of participative, shared decision-making in
potential dynamics and reciprocal relationships among relevant schools. Teacher participation in school decision-making processes
variables and providing more insights into collective is likely to give teachers a sense of more professional control that is
innovativeness. the core characteristic of autonomy (Dikilitaş & Mumford, 2019;
As noted in the Methods section, the current article draws upon Evans & Fischer, 1992). Similarly, teachers with a greater sense of
an analysis, using sophisticated statistical methods, of available classroom autonomy tend to be more open to get involved and
large-scale data. Each construct in this article was measured by one share responsibilities for school-wide issues (Buske, 2018). These
single scale. Although the scales used in this article are supported results allow a theoretical implication that autonomy and collabo-
by the literature as discussed in the previous sections, it is unlikely ration are two distinguishable albeit interrelated constructs.
that they cover all factors or facets of each construct. For example, Despite the interconnectedness between teacher autonomy and
the construct of teacher autonomy in this study measured teachers’ collaborative culture, their pathways and extent of effects on col-
perceptions of five domains within classroom autonomy, and thus lective teacher innovativeness are differentiated. The statistical
excluded the other levels and domains of autonomy (see Salokangas results suggest a positive albeit small effect of teacher classroom
& Wermke, 2020). This article follows a good practice of providing autonomy on collective teacher innovativeness at both the teacher
clarity on which factors, variables, and items were examined. and school levels. Teachers with higher perceptions of their own
Future research could build on this article to measure each classroom autonomy tend to rate their school's collective innova-
construct in more depth and comprehensiveness. tiveness stronger. The contextual effect of teacher autonomy on
In addition, this article centres on investigating the effect collective innovativeness, as identified at the school level, suggests
mechanisms of teacher autonomy, collaborative culture, and pro- that working with colleagues who have higher perceptions of their
fessional learning on collective teacher innovativeness, given the own classroom autonomy would increase teachers' collective
complexity of the current conceptual model. Future research could innovativeness perceptions of their school. This finding implies the
examine the effects of other potential factors, for example, school need of defending or enhancing teacher autonomy in making de-
leadership (Robinson et al., 2008), teacher efficacy (Schwabsky cisions on the issues of subject content, instructional methods, and
et al., 2019) and trust (Schwabsky et al., 2019) on collective inno- assessment. The degree of autonomy given to teachers should be
vativeness in school, using TALIS datasets and beyond. sensitive to that of teachers' readiness and each country's context
The scope and space of this article justify a choice of focusing on (OECD, 2020). This implication is aligned with the literature on the
sophisticated analyses of the TALIS data as an aggregated, decon- benefits of teacher autonomy in promoting educational initiatives
textualised dataset across 48 countries. This approach is similar to and teacher job satisfaction, commitment, and professional
that of some previous articles (e.g., Fackler & Malmberg, 2016; engagement (Cribb & Gewirtz, 2007; Deci & Ryan, 2008; Skaalvik &
Vieluf, Kunter, & Van de Vijver, 2013). For example, Fackler and Skaalvik, 2014; Worth & Van den Brande, 2020).
Malmberg (2016) analysed TALIS 2007 data collected from 14 It is noteworthy that the analyses at both the teacher and school
OECD countries to examine the factors influencing teacher self- levels show no significant effects of teachers' perceived classroom
efficacy. While the approach of the current study extends the autonomy on their participation in professional learning activities
geographical locus of previous research on teacher innovativeness, within and outside their schools. Interpretation of this finding
subsequent studies could further leverage OECD datasets by con- should take two methodological factors into account. Firstly, the
ducting methodologically comparative analyses around this topic scale of teacher autonomy in TALIS 2018 focuses on measuring
between countries and regions. teachers’ perceptions of autonomy in the issues of their course
All controlling variables at each level of analysis in this study, as content, instructional methods, and assessment, and thus excludes
reported in the Results section, were found to be significantly the issues beyond classroom level that may influence their partic-
related to teachers’ perceptions of collective innovativeness in their ipation in professional learning such as teacher autonomy in
schools. Although the current survey data offer no firm basis for any participating professional development types and activities. Sec-
meaningful explanations of such relationships, these findings may ondly, this article centres on an integration of formal and job-
provide clues for future research focusing on the interactions of embedded teacher professional learning, as elaborated in the pre-
demographic factors and school types, location, and level and vious sections. An alternative approach of measuring these two
teacher innovativeness. types of professional learning distinctively may result in a different
pattern of findings. For example, Kwakman (2003) found a small,
6.2. Interpretation and implications of findings positive effect of teacher professional autonomy on their partici-
pation in job-embedded professional learning in schools.
The current article supports the aforementioned proposition The analyses at both the teacher and school levels show that the
that teacher classroom autonomy and collaborative school culture extent of direct, positive effects of collaborative school culture on
are positively interconnected. The empirical literature has docu- collective teacher innovativeness is substantially greater than that
mented classroom autonomy and collaboration as important of teacher classroom autonomy. To put it simply, it is implied that
stimulants to teachers' positive affective factors such as teacher teachers are more receptive to innovations in schools with a strong
professional engagement, teacher job satisfaction and motivation, culture of teacher collaboration and participation in school-wide
and teacher efficacy (Dou et al., 2017; Scheopner, 2010; Skaalvik & decision-making processes. This finding is logically meaningful
Skaalvik, 2011, 2014; Strong & Yoshida, 2014). Nevertheless, there and critical in promoting implementation of initiatives in schools. It
has been a tendency to conceptualise teacher autonomy and reinforces the vitality of building and maintaining a positive culture
collaboration as two opposite constructs: autonomy tends to be of mutual support, cognitively and affectively, in promoting initia-
connoted with independence or individualism (Street & Licata, tion, implementation and diffusion of innovations (Nguyen et al.,
1989) that is theoretically opposite to the interdependence of 2019; Nguyen & Ng, 2020).
collaboration (Vangrieken et al., 2017). The results at both the This article evidences complexities in the pathways that inte-
teacher and school levels show that a culture of collaboration grated professional learning mediates the relationship between
9
collaborative culture and collective teacher innovativeness. Teach- Acknowledgements

ers' integrated professional learning would mediate the relation-
ship between collaborative school culture and collective teacher The authors wish to thank the University of Glasgow, a DFG
innovativeness positively at the school level but negatively at the Heisenberg grant to Marcus Pietsch (PI 618/4-1, Quantitative School
teacher level. More specifically, the analyses at the school level Leadership Research), and Aarhus University for supporting this
indicate that teachers’ average rating of collaborative culture in a research.
school is positively related to their participation in integrated
professional learning, which in turn is positively associated with
their average perceptions of innovativeness of their school. Appendix A. Items and scales used in the current study
The contextual effect of integrated professional learning, as re-
ported in the Results section, suggests that teachers working with Teacher Autonomy
colleagues who have a higher level of participation in professional
learning activities tend to be more positive about the degree of TT3G40: How strongly do you agree or disagree that you have
their schools' collective innovativeness. The analyses at the teacher control over the following areas of your planning and teaching in
level, however, show that teachers' perception of collaborative this < target class>?
school culture is positively linked with their participation in inte-
grated professional learning, but teachers’ individual participation
in integrated professional learning is negatively related to their
perceptions of collective innovativeness. Response options: “Strongly disagree” (1), “Disagree” (2), “Agree” (3), “Strongly
These results could be interpreted as firstly, the collaborative agree” (4).
culture of a school would encourage more teachers of the school to Items:
participate in professional learning activities, which would increase TT3G40A. Determining course content
TT3G40B. Selecting teaching methods
collective teacher innovativeness; and secondly, teachers' individ- TT3G40C. Assessing students' learning
ual participation in integrated professional learning would have no TT3G40D. Disciplining students
significant effects or a minimal, negative effect on their perceptions TT3G40E. Determining the amount of homework to be assigned
of collective teacher innovativeness in their school. This negative Collaborative Culture
TT3G48: How strongly do you agree or disagree with these statements, as
effect could be attributable to the possibilities: (i) that a few
applied to this school?
teachers exposed to innovative ideas as a result of their individual Response options: “Strongly disagree” (1), “Disagree” (2), “Agree” (3), “Strongly
participation in learning opportunities feel the need of enhancing agree” (4).
the collective innovativeness of their own school; or (ii) that those Items:
individual teachers face their colleagues’ resistance to their in- TT3G48A. This school provides staff with opportunities to actively participate in
school
novations inspired from their participation in professional learning decisions.
activities. Although this interpretation could be specifically verified TT3G48D. This school has a culture of shared responsibility for school issues.
in future research, it does imply the benefit of collective partici- TT3G48E. There is a collaborative school culture which is characterised by
pation of groups of teachers from the same school in professional mutual
support.
learning activities to promote collective teacher innovativeness.
Integrated Professional Learning
Such collective participation would be helpful in establishing TT3G22: During the last 12 months, did you participate in any of the following
receptivity and sustain implementation of innovations in a school professional development activities?
since teachers are exposed to the common types and content of Response options: “Yes” (1), “No” (2)
learning opportunities and more likely to reach a consensus about Items:
TT3G22A. In-person courses seminars
an innovative approach for improvement (Desimone, 2009; Garet
TT3G22B. Online courses seminars
et al., 2001). Collective teacher participation in job-embedded TT3G22C. Education conferences
professional learning is potentially promoted through developing TT3G22D. Formal qualification programme
inclusive professional learning communities within and across TT3G22E. Observation visits to other schools.
TT3G22F. Observation visits to business premises.
schools (Nguyen & Ng, 2020; Stoll et al., 2006) and, where possible,
TT3G22G. Peer and/or self-observation and coaching.
organising effective professional development courses and work- TT3G22H. Participation in a network of teachers formed specifically for the
shops for groups of teachers from the same school (Garet et al., professional
2001). development of teachers.
In conclusion, the current article calls attention for developing TT3G22I. Reading professional literature.
Collective Teacher Innovativeness
and maintaining a positive culture of collaboration, mutual support,
TT3G32: Thinking about the teachers in this school, how strongly do you agree
collective teacher participation in professional learning, and or disagree with the following statements?
respecting teacher classroom autonomy while maintaining a good Response options: “Strongly disagree” (1), “Disagree” (2), “Agree” (3), “Strongly
level of support for teacher development, to promote collective agree” (4).
Items:
innovativeness in schools. Collective teacher innovativeness is a
TT3G32A. Most teachers in this school strive to develop new ideas for teaching
construct of international relevance and a critical condition for and
transforming schools into innovative organisations that nurture learning.
creativity of teachers and students. Looking forwards, more TT3G32B. Most teachers in this school are open to change.
rigorous empirical and theoretical research relevant to collective TT3G32C. Most teachers in this school search for new ways to solve problems.
TT3G32D. Most teachers in this school provide practical support to each other
teacher innovativeness should be developed timely to establish a
for the application of new ideas.
strong knowledge base to support innovative teaching and learning
in schools.
10
Appendix B. Example of MPLUS Syntax (continued )
auton_b by TT3G40B (402);

auton_b by TT3G40C (403);
auton_b by TT3G40D (404);
auton_b by TT3G40E (405);
Title: Collective Teacher Innovativeness
clima_b by TT3G48A@1 (501);
Data: File is talis.dat;
clima_b by TT3G48D (504);
Format is f10.0, f8.0, f6.0, 21f1.0, f3.0;
clima_b by TT3G48E (505);
Variable: Names are
inno_b on devcom (b901);
IDTEACH IDSCHOOL IDCNTRY
inno_b on auton_b (b903);
TT3G01
inno_b on clima_b (b905);
TT3G32A TT3G32B TT3G32C TT3G32D
devcom on auton_b (b910);
devcom on clima_b (b911);
TT3G40E
inno_b on IDPOP SCHLOC TC3G12 NENRSTUD;
TT3G48A TT3G48D TT3G48E
auton_b on IDPOP SCHLOC TC3G12 NENRSTUD;
devcom
clima_b on IDPOP SCHLOC TC3G12 NENRSTUD;
IDPOP
devcom on IDPOP SCHLOC TC3G12 NENRSTUD;
SCHLOC
auton_b with clima_b;
TC3G12
inno_b (varop);
NENRSTUD
model indirect:
TCHAGEGR
inno_b ind auton_b;
devinfo
inno_b ind clima_b;
devform
inno_w ind auton_w;
TT3G11B;
inno_w ind clima_w;
within ¼ TCHAGEGR TT3G01 TT3G11B;
model constraint:
between ¼ IDPOP SCHLOC TC3G12 NENRSTUD;
new(c_effect);
Usevariables are
c_effect ¼ b901-w901;
new (devc_w);
new (devc_b);
TT3G40E
devc_w ¼ 0.574;
TT3G48A TT3G48D TT3G48E
devc_b ¼ 0.112;
IDPOP SCHLOC TC3G12 NENRSTUD
new(c_effect_stdxy);
TCHAGEGR TT3G01 TT3G11B
c_effect_stdxy ¼ c_effect*(sqrt(varop)/sqrt(devc_w þ devc_b));
devcom;
new (c1_effect);
IDVariable is IDTEACH;
c1_effect ¼ b903-w903;
Cluster is IDSCHOOL;
new (aut_w);
Missing are all (7, 8, 9, 997e999);
new (aut_b);
Define:
aut_w ¼ 0.652;
Standardize
aut_b ¼ 0.090;
new(c1_effect_stdxy);
c1_effect_stdxy ¼ c1_effect*(sqrt(varop)/sqrt(aut_w þ aut_b));
TT3G40E TT3G48A TT3G48D TT3G48E;
Output: Tech1 Tech4 Tech8 Standardised;
centering ¼ Grandmean(TCHAGEGR TT3G01 TT3G11B);
Analysis:
Processors ¼ 3;
Type ¼ twolevel; References
Model:
%within% Agarwal, R., & Prasad, J. (1998). A conceptual and operational definition of personal
inno_w by TT3G32A@1 (101); innovativeness in the domain of information technology. Information Systems
inno_w by TT3G32B (102); Research, 9(2), 204e215.
inno_w by TT3G32C (103); Akar, S. G. M. (2019). Does it matter being innovative: Teachers' technology
inno_w by TT3G32D (104); acceptance. Education and Information Technologies, 24(6), 3415e3432.
Aldridge, J. M., & Fraser, B. J. (2016). Teachers' views of their school climate and its
auton_w by TT3G40A@1 (401);
relationship with teacher self-efficacy and job satisfaction. Learning Environ-
auton_w by TT3G40B (402);
ments Research, 19(2), 291e307.
auton_w by TT3G40C (403); Arens, A. K., & Morin, A. J. (2016). Relations between teachers' emotional exhaustion
auton_w by TT3G40D (404); and students' educational outcomes. Journal of Educational Psychology, 108(6),
auton_w by TT3G40E (405); 800.
clima_w by TT3G48A@1 (501); Avalos, B. (2011). Teacher professional development in teaching and teacher edu-
clima_w by TT3G48D (504); cation over ten years. Teaching and Teacher Education, 27(1), 10e20.
clima_w by TT3G48E (505); Bakkenes, I., Vermunt, J. D., & Wubbels, T. (2010). Teacher learning in the context of
inno_w on devcom (w901); educational innovation: Learning activities and learning outcomes of experi-
inno_w on auton_w (w903); enced teachers. Learning and Instruction, 20(6), 533e548.
inno_w on clima_w (w905); Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in
devcom on auton_w (w910); social psychological research: Conceptual, strategic, and statistical consider-
devcom on clima_w (w911); ations. Journal of Personality and Social Psychology, 51, 1173e1182.
Bektaş, F., Kılınç, A.Ç., & Gümüş, S. (2020). The effects of distributed leadership on
inno_w on TCHAGEGR TT3G01 TT3G11B;
teacher professional learning: Mediating roles of teacher trust in principal and
auton_w on TCHAGEGR TT3G01 TT3G11B;
teacher motivation. Educational Studies, 1e23. https://doi.org/10.1080/
clima_w on TCHAGEGR TT3G01 TT3G11B; 03055698.2020.1793301
devcom on TCHAGEGR TT3G01 TT3G11B; Benson, P. (2010). Teacher education and teacher autonomy: Creating spaces for
auton_w with clima_w; experimentation in secondary school English language teaching. Language
!model indirect: Teaching Research, 14(3), 259e275.
!inno_w ind devcom; Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability:
%between% Implications for data aggregation and analysis. In K. J. Klein, & S. W. J. Kozlowski
inno_b by TT3G32A@1 (101); (Eds.), Multilevel theory, research, and methods in organizations (pp. 349e381).
inno_b by TT3G32B (102); San Francisco, CA: Jossey-Bass.
inno_b by TT3G32C (103); Breaugh, J. A. (1985). The measurement of work autonomy. Human Relations, 38,
inno_b by TT3G32D (104); 551e570.
auton_b by TT3G40A@1 (401); Brown, R. D., & Hauenstein, N. M. A. (2005). Interrater agreement reconsidered: An
alternative to the rwg indices. Organizational Research Methods, 8(2), 165e184.
(continued on next column)
11
Bryant, D. A., Wong, Y. L., & Adames, A. (2020). How middle leaders support in- Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure
service teachers' on-site professional learning. International Journal of Educa- analysis: Conventional criteria versus new alternatives. Structural Equation
tional Research, 100. https://doi.org/10.1016/j.ijer.2019.101530 Modeling, 6, 1e55.
Buckler, N., Cordingley, P., & Temperley, J. (2009). Professional learning and the role of Jak, S., Oort, F. J., & Dolan, C. V. (2013). A test for cluster bias: Detecting violations of
the coach in the new masters in teaching and learning (MTL); technical report measurement invariance across clusters in multilevel data. Structural Equation
(Coventry: CUREE [Online]). Modeling: A Multidisciplinary Journal, 20(2), 265e282.
Bureau, J. S., Gagne , M., Morin, A. J., & Mageau, G. A. (2017). Transformational Johnson, B., Stevens, J. J., & Zvoch, K. (2007). Teachers' perceptions of school climate:
leadership and incivility: A multilevel and longitudinal test. Journal of Inter- A validity study of scores from the revised school level environment ques-
personal Violence. doi: 10.117/0886260517734219. tionnaire. Educational and Psychological Measurement, 67(5), 833e844.
Buske, R. (2018). The principal as a key actor in promoting teachers' Jurasaite-Harbison, E., & Rex, L. A. (2010). School cultures as contexts for informal
innovativenesseanalyzing the innovativeness of teaching staff with variance- teacher learning. Teaching and Teacher Education, 26(2), 267e277.
based partial least square modeling. School Effectiveness and School Improve- Kern, B. D., & Graber, K. C. (2018). Understanding teacher change: A national survey
ment, 29(2), 262e284. of US physical educators. Research Quarterly for Exercise and Sport, 89(1), 80e90.
Carpenter, D. (2015). School culture and leadership of professional learning com- King, M. B., & Newmann, F. M. (2001). Building school capacity through professional
munities. International Journal of Educational Management, 29(5), 682e694. development: Conceptual and empirical considerations. International Journal of
Clarke, D., & Hollingsworth, H. (2002). Elaborating a model of teacher professional Educational Management, 15(2), 86e94.
growth. Teaching and Teacher Education, 18(8), 947e967. Lamb, T. E., & Reinders, H. (2008). Learner and teacher autonomy: Concepts, realities
Cochran-Smith, M., & Lytle, S. (1999). Relationships of knowledge and practice: and responses. Amsterdam: John Benjamins.
Teacher learning in communities. Review of Research in Education, 24, 249e305. Leithwood, K., Leonard, L., & Sharratt, L. (1998). Conditions fostering organizational
Cohen-Vogel, L., Cannata, M., Rutledge, S., & Socol, A. R. (2016). A model of learning in schools. Educational Administration Quarterly, 34(2), 243e276.
continuous improvement in high schools: A process for research, innovation Lieberman, A. (1990). Schools as collaborative cultures: Creating the future now.
design, implementation, and scale. Teachers College Record, 116(13), 1e26. Bristol, PA: The Falmer Press, Taylor and Francis.
Cribb, A., & Gewirtz, S. (2007). Unpacking autonomy and control in education: Some Lieberman, A., & Pointer Mace, D. H. (2008). Teacher learning: The key to educa-
conceptual and normative groundwork for a comparative analysis. European tional reform. Journal of Teacher Education, 59(3), 226e234.
Educational Research Journal, 6(3), 203e213. Little, D. (1995). Learning as dialogue: The dependence of learner autonomy on
Darling-Hammond, L., Hyler, M. E., & Gardner, M. (2017). Effective teacher profes- teacher autonomy. System, 23(2), 175e181.
sional development. Palo Alto, CA: Learning Policy Institute. Liu, S., & Hallinger, P. (2018). Principal instructional leadership, teacher self-efficacy,
David, D., & Sava, F. A. (2015). Designs for studying mediation. In R. Cautin, & and teacher professional learning in China: Testing a mediated-effects model.
S. Lilienfeld (Eds.), The encyclopedia of clinical psychology. Hoboken, NJ: Wiley- Educational Administration Quarterly, 54(4), 501e528.
Blackwell. https://doi.org/10.1002/9781118625392.wbecp516. Lüdtke, O., Marsh, H. W., Robitzsch, A., & Trautwein, U. (2011). A 2 2 taxonomy of
Deci, E. L., & Ryan, R. M. (2008). Self-determination theory: A macrotheory of hu- multilevel latent contextual models: Accuracyebias trade-offs in full and partial
man motivation, development, and health. Canadian Psychology, 49, 182e185. error correction models. Psychological Methods, 16(4), 444.
Desimone, L. M. (2009). Improving impact studies of teachers' professional devel- Lüdtke, O., Marsh, H. W., Robitzsch, A., Trautwein, U., Asparouhov, T., & Muthe n, B.
opment: Toward better conceptualizations and measures. Educational (2008). The multilevel latent covariate model: A new, more reliable approach to
Researcher, 38(3), 181e199. group-level effects in contextual studies. Psychological Methods, 13, 203e229.
Dikilitaş, K., & Mumford, S. E. (2019). Teacher autonomy development through Lu, J., Jiang, X., Yu, H., & Li, D. (2015). Building collaborative structures for teachers'
reading teacher research: Agency, motivation and identity. Innovation in Lan- autonomy and self-efficacy: The mediating role of participative management
guage Learning and Teaching, 13(3), 253e266. and learning culture. School Effectiveness and School Improvement, 26(2),
Dou, D., Devos, G., & Valcke, M. (2017). The relationships between school autonomy 240e257.
gap, principal leadership, teachers' job satisfaction and organizational Lundstro €m, U. (2015). Teacher autonomy in the era of new public management.
commitment. Educational Management Administration & Leadership, 45(6), Nordic Journal of Studies in Educational Policy, 2015(2), 73e85.
959e977. Manzano Va zquez, B. (2018). Teacher development for autonomy: An exploratory
Evans, B. K., & Fischer, D. G. (1992). A hierarchical model of participatory decision- review of language teacher education for learner and teacher autonomy.
making, job autonomy, and perceived control. Human Relations, 45(11), Innovation in Language Learning and Teaching, 12(4), 387e398.
1169e1189. Marsh, H. W., Hau, K.-T., & Wen, Z. (2004). In search of golden rules: Comment on
Fackler, S., & Malmberg, L. E. (2016). Teachers' self-efficacy in 14 OECD countries: hypothesis-testing approaches to setting cut-off values for fit indexes and
Teacher, student group, school and leadership effects. Teaching and Teacher dangers in overgeneralizing Hu and Bentler's (1999) findings. Structural Equa-
Education, 56, 185e195. tion Modeling, 11, 320e341.
Fullan, M. (2007). The new meaning of educational change (4th ed.). New York, NY: Marsh, H. W., Lüdtke, O., Nagengast, B., Trautwein, U., Morin, A. J., Abduljabbar, A. S.,
Teachers College Press. et al. (2012). Classroom climate and contextual effects: Conceptual and meth-
Garet, M. S., Porter, A. C., Desimone, L., Birman, B. F., & Yoon, K. S. (2001). What odological issues in the evaluation of group-level effects. Educational Psychol-
makes professional development effective? Results from a national sample of ogist, 47(2), 106e124.
teachers. American Educational Research Journal, 38(4), 915e945. Marsh, H. W., Lüdtke, O., Robitzsch, A., Trautwein, U., Asparouhov, T., Muthe n, B.,
Geijsel, F., Sleegers, P. J. C., Stoel, R., & Krüger, M. L. (2009). The effect of teacher et al. (2009). Doubly-latent models of school contextual effects: Integrating
psychological, school organizational and leadership factors on teachers' pro- multilevel and structural equation approaches to control measurement and
fessional learning in Dutch schools. The Elementary School Journal, 109(4), sampling error. Multivariate Behavioral Research, 44(6), 764e802.
406e427. Mausethagen, S., & Mølstad, C. E. (2015). Shifts in curriculum control: Contesting
Gilad-Hai, S., & Somech, A. (2016). “The day after” the organizational consequences ideas of teacher autonomy. Nordic Journal of Studies in Educational Policy,
of innovation implementation in experimental schools. Journal of Educational 2015(2), 30e41.
Administration, 54(1), 19e40. McGeown, V. (1980). Dimensions of teacher innovativeness. British Educational
Goddard, Y. L., Goddard, R. D., & Tschannen-Moran, M. (2007). A theoretical and Research Journal, 6(2), 147e163.
empirical investigation of teacher collaboration for school improvement and Moolenaar, N. M., Daly, A. J., Cornelissen, F., Liou, Y. H., Caillier, S., Riordan, R., &
student achievement in public elementary schools. Teachers College Record, Cohen, N. A. (2014). Linked to innovation: Shaping an innovative climate
109(4), 877e896. through network intentionality and educators' social network position. Journal
Goldsmith, R. E. (1986). Convergent validity of four innovativeness scales. Educa- of Educational Change, 15(2), 99e123.
tional and Psychological Measurement, 46, 81e87. Morin, A. J., Marsh, H. W., Nagengast, B., & Scalas, L. F. (2014). Doubly latent
Gopalakrishnan, S., & Damanpour, F. (1997). A review of innovation research in multilevel analyses of classroom climate: An illustration. The Journal of Exper-
economics, sociology and technology management. Omega, 25(1), 15e28. imental Education, 82(2), 143e167.
Guenole, N. (2018). Reconsidering cluster bias in multilevel data: A Monte Carlo Muthe n, B. O., & Muthe n, L. K. (2019). Mplus (version 8.3.) [Computer software]. Los
comparison of free and constrained baseline approaches. Frontiers in Psychology, Angeles, CA: Muthe n & Muthe n.
9, 255. Newman, A., Round, H., Wang, S., & Mount, M. (2020). Innovation climate: A sys-
Gümüş, S. (2013). The effects of teacher-and school-level factors on teachers tematic review of the literature and agenda for future research. Journal of
participation in professional development activities: The role of principal Occupational and Organizational Psychology, 93(1), 73e109.
leadership. Journal of International Education Research, 9(4), 371e380. Nguyen, D., Harris, A., & Ng, D. (2019). A review of the empirical research on teacher
Guskey, T. R. (1986). Staff development and the process of teacher change. Educa- leadership (2003-2017): Evidence, patterns and implications. Journal of
tional Researcher, 15(5), 5e12. Educational Administration, 58(1), 60e80.
Hallinger, P., & Kulophas, D. (2019). The evolving knowledge base on leadership and Nguyen, D., & Ng, D. (2020). Teacher collaboration for change: Sharing, improving,
teacher professional learning: A bibliometric analysis of the literature, 1960- and spreading. Professional Development in Education, 46(4), 638e651.
2018. Professional Development in Education, 46(4), 521e540. Nguyen, D., Ng, D., Luo, W., & Mansor, S. (2020). Exploring the relationships be-
Harris, A., Jones, M., Adams, D., Perera, J., & Sharma, S. (2014). High performing tween instructional leadership and teacher competences: Singapore primary
education systems in asia: Leadership art meets implementation science. Asia- school teachers' perceptions. International Journal of Leadership in Education.
Pacific Education Researcher, 23(4), 861e869. https://doi.org/10.1080/13603124.2020.1731765
Harris, A., Jones, M., Ismail, N., & Nguyen, D. (2019). Middle leaders and middle OECD. (2019). TALIS 2018 technical report. https://www.oecd.org/education/talis/
leadership in schools: Exploring the knowledge base (2003e2017). 39 pp. TALIS_2018_Technical_Report.pdf on 20.05.2020.
255e277). School Leadership & Management, 3e4. OECD. (2020). TALIS 2018 insights and interpretations. https://www.oecd.org/
12
education/talis/TALIS2018_insights_and_interpretations.pdf. on 20.05.2020. Stress & Coping, 24(4), 369e385.

Opfer, V. D., & Pedder, D. (2011). Conceptualizing teacher professional learning. Skaalvik, E. M., & Skaalvik, S. (2014). Teacher self-efficacy and perceived autonomy:
Review of Educational Research, 81(3), 376e407. Relations with teacher engagement, job satisfaction, and emotional exhaustion.
Parise, L. M., & Spillane, J. P. (2010). Teacher learning and instructional change: How Psychological Reports, 114(1), 68e77.
formal and on-the-job learning opportunities predict change in elementary Somech, A. (2010). Participative decision making in schools: A mediating-
school teachers' practice. The Elementary School Journal, 110(3), 323e346. moderating analytical framework for understanding school and teacher out-
Pearson, L. C., & Hall, B. C. (1993). Initial construct validation of the teaching au- comes. Educational Administration Quarterly, 46(2), 174e209.
tonomy scale. The Journal of Educational Research, 86, 172e177. Stoll, L., Bolam, R., McMahon, A., Wallace, M., & Thomas, S. (2006). Professional
Podsakoff, P., MacKenzie, S., Lee, J., & Podsakoff, N. (2003). Common method biases learning communities: A review of the literature. Journal of Educational Change,
in behavioral research: A critical review of the literature and recommended 7(4), 221e258.
remedies. Journal of Applied Psychology, 88, 879e903. Strong, L. E., & Yoshida, R. K. (2014). Teachers' autonomy in today's educational
Printy, S. M. (2008). Leadership for teacher learning: A community of practice climate: Current perceptions from an acceptable instrument. Educational
perspective. Educational Administration Quarterly, 44(2), 187e226. Studies, 50(2), 123e145.
Putnam, R. T., & Borko, H. (2000). What do new views of knowledge and thinking Thurlings, M., Evers, A. T., & Vermeulen, M. (2015). Toward a model of explaining
have to say about research on teacher learning? Educational Researcher, 29(1), teachers' innovative behaviour: A literature review. Review of Educational
4e15. Research, 85(3), 430e471.
Robinson, V. M., Lloyd, C. A., & Rowe, K. J. (2008). The impact of leadership on Timperley, H., Wilson, A., Barrar, H., & Fung, I. (2007). Teacher professional learning
student outcomes: An analysis of the differential effects of leadership types. and development: Best evidence synthesis iteration (BES). Auckland, New Zealand:
Educational Administration Quarterly, 44(5), 635e674. University of Auckland. http://educationcounts.edcentre.govt.nz/goto/BES.
Scheopner, A. (2010). Irreconcilable differences: Teacher attrition in public and Van Houtte, M. (2005). Climate or culture? A plea for conceptual clarity in school
catholic schools. Educational Research Review, 5, 261e277. effectiveness research. School Effectiveness and School Improvement, 16(1),
Schoen, L. T., & Teddlie, C. (2008). A new model of school culture: A response to a 71e89.
call for conceptual clarity. School Effectiveness and School Improvement, 19(2), Vangrieken, K., Grosemans, I., Dochy, F., & Kyndt, E. (2017). Teacher autonomy and
129e153. collaboration: A paradox? Conceptualising and measuring teachers' autonomy
Schwabsky, N., Erdogan, U., & Tschannen-Moran, M. (2019). Predicting school and collaborative attitude. Teaching and Teacher Education, 67, 302e315.
innovation: The role of collective efficacy and academic press mediated by Vescio, V., Ross, D., & Adams, A. (2008). A review of research on the impact of
faculty trust. Journal of Educational Administration, 58(2), 246e262. professional learning communities on teaching practice and student learning.
Schwarz, N. (2007). Retrospective and concurrent self reports. The rationale for Teaching and Teacher Education, 24(1), 80e91.
real-time data capture. In A. A. Stone, S. Shiffman, A. A. Atienza, L. Nebeling, & Vieluf, S., Kunter, M., & Van de Vijver, F. J. (2013). Teacher self-efficacy in cross-
Hrsg (Eds.), The science of real-time data capture (S. 11e26). New York: Oxford national perspective. Teaching and Teacher Education, 35, 92e103.
University Press. Wang, M. T., & Degol, J. L. (2016). School climate: A review of the construct, mea-
Serdyukov, P. (2017). Innovation in education: What works, what doesn't, and what surement, and impact on student outcomes. Educational Psychology Review,
to do about it? Journal of Research in Innovative Teaching & Learning, 10(1), 4e33. 28(2), 315e352.
Shirrell, M., Hopkins, M., & Spillane, J. P. (2019). Educational infrastructure, pro- Wang, Q., & Zhang, H. (2014). Promoting teacher autonomy through
fessional learning, and changes in teachers' instructional practices and beliefs. universityeschool collaborative action research. Language Teaching Research,
Professional Development in Education, 45(4), 599e613. 18(2), 222e241.
Skaalvik, E. M., & Skaalvik, S. (2011). Teachers' feeling of belonging, exhaustion, and Worth, J., & Van den Brande, J. (2020). Teacher autonomy: How does it relate to job
job satisfaction: The role of school goal structure and value consonance. Anxiety, satisfaction and retention? Slough. England: NFER.
13

1 s2.0 S0742051X21001876 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S0742051X21001876 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S0742051X21001876 Main

Uploaded by

Copyright:

Available Formats

Teaching and Teacher Education 106 (2021) 103463

Contents lists available at ScienceDirect

Teaching and Teacher Education

Collective teacher innovativeness in 48 countries: Effects of teacher

Teacher autonomy and school culture inﬂuence collective teacher innovativeness.

1. Introduction functions as a critical factor inﬂuencing teacher innovative behav-

1. Teacher Collective Innovativeness 2.99 .64 1/1

Table 2 (Collaborative Culture: 0.82; Teacher Autonomy: 0.87; Collective

(i) values ranging from 0 to 0.59 suggest an unacceptable level of

5.1. Results from analyses at teacher level (level one)

Within (teacher level)

collaborative culture and collective teacher innovativeness. Teach- Acknowledgements

Appendix B. Example of MPLUS Syntax (continued )

auton_b by TT3G40B (402);

education/talis/TALIS2018_insights_and_interpretations.pdf. on 20.05.2020. Stress & Coping, 24(4), 369e385.

You might also like