Measuring collective efficacy for inclusion in a global context

ABSTRACT Previous research has identified the importance of teacher attitudes and self-efficacy in supporting inclusive education. This study involved a multi-national exploration of a further dimension of inclusive education, collective efficacy, through the testing of a new tool, the Teacher Efficacy for Inclusive Practice-Collective (TEIP-C) Scale. The study also aimed to investigate whether teacher attitudes, self-efficacy, collective efficacy and intention to teach in inclusive classrooms differ across countries. Participants included 1,523 teachers from Canada, Greece, Italy and Switzerland. Results suggested a two-factor structure for the TEIP-C, Engagement, and Inclusive Pedagogies, with strong internal consistency for the scale. Several differences across countries were identified, with teachers from Italy reporting more positive attitudes towards inclusion and a greater intention to teach in inclusive classrooms. Implications of the study in terms of further strengthening inclusive practice are discussed.

Countries around the globe have recognised the value of inclusive education (Ainscow 2020).While some countries now have legislation, others have formulated policies to support the inclusion of learners with diverse abilities in regular classrooms (de Bruin 2019).Although policies and legislation are necessary, they do not provide sufficient guidance on how inclusion could be achieved (Forlin et al. 2013).Research has found that successful implementation of inclusive practices is largely dependent on teachers' commitment or positive attitudes, their teaching efficacy, and the availability of support for educators to include all learners (Ainscow, Booth, and Dyson 2004;Opoku et al. 2020;Sharma and Jacobs 2016;Sharma, Loreman, and Forlin 2012;Sharma et al. 2021).
In addition to these factors, a school's overall collective efficacy to include learners with diverse abilities may be equally critical and requires further investigation.Although the role of collective efficacy in schools more broadly has been investigated (Datnow 2012),

Teaching efficacy for inclusion
Interest in examining educators' self-efficacy beliefs to teach in inclusive classrooms has also seen a significant increase over the years (Wilson, Marks Woolfson, and Durkin 2020;Woodcock et al. 2022).In a recently published systematic review, Wray, Sharma, and Subban (2022)reported that there were 71 studies that have examined teacher efficacy to teach in inclusive classrooms.Findings from these studies suggest a significant association between teacher self-efficacy and the use of inclusive teaching practices (Savolainen, Malinen, and Schwab 2022;Sharma and Sokal 2015;Sharma et al. 2021).There is also growing evidence regarding the impact of a range of factors on teacher self-efficacy for inclusive education, including pre-service teacher education (Loreman, Sharma, and Forlin 2013;Sharma and Sokal 2015), teacher knowledge of inclusive education policy and legislation (Ahsan, Sharma, and Deppeler 2012;Chao, Forlin, and Ho 2016) and other demographic and contextual factors such as teaching experience and age (Subban, Round, and Sharma 2021), school environment (Wilson, Marks Woolfson, and Durkin 2020) and policy context (Cappe et al. 2021;Schwab, Hellmich, and Goerel 2017;Sharma et al. 2018).

Intention to teach in inclusive classrooms
Although it is difficult to measure the use of inclusive practices, it can be predicted based on individual teacher's intention to include all learners in a school.This is consistent with Ajzen's (1991) theory of planned behaviour which suggests that intentions to undertake a specific action are influenced by a range of variables including attitudes, perceived behavioural control and subjective norms, which in turn are highly correlated with actual rates of behaviour.As applied to the inclusive education context, a teacher will have stronger intentions to use inclusive practices if they hold positive attitudes towards inclusion, experience support from school leaders, colleagues and the school culture more broadly and have a high degree of perceived control over their behaviour.Ahmmed, Sharma, and Deppeler (2013) applied the theory of planned behaviour to their study of teacher attitudes, perceived school support and teacher self-efficacy for inclusion on intention to teach in inclusive classrooms in a sample of 738 teachers in Bangladesh, with results suggesting a strong relationship between intentions and teacher attitudes, efficacy and perceived school support.

Collective efficacy for inclusion
Collective efficacy has been defined as 'a group's shared belief in its conjoint capability to organise and execute the courses of action required to produce given levels of attainment' (Bandura 1997, 477), extending Bandura's concept of self-efficacy to a group or organisation.Building on research examining the role of collective efficacy in schools more broadly (Bandura 1993;Datnow 2012), Lyons, Thompson, and Timmons (2016) conducted a qualitative study investigating effective inclusive practices within four Canadian primary schools, with results suggesting that teachers' use of inclusive practices was supported by collaborative teamwork driven by a belief in collective efficacy.Other researchers have examined the role of collective efficacy in supporting teacher self-efficacy and inclusive teaching practices.For example, a recent study by Wilson, Marks Woolfson, and Durkin (2020) identified school environment (including perceptions of collective efficacy) as a significant predictor of teacher self-efficacy for inclusive education in a sample of 148 primary school teachers from Scotland.
Although several validated measures of collective efficacy exist, research in the area of collective efficacy specific to inclusive education has been hindered by a lack of consistency in measurement.Having access to a reliable and valid tool to measure teachers' perceptions of their school's collective efficacy for inclusion would help advance research in this area and support the identification of a possible additional lever for intervention to enable the inclusion of all learners.Given collective efficacy is also likely to be influenced by contextual factors such as the jurisdictional environment, further research investigating its role in inclusive education across countries is also required.

Canada
Inclusive practice varies widely across Canada, with each Canadian province adopting its own set of philosophies, policies, and practices (Loreman 2014).Within each province, education jurisdictions are divided up into what are known as School Districts that may comprise a handful of schools or over 200 schools spread over a geographic area.Each School District may also adopt its own set of inclusive approaches and supports.It is, therefore, not possible to accurately describe inclusion in Canada as being one thing.For the purposes of the current study, a sample of teachers from Alberta and British Columbia were surveyed.Practice in these provinces for children with disabilities tends to continue to be focused on the individual learner through individual programs intended to dovetail with the regular curriculum.However, over the past 10 years increased emphasis has been placed on the use of Universal Design for Learning in order to move away from such an individualistic approach.Some School Districts in Alberta and British Columbia have, to a degree, started to adopt Jennifer Katz's (2013) 'Three Block Model' of UDL as a recommended approach.This is the case with some of the teachers who comprise the sample used in this study.

Greece
Greece, like many other western countries, signed the Salamanca Statement and Framework for Action on Special Needs Education (UNESCO 1994) in relation to the inclusion of students with disabilities (SWD) in regular schools.As a result, the movement of inclusive education has gained significant momentum in Greece over the last thirty years.The progress achieved to date is most vividly illustrated by the steady reduction of placements in special segregated provision and the restructuring of regular schools to become more effective in meeting the needs of an increasingly diverse student population.
Inclusion in Greece is mainly implemented through the operation of 'inclusion units' in regular schools, which involves withdrawing SWD from the regular class for individualised tuition for a maximum of 2 hours per day.Inclusion units are operated by special teachers who either hold a degree in special education or have other formal qualifications in special education.More recently, special teachers have been deployed in schools to provide learning support to individual SWD within their regular class and enable their access to the general curriculum.This development was significant as in-class support ensures that SWD are not deprived from valuable learning experiences alongside their peers.
Notwithstanding the importance of the abovementioned policy initiatives, inclusion in Greece still faces significant obstacles, such as the development of new organisational structures, the differentiation of curriculum, the reforming of the educational environment, a continuing emphasis on individual deficits reflecting a medical conceptualisation of SWD, and a reluctance by teachers to implement innovative inclusive arrangements in their classrooms (Avramidis et al. 2019).The reservations often voiced by Greek teachers could be attributed to their lack of knowledge and skills to effectively meet the needs of a diverse class population, which is further intensified by the shortage of relevant training opportunities (Coutsocostas and Alborz 2010).

Italy
The Italian school context is considered one of the most 'inclusive' with respect to European countries (Sibilio and Aiello 2018).Since 1971, the Italian regulation system has provided a rich framework which imposed the abolition of special classes and the integration of 'handicapped pupils' in regular primary and middle classes.In 1977, the law 517 introduced the role of the 'special needs teacher' with primary responsibility for the participation of students with disabilities in the classroom.The last forty years saw the evolution of the decrees for the implementation of inclusive practices in schools (De Anna, Gaspari, and Mura 2015), including Law 104/1992 which stipulates the full integration of students; and Decree 66 which enhances individual education plans (IEPs).
Since 2011, the government has imposed a compulsory initial and in-service training for special education teachers (Aiello 2019;D'alonzo 2022).The initial certification course is required to be employed with a permanent position as a special education teacher and lasts at least eight months per year: it is in the context of the last two editions of this mandatory certification course that the Italian research team collected data with the Italian sample.Despite the many positive developments, the road to school inclusion is still complicated and impeded in Italy: in the ordinary 'modus of operandi' at school, an individualistic approach still prevails with a focus on strategies to support the learner with disabilities.Some distorted assumptions also continue to exist about the role of special education teachers and the importance of the bio-psycho-social and intersectional perspective on inclusion in practice (Aiello, Sharma, and Sibilio 2016).

Switzerland
Switzerland is a small country with 26 Cantons and about 8.5 million inhabitants.Since 2004, various federal laws and agreements between Swiss Cantons have required that integrative ('inclusion' is explicitly not mentioned) school forms should be preferred to separate school forms (EDK Swiss Conference of Cantonal Ministers of Education 2007).In 2014, Switzerland ratified the UN Convention on the Rights of Persons with Disabilities (UN General Assembly 2006), but without signing the optional protocol allowing individual actions and claims.As a result, the Swiss Federal Council does not interpret disability laws to include any obligation to integrate [or include] children with disabilities (Hollenweger 2016).
Following the development of federal and cantonal legislation as well as demands for more integrative education, there has been a steady reduction in the number of separate special classes for students with so called 'mild learning and behavioural disabilities' since 2005.Paradoxically, however, the number of students in special schools for those with 'severe disabilities' has been increasing in many cantons in over the last few years (Bundesamt für Statistik, 2020; Kronenberg 2021).In addition, it is observed that the number of integrated children with severe disabilities decreases as grade level increases.
In Switzerland, there are separate pathways for the preparation of special and mainstream teachers.The National Minimum Standards for teacher education are set by the Swiss Conference of Cantonal Directors of Education (Criblez 2012).Until 2010, no explicit content on school inclusion was covered in the curricula across Swiss teacher education programs in the country.In response to legal requirements and the increasing integration of children with disabilities, the Swiss Conference of Cantonal Directors of Education in 2016 recommended to cover key aspects about special education/inclusive education in regular teacher education programs (Chamber of Universities of Teacher Education 2016).However, it is important to note that there is no mandatory requirement if and about how university and teacher education institutes embed and cover aspects of special education/inclusive education in their programs.

Current study
Although a growing body of research has examined the role of teacher attitudes and selfefficacy in supporting inclusive education, there has been limited research investigating how collective efficacy can influence a school's overall implementation of inclusive practices.Knowledge of factors that closely relate to the use of inclusive practices (or intention to include all learners) in a school can allow policymakers to identify areas where more resources could be devoted.Research of this nature and that compares the experiences of teachers across countries can also advance the theoretical understanding of key constructs that directly relate to the implementation of inclusive practices.
Thus, the key objectives of this research are twofold: (1) To design and test a new collective efficacy scale for inclusive education that has international validity (2) To undertake a cross-country comparison to explore differences in attitudes, teaching efficacy, collective efficacy and intention to teach in inclusive classrooms.

Measures
Data were collected using a five-part online survey.

Part one
Part One consisted of the eight-item Attitudes to Inclusion Scale (AIS, Sharma and Jacobs 2016) which is rated on a seven-point Likert scale from Strongly Disagree (1) to Strongly Agree (7) and includes items such as 'I believe that inclusion benefits all students academically'.The scale consists of a total score which involves the sum of all items and two subscales: Beliefs and Feelings.The AIS has been demonstrated in previous studies as having strong internal consistency with a Cronbach's alpha of 0.80 (Sharma and Jacobs 2016).For the current study, the Cronbach's alpha for the AIS total score for the whole sample was 0.93, 0.93 for teachers from Canada, 0.91 for teachers from Greece, 0.85 for teachers from Italy, and 0.90 for teachers from Switzerland.

Part two
Participants' intention to teach in inclusive classrooms was measured using the Intention to Teach in Inclusive Classroom Scale (ITICS, Sharma and Jacobs 2016).The ITICS consists of seven items (e.g.'Change the curriculum to meet the learning needs of a student with learning difficulty enrolled in your class') rated on a seven-point Likert scale from Extremely Unlikely (1) to Extremely Likely (7).The scale includes a total score in addition to two subscales: Change Curriculum and Consulting.The ITICS has strong internal consistency as demonstrated in previous studies with an alpha coefficient ranging from 0.73 to 0.83 (Sharma and Jacobs 2016).The alpha co-efficient for the current study was 0.88 for the ITICS total score for the whole sample, 0.86 for teachers from Canada, 0.93 for teachers from Greece, 0.85 for teachers from Italy, and 0.81 for teachers from Switzerland.

Part three
Part Three consisted of the Teacher Efficacy for Inclusive Practices (TEIP) Scale (Sharma, Loreman, and Forlin 2012).The TEIP consists of 18-items in which participants rate their responses on a six-point Likert scale ranging from Strongly Disagree (1) to Strongly Agree ( 6).An example item from the TEIP is 'I am confident in designing learning tasks so that the individual needs of students with disabilities are accommodated'.The scale consists of a total score and three subscales: Efficacy in Inclusive Instruction, Efficacy in Managing Behaviour, and Efficacy in Collaboration.The scale has been used across a number of countries with demonstrated strong reliability (e.g.Savolainen et al. 2012;Schwab, Sharma, and Hoffmann 2022).The alpha coefficient for the current study was 0.93 for the TEIP total score for the whole sample, 0.94 for teachers from Canada, 0.92 for teachers from Greece, 0.93 for teachers from Italy, and 0.88 for teachers from Switzerland.

Part four
Part Four of the survey consisted of the newly developed Teacher Efficacy for Inclusive Practice -Collective (TEIP-C) Scale which was designed to elicit teacher views regarding the capacity of their peers to influence the success of routine classroom activities in creating an inclusive classroom environment.The scale was adapted from items in the TEIP (Sharma, Loreman, and Forlin 2012) developed on the basis of a review of the literature regarding inclusive teaching practices and using a Delphi approach to validate items.The TEIP-C incorporates 18 items rated on a 6-point Likert scale ranging from Strongly Disagree (1) to Strongly Agree (6).Teachers are asked to consider the general or average capacities of their colleagues as a group, rather than for themselves as an individual, in addressing a range of teaching skills.An example item is 'Teachers in my cohort/school can collaborate with other professionals (e.g.itinerant teachers or speech pathologists) in designing educational plans for students with disabilities'.TEIP-C items are summarised in Table 3.

Part five
This part of the survey collected participants' brief demographic information (e.g.age, gender, school setting, teacher level).

Procedure
Ethical approval for the conduct of this project was granted from the relevant ethics board in each participating country.To ensure the social validity and acceptability of items, the survey was pilot tested with a small group of educational professionals prior to data collection.The online survey was distributed in each country using a range of strategies, including social media posts (i.e.Facebook, Instagram, LinkedIn and Twitter), through direct invitation to schools in the researchers' networks and via invitation sent to a randomly selected sample of schools.The survey included an explanatory statement outlining the purpose of the study and what was involved for participants and informed consent was obtained from participants by selecting a button prior to undertaking the survey.Data was collected over a time period ranging from 2019 to 2022.

Data analysis
To investigate the underlying factor structure of the newly developed TEIP-C, its 18 items were subjected to principal components analysis (PCA) using SPSS Version 27.PCA was selected as a parsimonious approach to reducing TEIP-C items into a smaller set of linear combinations, drawing on all of the variance in the original variables.PCA has been identified as a preferred approach to Factor Analysis as it can avoid issues associated with factor indeterminacy (Stevens 2012).Parallel analysis was used to guide comparison of model fit indices, with oblimin rotation used to support the interpretation of identified factors which were assessed for both statistical and conceptual fit.Reliability of the TEIP-C was calculated using Cronbach's alpha to provide a measure of internal consistency of the scale.
To determine whether there were any significant differences across countries for each measure (AIS, ITICS, TEIP and TEIP-C), a series of one-way between groups analysis of variance (ANOVA) were conducted.Prior to conducting the analyses, preliminary tests were undertaken to inspect the normality of distributions of the individual items and total scores for each measure, which revealed relatively normal distributions.

Psychometric properties of the TEIP-C
The 18 items of the TEIP-C were subjected to principal components analysis (PCA) using SPSS Version 27.Prior to conducting the PCA, the suitability of the data was assessed, which involved consideration of two key areas: the strength of the relationship among items, and sample size.With respect to the strength of the relationship among items, visual inspection of the correlation matrix revealed the presence of many coefficients of 0.3 and above.The Kaiser-Meyer-Oklin value was 0.96, exceeding the recommended value of 0.6 (Kaiser 1974), and Bartlett's Test of Sphericity reached statistical significance supporting the factorability of the correlation matrix (Bartlett, 1954).With respect to sample size, inspection of the loading of items indicated several high loading marker variables (above 0.8) (Stevens 2012;Tabachnick, Fidell, and Ullman 2007) and the ratio of participants to items was 84.61:1, well exceeding the 10:1 ratio recommended by Nunnally (1978).
PCA revealed the presence of two components with eigenvalues exceeding 1, accounting for 55.51% and 5.56% of the variance respectively.Inspection of the screeplot revealed a clear break after the second component.Based on Cattell's (1966) scree test, and results of Parallel Analysis which revealed only two components with eigenvalues exceeding the corresponding criterion values for a randomly generated data matrix of the same size (18 variables x 1523 respondents), two components were retained for further investigation.
To support the interpretation of these two components, oblimin rotation was conducted.As summarised in Table 2, the rotated solution revealed the presence of a simple structure (Thurstone 1947), with both components revealing a number of strong loadings and all variables loading substantially on only one component.On the basis of these findings, no items were removed from the scale.Inspection of the items included in each component suggested the following two factors: Component 1 -Engagement; and Component 2 -Inclusive Pedagogies, named on the basis of the core concepts represented in each component.The correlation between the two identified components was .83.
Scores on the TEIP-C range from 18 to 108, with higher scores indicative of higher ratings of collective efficacy.Teacher ratings on the scale overall suggested that most participants rated the collective efficacy for inclusive practices at their school in the mid-range (M = 77.31;SD = 14.71).Teacher ratings of collective efficacy in engaging and supporting student behaviour were also in the mid-range as measured by the Engagement subscale (M = 47.26;SD = 8.85), given responses on this subscale can range between 11-66.Teacher ratings of collective efficacy in using inclusive pedagogies at their school were also in the mid-range as measured by the Inclusive Pedagogies subscale (M = 30.05;SD = 6.51), given this subscale can have a range of responses between 7-42.
The internal consistency of the scale overall and the two identified subscales for the total sample and for each country was investigated by calculating Cronbach alpha coefficients, as reported in Table 2.

Differences in attitudes, efficacy, collective efficacy and intention to teach in inclusive classroom across countries
To determine whether there were any significant differences in teacher ratings of attitudes, efficacy, collective efficacy and intention to teach in inclusive classrooms across countries, a series of one-way between groups analysis of variance (ANOVA) were conducted.To control for Type 1 errors across multiple tests, Bonferroni's adjustment was applied, resulting in a new alpha cut off value of 0.01.

Teacher Attitudes towards Inclusive Education (AIS)
A significant difference between countries was identified for teacher attitudes towards inclusion as measured by the AIS, F (3, 1517) = 285.52,p = <.001.The effect size calculated using eta squared was 0.51 representing a very large effect (Cohen 1988).Post-hoc comparisons using the Tukey HSD test indicated that the mean teacher attitudes towards inclusive education for participants from Italy (M = 52.42;SD = 5.47) were significantly higher compared to participants from Canada (M = 46.23;SD = 9.91), Greece (M = 43.29;SD = 7.86) and Switzerland (M = 35.81;SD = 10.92).Teacher attitudes for participants from Canada were also significantly higher than mean teacher attitudes for participants from Greece and Switzerland.Teacher attitudes for participants from Greece were also significantly higher than attitudes for participants from Switzerland.

Teacher Intention to Teach in Inclusive Classrooms (ITICS)
A significant difference between countries was identified for teacher intentions to teach in inclusive classrooms as measured by the ITICS, F (3, 1518) = 111.64,p = <.001.The effect size was 0.18 representing a large effect (Cohen 1988).Post-hoc comparisons using the Tukey HSD test indicated that intentions to teach in inclusive classrooms for participants from Italy (M = 44.02;SD = 5.47) and Canada (M = 44.70;SD = 4.71) were significantly higher than ratings by participants from Greece (M = 41.23;SD = 6.90) and Switzerland (M = 36.30;SD = 6.48).Intentions to teach in inclusive classrooms for participants from Greece were also significantly higher than ratings by participants from Switzerland.

Teacher Efficacy for Inclusive Practices (TEIP)
A significant difference between countries was identified for teacher ratings of their sense of efficacy in using inclusive practices as measured by the TEIP, F (3, 1519) = 37.37, p = <.001.The effect size was 0.07 representing a medium effect (Cohen 1988).Post-hoc comparisons using the Tukey HSD test indicated that teacher ratings of their sense of efficacy in using inclusive practices for participants from Canada (M = 89.43;SD = 12.15) and Greece (M = 87.73;SD = 10.03) were significantly higher than ratings for participants from Italy (M = 81.97;SD = 11.77) and Switzerland (M = 82.92;SD = 10.39).

Teacher Efficacy for Inclusive Practice -Collective (TEIP-C)
A significant difference between countries was identified for teacher ratings of collective efficacy in using inclusive practices in their school as measured by the TEIP-C, F (3, 1519) = 39.51,p = <.001.The effect size was 0.08 representing a medium effect (Cohen 1988).Post-hoc comparisons using the Tukey HSD test indicated that teacher ratings of collective efficacy in using inclusive practices for participants from Canada (M = 81.37;SD = 15.44),Greece (M = 83.54;SD = 12.93) and Switzerland (M = 80.24; SD = 10.32) were significantly higher than ratings by participants from Italy (M = 73.92;SD = 14.71).

Discussion
This is a complex study that examines teachers and inclusive education from a variety of related angles, including teacher attitudes towards inclusive education and their teaching intentions.However, holding certain attitudes and intending to teach in a particular way does not imply that teachers necessarily feel that they have the ability to actually do so.The TEIP scale is intended to measure an individual's sense of efficacy in this area, while the newly developed TEIP-C goes one step further and measures teachers' views of the efficacy of their teaching colleagues.In other words, how effective is the community when it comes to inclusive practice.Additionally, this study sought to identify any differences in these areas between countries.
Results of PCA suggested a two-factor structure for the TEIP-C: Engagement, and Inclusive Pedagogies, with the scale overall and each of the two subscales possessing strong internal consistency.In addition to providing researchers with a consistent and internationally validated measure of collective efficacy for inclusion, the TEIP-C may also provide a useful measure for school leadership teams to identify opportunities to build a school's collective commitment and confidence in using inclusive practices.Research suggests that providing teachers with adequate training and support, in addition to opportunities to apply new skills, can positively influence teachers' self-efficacy (Forlin and Chambers 2011;Sharma, Loreman, and Forlin 2012).It would be valuable for future research to explore the impact of other modifiable factors such as effective leadership, inclusive school cultures and systemic support and structures (Saloviita and Schaffus 2016) on teacher attitudes, selfefficacy, collective efficacy and intention to teach in inclusive classrooms.
In terms of teacher attitudes and intentions to teach in inclusive classrooms, results suggest that countries could be ranked in order from most to least positive as follows: Italy, Canada, Greece, and Switzerland.This finding is not particularly surprising given the long-term commitment to inclusive education in Italy that now spans decades (De Anna, Gaspari, and Mura 2015;D'alonzo 2022).It is also consistent with the findings of the pilot study (Romano, Rullo, and Petruccioli 2021).Switzerland, on the other hand, still operates a comparatively rigid education system that is streamed by ability and that contains numerous segregated special education options (Hollenweger 2016).
The findings using the TEIP-C scale were of interest.Despite ranking their own attitudes, intentions, and efficacy highly, it appears that Italian teachers rank the efficacy of their peers as being low in comparison to teachers from the other three countries, who were all quite similar in their responses.While a lower comparative rating does not imply that Italian teachers think their colleagues have low efficacy for inclusive teaching (they were still positive), it is a curious result.It would be interesting to investigate further to see if this is a cultural phenomenon unique to Italy or if some other factor is at play.Thankfully for those who support inclusive practice in Italy, this result had no bearing on the attitudes, feelings of efficacy, and inclusive teaching intentions of the individual teachers.

Limitations
The findings of this study need to be considered within the context of several limitations.The convenience sampling approach to recruitment may have resulted in a biased sample, such that teachers with greater levels of self-efficacy and more positive attitudes towards inclusion may have been more likely to participate in the study.Similarly, it was not possible to report on the representativeness of the sample, due to variations in the demographic data collected from each country.As participant self-report was used to measure the constructs of interest, social desirability may have influenced teachers' responses.COVID-19 may also have impacted teachers' responses, considering the significant changes to teaching and learning experienced by many teachers around the world throughout the pandemic (Dabrowski 2021).Given the loss of face-to-face contact and opportunities for routine interactions with colleagues during periods of remote learning (Hargreaves and Blais 2021), it would be valuable to explore any variations in teacher ratings of collective efficacy post-pandemic.The two TEIP-C components identified in the current study were highly correlated.Future research may further examine the psychometric properties of the scale including further testing of the validity of the proposed factor structure through Confirmatory Factor Analysis (CFA).Also, given collective efficacy can be considered a school-level concept, future research should extend application of the TEIP-C from the individual-level to the between school-level.Finally, it will be important for future research to examine the relationship between collective efficacy, teacher attitudes and teacher efficacy and their impact on intention to teach in inclusive classrooms.

Conclusion
This study sought to explore a further dimension of inclusive education, collective efficacy, through the testing of a new tool, the TEIP-C.In addition, the study aimed to investigate whether teacher attitudes, teacher self-efficacy, collective efficacy and intention to teach in inclusive classrooms differ across countries.Results suggested the TEIP-C provides a valid tool for the measurement of collective efficacy.Several differences across countries were observed, including more positive attitudes and a greater intention to teach in inclusive classrooms in teachers from Italy as compared to teachers from Switzerland, with the responses from teachers from Canada and Greece falling somewhere in the middle of this spectrum.Such differences appear to reflect the longer-term commitment of countries such as Italy towards inclusive education and underscore the importance of system-level influences in supporting inclusive teacher practice.It would be of value to further explore the role of collective efficacy and its relationship with teacher attitudes and self-efficacy on intention to teach in inclusive classrooms in future research.Such research may be useful in identifying further opportunities to build teachers' collective confidence and commitment towards inclusive practices in order to increase the inclusion of all learners.

Table 1 .
Demographic characteristics of participants.
a Response option in the Canadian survey.b NA = No response c Response option in the Swiss survey.de Response options in the Italian survey.

Table 2 .
Internal consistency of the TEIP-C scale and subscales.

Table 3 .
Pattern and structure matrix for PCA with oblimin rotation of two factor solution of the TEIP-C.