Assessing the development of global competence in teacher education programmes: internal consistency and reliability of a set of rubrics

Global competence is a complex concept as it is multifaceted, composite, multi-layered, multidimensional


Introduction
Teachers at all school levels currently face educational issues arising from global challenges: intercultural matters, economic questions, and political disputes.Creating and managing a learning environment, even in a small classroom located in a remote village, depends increasingly on the teacher's global competence (GC).This reality means that teachers are working in school contexts where both local needs and global challenges are present, and teachers must be prepared to facilitate the 'development of young people to become informed, engaged, and globally competent citizens' (Kopish, 2016, p. 76).In their futures, pre-service teachers will have to work in complex educational systems composed of many cultures, ideas and perspectives.GC represents the concept/paradigm that can help these future teachers to build broad and inclusive educational paths where all pupils will have the opportunity to discuss and experience several forms of learning (OECD PISA, 2018;van Werven, Coelen, Jansen, & Hofman, 2021).
This article presents the third step of a study focused on developing and assessing GC within teacher education programmes.The first phase was dedicated to identifying which aspects of GC should be integrated into initial teacher education programmes across Europe, as well as how to do so (Parmigiani, Jones, Kunnari, & Nicchia, 2022).The second stage was aimed at designing, creating, and validating a set of rubrics to be used by either teacher educators or pre-service teachers to assess pre-service teachers' development of GC (Parmigiani et al., 2022).Through a Delphi method and the involvement of 31 international experts, we created a specific set of rubrics (Appendix A) to be used in teacher education programmes to help both teacher educators and pre-service teachers assess and monitor individual levels of GC development.
There were three main objectives in the third step.The first was to analyse the structure and features of the rubrics in order to identify positive aspects and inadequacies.The second aim was to test the internal consistency and reliability of the rubrics in order to validate them within an international context that would be considered robust from a methodological point of view.Finally, the third goal was to offer a consistent instrument for the assessment of pre-service teachers' development of GC within teacher education programmes.To do so, the set of rubrics was administered to pre-service teachers from several countries from all around the world.

Global competence as a theoretical framework
There are a number of GC frameworks, with many of these sharing similarities with cultural competence, global citizenship and a host of related terms (Parmigiani, Jones, Kunnari, & Nicchia, 2022).In particular, most relevant frameworks include components related to cultural awareness, knowledge, skills and values (Kahn & Agnew, 2016;Reimers et al., 2010;Schleicher, 2018).The research leading to the rubrics presented in this paper has been guided by the Asia Society/OECD (2018) definition of GC as a theoretical framework, as it refers not only to understandings, but also a capacity 'to act for collective well-being and sustainable development' (p.7), which is essential for teachers in the 21 st century.

Fostering pre-service teachers' global competence
Three main approaches to fostering GC in post-secondary students have been broadly supported (Shultz, 2007).The first approach -a neo-liberal approach -focuses on students acquiring the skills necessary for global economic participation and mobility.The second approach focuses on student activism and critique of existing global systems.The third and most common approach within post-secondary institutions is the transformational approach (Aboagye & Dlamini, 2021).Transformational learning allows students to develop 'attitudes and values that shift their meaning perspective [. . .] towards humanity, critical thinking, cross-cultural understanding, and orientation toward social justice' (Stanlick, 2021, p. 43) through active processes based on disorienting events and reflection (Mezirow, 1991).Indeed, Asia Society/OECD (2018) definition of GC captures all three post-secondary approaches to GC: Global Competence is a multi-dimensional construct that requires a combination of knowledge, skills, attitudes and values successfully applied to global issues or intercultural situations.Global issues refer to those that affect all people, and have deep implications for current and future generations.Intercultural situations refer to face-to-face, virtual or mediated encounters with people who are perceived to be from a different cultural background (p.4).
Given the importance of experience and reflection as central to the development of students' GC (Stanlick, 2021), teaching practices related to its development across disciplines within post-secondary education have included service learning, signature projects, and international projects (for examples, see Aboagye, 2021;Broom & Bai, 2021;Gisolo & Stanlick, 2021;Naidoo & Benjamin, 2021).Moreover, research supports curricular inclusion of GC pedagogy and content across courses in a habitual, integrated, and active learning format rather than a stand-alone course structure (Mostafa, 2020).Of particular note are the additional needs of post-secondary students who intend to become teachers, as these students must not only develop their own GC but must also develop the practical skills to foster it within their future students (Kerkhoff & Cloud, 2020).Recent research has confirmed that the same active learning processes that foster GC in post-secondary students are also useful with school children; these strategies include debates, discussions, games, project-based learning, and service learning (Asia Society/OECD, 2018).

Assessing the development of global competence in teacher education programmes: State of the art
Assessing GC is not an easy task, especially as there is no common understanding among the scientific community of what GC is exactly (Sälzer & Roczen, 2018).Thus, 'scientific theory-building is in this regard relatively young and undeveloped' (Sälzer & Roczen, 2018, p. 7).Often no clear distinction exists between connected concepts such as global competence, global citizenship education, or intercultural competence to name just a few.Intercultural competence focuses mainly on an individual's communication and behaviour in intercultural contexts (Deardorff, 2006) whereas global citizenship education 'refers to a sense of belonging to a broader community and common humanity.It emphasises political, economic, social and cultural interdependence and interconnectedness between the local, the national and the global' (UNESCO, 2015, p. 14).In contrast, GC refers to competencies connected to global issues and intercultural situations (Asia Society/OECD, 2018).Another difference between GC and global citizenship education is the latter's focus on conceptions and frameworks of citizenship education and hence, its roots in the context of education policies and curricula, as well as in teaching and learning (UNESCO, 2015).
These different views on what constitutes GC have also led to disputes on how to measure GC.The PISA study presents one prominent example, as more than 30 countries refrained from measuring GC in (OECD PISA, 2018;Sälzer & Roczen, 2018).Despite these problems with developing assessment instruments for GC, there are a range of tools that are used in practice.In their overview of existing assessment tools, Conolly, Lethomäki, and Scheunpflug (2019) concluded that most existing measurement tools focus on knowledge and skills on an individual level.In contrast, the rubrics tested in this study also include a scale about action.There are only a few GC assessment tools adapted for teachers.One of them is the Global Readiness Scale (Kerkhoff, 2017), which is a measurement model and scale of teacher practices related to global readiness instruction.This scale can be used in teaching situations and was developed in the US context.Another example is the Globally Competent Teaching Continuum (Tichnor-Wagner, Parkhouse, Glazier, & Cain, 2019), an interactive tool to help teachers develop GC and connected teaching practices.Overall, this tool is a self-reflection tool rather than an assessment tool and, like the Global Readiness Scale, was developed for the US context.Both tools are highly connected to classroom practices of in-service teachers and thus, cannot be easily adopted to initial teacher education.Moreover, they do not account for other national contexts.The set of rubrics tested in this study is hence a rare example of an assessment tool for GC that can be used within initial teacher education and has been validated in an international context.

Global competence in the different contexts of the study
Across the various countries surveyed for the purpose of the current study, there exists a wide range of approaches to GC in teacher education and school education.In some nations there is a clear focus on GC within teacher education and accreditation approaches, while in others GC is more evident in terms of student outcomes and school curricula.Some countries have a combined approach, where GC is a focus in both teacher accreditation and school curriculum.Additionally, aspects of GC are explicitly outlined in some countries, while in others, principles of GC can be inferred through close inspection of relevant documents.
The following countries in the current study have formal Teacher Standards: Australia (Australian Institute for Teaching and School Leadership, 2022); U.S.A (Council of Chief State School Officers, 2013); Canada, by province, for example New Brunswick (Government of New Brunswick, n.d..); Slovakia (University portal of the Ministry of Education, Science, Research and Sport of the Slovak Republic 2019); Germany (KMK Kultusministerkonferenz, 2019); and France (Ministère de l'éducation nationale et de la jeunesse, 2013).Typically, Teacher Standards outline the knowledge and skills which teachers are expected to gain through teacher education programmes and demonstrate prior to being accredited/qualified (Santoro & Kennedy, 2016).Within these various standards, it is possible to identify aspects related to GC either specifically or tangentially.For example, one Australian Standard includes that teachers must 'Demonstrate knowledge of teaching strategies that are responsive to the learning strengths and needs of students from diverse linguistic, cultural, religious and socioeconomic backgrounds' (Australian Institute for Teaching and School Leadership, 2022, p. 10).In the U.S.A Standards, reference is made to teacher candidates developing their students' cultural competence, for example: "The teacher promotes an understanding of inter-and intragroup diversity to facilitate learners' development of cultural competence and build respect across communities" (Council of Chief State School Officers, 2013, p. 19).While Norway's Teacher Standards make no specific reference to GC, these are addressed in the Norwegian Qualifications Framework (Ministry of Education and Research, 2014).
In Portugal, there are no specific teacher standards, but similar competencies are identified in Decree-Law no.240/2001 outlining a teacher profile which includes an expectation to identify and respect cultural differences among students and value various knowledge and cultures (Ministry of Education Portugal, 2001).GC can also be identified in documentation related to what school students should be learning (Ministry of Education Portugal, 2017).In Italy, despite not having specific teacher standards, representatives of education ministries from many regions have campaigned for the national Ministry of Education to include global citizenship in teacher education programmes (AICS -Agenzia italiana per la cooperazione allo sviluppo 2018).In Israel, a document outlining teacher education requirements (The Council of Higher Education, 2020) infers the inclusion of GC through aspects such as multiculturalism, anti-racism and civil education.

Aims and research question
This study was aimed at testing the internal consistency and reliability of a set of rubrics designed to assess pre-service teachers' development of GC.The rubrics are available in the appendix (Appendix A).The study sought to answer the following research question: what are the levels of internal consistency and reliability of a set of rubrics about preservice teachers' development of GC?As mentioned previously, the purpose of the study was to deeply analyse the characteristics of the rubrics in order to identify strengths and weaknesses and offer a reliable and coherent instrument to be used within teacher education programmes.

Participants and data collection procedure
To pursue the aims of the study and answer the research question, we involved 12 institutions located in 10 countries around the world.These institutions offer at least one of the following teacher education programmes for pre-service teachers at different levels: pre-primary/kindergarten, primary, lower/upper secondary, vocational or special education.Table 1 shows the number of participants from each institution.
Table 2 reports the demographic and the school/educational variables: participants' birth country, gender, birth year categories, future intentions of grade/school levels and English level.
Tables 1 and 2 indicate some differences between the number of participants involved by each institution engaged in the study.The differences are due mainly to two factors.The first element is the total number of pre-service teachers recruited in each institution.The second element is the number of pre-service teachers able to complete an assessment tool written in English.
The set of rubrics was administered online by the representative of each teacher education programme involved in the study.Thus, the set of rubrics was completed by the respondents without the intervention of the research team collecting the data.The research procedure was strictly governed by ethical codes of conduct.The procedure was approved by the ethical committee of the principal investigator's university and permissions to conduct the research were obtained from each institution involved in the project.The set of rubrics were completed anonymously, and the preamble stated that the researchers would not have access to the identities of the people who completed the tool.
Respondents were free to participate in the study or not.They were also given the option to cease involvement in the study at any time.No risks were associated with completing the anonymized survey outside of those of daily life.No incentives were offered, but respondents who completed the instrument may have benefitted from knowing that they contributed to the understanding of global competence in pre-service teachers.
The rubrics were written and administered in English, because we wanted to create an instrument available potentially all over the world.For this reason, all pre-service teachers who filled in the rubrics were requested to declare their level of English language skills in order to check any significant difference linked to this factor.

Instrument and quantitative data analysis procedure
As mentioned previously, a specific set of rubrics was used to assess pre-service teachers' development of GC.The rubrics were designed and created in two previous research steps (Parmigiani et al., 2022;Parmigiani, Jones, Kunnari, & Nicchia, 2022).The rubrics are composed of three areas: • Exploring: 4 dimensions and 6 indicators/criteria • Engaging: 5 dimensions and 7 indicators/criteria • Acting: 7 dimensions and 19 indicators/criteria.The three areas combined contain a total of 32 indicators.The first area, Exploring, is aimed at assessing initial approaches to GC issues.The dimensions, such as 'openness' or 'intent to experience/interact' indicate the first steps in facing global issues.Sample indicators/criteria of this area include: 'I feel responsibility to address ethical, social, economic and environmental challenges' (dimension: global responsibility) and 'I support rights, equity and social justice in different sectors such as gender, racial, religion, disability, etc.' (dimension: ethical orientation).
The second area, Engaging, focuses more precisely on pre-service teacher's reflections on their development and their attitudes regarding 'global self-awareness' or 'inclusion and diversity.'It is important to emphasise that the first area, Exploring, can be used with all higher education students enrolled in a variety of disciplines, as the dimensions are general.The second area also focuses on overall dimensions of GC, but the criteria become more specific to higher education students who are undertaking a teacher education programme.Sample indicators/criteria for this area include: 'I demonstrate awareness of diverse and multiple perspectives when teaching/practising in classrooms with students from diverse backgrounds' (dimension: world views, perspectives and cultural diversity) and 'I try to contribute to the development of a more just, peaceful, and sustainable world' (dimension: sustainability).
The third area, Acting, is specific to pre-service teachers and all dimensions and indicators/criteria focus on school and educational contexts and situations (e.g.'managing complex learning environments' or 'active teaching strategies').Sample indicators/ criteria of this area include: 'I'm able to create effective learning environments and manage classes with students from diverse backgrounds' (dimension: managing complex learning environments) and 'I'm able to design and implement formative assessment methods to inform instruction with students from diverse backgrounds (self and peer assessment, portfolios, etc.)' (dimension: interactive assessment strategies).
The indicators/criteria are assessed using four levels: emerging; developing; achieving; extending.For each level, a descriptive caption is provided.The captions are different for each area.For instance, the caption for the level Emerging in the Exploring area is: 'I show a low willingness to explore the criterion'.However, the caption for the same level in the Acting area is: 'I show a low willingness to act in relation to the criterion'.There is also the option to not indicate any level by selecting 'Not applicable' in cases where the pre-service teacher is not involved in that criterion or does not want to respond.The rubrics also include text boxes for additional qualitative comments.
The set of rubrics has been designed in two versions: 'self-administered' and 'administered by the teacher educators'.The first version is available in the appendix, and can be used by each pre-service teacher.The second version can be used by the teacher educators observing the pre-service teachers while they are acting in an educational context/situation, such as: teaching practice, internship, workshop, etc.
From a technical point of view, a rubric is not a questionnaire but there are some similarities.First of all, the areas can be considered as subscales.Secondly, the indicators/ criteria can be considered as items and, ultimately, the levels can be considered as Likert scales.But, in this case, it is necessary to consider them as ordinal scales and to conduct the data analyses with non-parametric tests.
The quantitative analysis concentrated on the instrument's reliability, so we used the following coefficients: Cronbach Alpha (α); McDonald's Omega (ω); average inter-item correlation.To explore the instrument's internal consistency more deeply, we performed both Exploratory (EFA) and Confirmatory Factor Analysis (CFA).Ultimately, to investigate potential significant differences between the groups of participants (based on the demographic and school/educational variables) involved in the study, we conducted nonparametric tests for independent samples: Kruskal-Wallis (used when the groups are more than two) and Mann-Whitney U. The quantitative analyses were conducted using SPSS 26 and JASP 0.16.2.
Because the rubrics were written in English, the participants were asked to indicate their English level as shown in Table 2. Consequently, the instrument's reliability was first calculated underlining the results of the participants' groups labelled as 'Beginner' (1), 'Intermediate' (2), 'Advanced' (3) and 'Native' (4).Table 3 shows low values of reliability in the Areas A and B for the pre-service teachers who indicated a 'Beginner' English level.By contrast, the coefficient values calculated for the participants who indicated 'Intermediate', 'Advanced,' or 'Native' as their English level were good or excellent, as indicated in the critical values.Therefore, to avoid bias, we decided not to consider the 26 rubrics filled in by pre-service teachers with a low English level.These rubrics originated as follows: 16 from Italy, 7 from Germany, 2 from Portugal and 1 from Israel.
Table 4 shows the overall coefficients calculated to investigate the reliability of the instrument after deleting the rubrics of pre-service teachers with an English level indicated as 'Beginner'.
The values of both coefficients α and ω indicate a good level of reliability for areas Exploring and Engaging and an excellent level of reliability for the third area: Acting.Table 5 confirms that the reliability is stable for all areas when a single item is dropped.
The average inter-item correlation also shows good levels of reliability.According to Clark and Watson (1995) and Zmnako and Chalabi (2019), the average inter-item correlation should be between .15 and .50.Spiliotopoulou (2009) proposes clarifications and specifications when considering broad versus narrow constructs.When the instrument 'measures broad characteristics (i.e.general constructs such as extraversion)' (p.152) the critical values should be within the range of .15 to .20 and when the instrument is focused on narrow characteristics and specific constructs, the critical values can be considered between .40 and .50.This second case is suitable for the constructs related with GC.Ultimately, DeVon et al. (2007) indicated studies where the average inter-item correlation was considered appropriate with scores between .30and .70,but the values generated by the GC set of rubrics met the more stringent standards proposed by Spiliotopoulou (2009).

Internal consistency: exploratory and confirmatory factor analysis
To test the internal consistency, we performed both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).The aim of these two analyses .400 to .500(Spiliotopoulou, 2009) .300 to .700(DeVon et al., 2007)  was to investigate the factorial structure of the rubrics to determine whether the areas included into the rubrics can represent the concept of GC effectively.From a methodological point of view, since we had a relatively large sample size of 703 participants, we randomly divided 'the group into half so that both an EFA and a CFA could be undertaken' (Willmer, Westerberg Jacobson, & Lindberg, 2019, p. 6).

EFA
Exploratory factor analysis (EFA) aims to identify 'the common factors that explain the order and structure among measured variables' (Watkins, 2018, p. 220).Firstly, we performed an EFA with promax rotation since the correlations among the factors are more than .15(Devellis, 2017;Finch, 2006).We used maximum likelihood extraction, because the data were normally distributed, and eigenvalues > 1 to highlight the number of significant factors related to GC.Then, we completed a further EFA with promax rotation again using maximum likelihood extraction but imposing a three-factor solution, because we needed to examine the original structure of the set of rubrics which was composed of three areas.Regarding the EFA without a forced solution, the result of the Kaiser-Meyer-Olkin test was .939,proving that the sample was adequate.In addition, Bartlett's Test of Sphericity showed a p-value of < .000df = 496).The goodness of fit test for this model was 641.581 (df 319; p < .000).
The analysis revealed five factors that explained 64.378% of the total variance.In particular, each factor explained the following percentages of total variance: F1 42.94%; F2 8.74%; F3 5.13%; F4 4.05%; F5 3.52%.Table 6 shows the factor loadings for each indicator/criterion.As shown in Table 6, Area C of the rubrics is separated into two factors.Factor 1 includes the first four indicators/criteria (from C1a to C4e) of Area C -Acting.In addition, Factor 2 is composed of the remaining indicators/criteria from the same area, being C5a to C7a.The indicators/criteria of Area A -Emerging are included into the Factor 3 (except A4a).Factors 4 and 5 include the indicators of Area B -Engaging.Figure 1 effectively represents the relationships among the factors in three dimensions.
For the EFA with a forced three-factor solution, the result of the Kaiser-Meyer-Olkin test was .918,proving that the sample was adequate.In addition, Bartlett's Test of Sphericity showed a p-value of < .001(Chi-square 4369.524;df = 496).The goodness of fit test for this model was 1058.163(df 403; p < .000).The extracted three factors explained 56.81% of the variance, specifically Factor 1 explained 42.99% of the variance, Factor 2 explained 8.74%, and Factor 3 explained 5.13%.In this case (see Table 7), Factor 1 includes all indicators/criteria of Area C plus two indicators/criteria from Area B (B2b   Jackson, Gillaspy, and Purc-Stephenson (2009) affirmed that 'confirmatory factor analysis (CFA) is a powerful statistical tool for examining the nature of and relations among latent constructs' (p. 6).They also note that 'CFA is often the analytic tool of choice for developing and refining measurement instruments and assessing construct validity [. ..] and plays an essential role in measurement model validation in path or structural analyses' (p. 6).

CFA
On one hand, the structure of the rubrics is based on three areas, suggesting a model with three factors.On the other hand, the EFA without a forced number of factors offered a five-factor solution with some peculiarities mainly for Area B and in part for Area A. For these reasons, we ran three CFAs following three models.The first one was based on the three original areas provided by the set of rubrics: Factor 1 with all indicators/criteria of Area C; Factor 2 with all indicators/criteria of Area A; Factor 3 with all indicators/ criteria Area B. The second model was based on five factors as indicated by the EFA without forced solution (as shown in Table 6) whilst the last model was based on the disposition of the indicators/criteria within the three factors highlighted by the EFA with three-factor forced solution (as shown in Table 7).For all models, we used standardised coefficients and maximum likelihood as estimation models.Lastly, we excluded the observations with missing values.
To investigate the models' goodness of fit, we calculated a series of statistics: overall chi-square (Hooper, Coughlan, & Mullen, 2008); root mean square error of approximation (RMSEA) (Hooper, Coughlan, & Mullen, 2008;Steiger, 1990), comparative fit index (CFI) (Bentler, 1990), Tucker-Lewis index (TLI) (Bentler, 1990;Tucker & Lewis, 1973), and the standardised root mean square residual (SRMR) (Bentler, 1995;Hooper, Coughlan, & Mullen, 2008).Tables 8 and 9 show, respectively, the summary of the CFA results for the three models and the CFA results and decision matrix.As indicated in Table 9, the model with five factors arising from the non-forced EFA can be accepted.Also, the factor loadings' analysis confirms this result since all loadings for those areas are acceptable, ranging from .573 to .720.

Differences among groups
As mentioned before, we considered the four levels to assess the indicators/criteria as ordinal scales, so we performed non-parametric tests to find potential significant differences between the groups of pre-service teachers involved in the study.In particular, we used the U Mann-Whitney test for dichotomous variables and the Kruskal-Wallis test for polytomous variables.Specifically, in the Kruskal-Wallis test, we performed the pairwise comparisons to assess the differences.We considered the following variables: gender; birth year categories; country where the pre-service teacher attended the teacher education programme; future intention of grade/school levels.We considered gender as a dichotomous variable since the number of pre-service teachers who chose the options 'other' (5) and 'I don't wish to say' (3) were too small for an effective calculation.Table 10 shows the results of the analyses, indicating the significant differences among groups.We indicated the area and the indicator/criterion; the variable and the test used for that variable.Ultimately, we indicated the direction of the result, underlining the groups who achieved higher scores in that indicator/criterion.

Discussion
As mentioned previously, the purpose of this study was to examine an instrument designed to assess pre-service teachers' development of GC, test its internal consistency and reliability validated within the international context, and make it robust and consistent from a methodological point of view.Before going into the technical and methodological details of the discussion, it is important to underline that, within teacher education programmes, teacher educators use many instruments to assess the professional growth of pre-service teachers.For instruments not validated at an international level, through extensive and deep data collection, accuracy and reliability may not be ensured.Consequently, some of these instruments risk not reflecting the content dimensions of the objectives of the instruments themselves (Benson & Clark, 1982).The crucial point of the discussion, broadly clarified in the following paragraphs, is that this set of rubrics represents a validated instrument that measures what is intended to measure.In this case, teacher educators can use this set of rubrics with the awareness of sustaining the professional development of preservice teachers in the issues related to GC in educational contexts.

Methodological issues
The results of the analysis indicate some important reflections and considerations both from a methodological and an educational point of view.In this section, we begin with the technical issues related to the structure of the rubrics.The reliability of the rubrics, measured with Cronbach's alpha (α), McDonald's omega (ω) and the average inter-item correlation, shows high scores in particular for Area C -  (Browne & Cudeck, 1992;Davidson, Tripp, Fabrigar, & Davidson, 2008).**value >.950 good model fit; .949<value<.800acceptable model fit; .799<value<.600 marginal model fit (Hu & Bentler, 1999;Xia & Yang, 2018).***value<.05good model fit; .06<value<.10 acceptable model fit; value > .11marginal model fit (Cangur & Ercan, 2015;Iacobucci, 2010;Kline, 2015).Acting, which represents the professional area to be developed specifically by the preservice teachers.It is important to underline that the rubrics, because they are written in English, must be used only with pre-service teachers who have, at least, an intermediate level of proficiency in English, otherwise the scores' reliability decreases in a significant way.
The internal consistency was calculated with both an EFA and a CFA.In particular, the EFA was performed in two main ways: without and with a forced solution.The first solution was aimed at investigating the potential latent factors included into the rubrics and not considered by the authors.The EFA without a forced solution revealed five factors, as shown in Table 6.Factor 1 explains the highest level of variance (42.94%) and is composed of most indicators/criteria of Area C (from C1a to C4e) except indicators/ criteria from C5a to C7 which are included into Factor 2 (8.74% of total variance).The first group of indicators/criteria can be named 'professional matters' since they refer to the ability of pre-service teachers to self-reflect, interact with colleagues and other school actors (parents, pupils, etc.), manage complex environments and design learning contexts founded on intercultural teaching.In contrast, the second group of indicators/ criteria is composed of very specific items, such as international experiences, teaching and assessment strategies.This factor can be called 'in the classroom' since the criteria concern the actions to be carried out in the classroom.In total, the first two factors explain more than 50% of the total variance, confirming the high reliability of Area C. Area A is completely included into Factor 3 (5.13%),whereas Area B is split into two factors (F4, 4.05% and F5, 3.52%).Factor 4 contains indicators B1a, B1b and B5.These indicators represent the importance of actions to support sustainability, and for this reason, it can be named 'actions for sustainability'.Factor 5 is composed of indicators from B2a to B4a.In this case, Factor 5 can be called 'multiple perspectives and inclusion' since the indicators refer to the capacity of pre-service teachers to consider multiple perspectives and include all pupils in their classrooms.Area B appears the most critical sector of the rubrics.The EFA with forced solution confirms the high reliability of Areas C and A whilst Area B is again split into two factors, overlapping with Area C (B2b and B3a).
The CFA was performed with three models since the indicators/criteria were distributed following: the original areas of the rubrics; the EFA without forced solution; the EFA with forced solution.As shown in Tables 8 and 9, only the second model shows an acceptable fit.
The combination of the results of reliability and internal consistency analysis implies some important considerations.The rubrics show a high reliability and a good internal consistency.We can state that pre-service teachers' development of GC can be assessed through use of the rubrics by teacher educators and/or the pre-service teachers themselves, but attention should be given to some aspects.First of all, it is necessary to consider the results of Area C, focusing and underlining the scores of the indicators/ criteria from C1 to C4, on the one hand, and C5 to C7, on the other.These two factors can be considered as two subscales of the Area C. Area A is quite consistent so it can be confirmed as it is.Area B should split into two subscales composed, respectively, by the indicators B1a, B1b and B5 (actions for sustainability) and B2, B3 and B4 (multiple perspectives and inclusion).
The final consideration is that Area C is specific to pre-service teachers but Areas A and B can be used for all higher education students.When used with post-secondary students not intending to be teachers, the indicator/criterion B3 requires slight modification, by deleting reference to the classroom.
The comparisons among groups show small differences for the variables: 'gender', 'birth year categories' and 'future intention of grade/school levels.'In contrast, the variable 'country where the pre-service teacher is attending the teacher education programme' reveals a high number of significant differences in all areas.This specific analysis does not invalidate and affect the reliability and the internal consistency of the rubrics.In future studies, it will be interesting to investigate the reasons why pre-service teachers in some countries, such as Italy, Norway, France, Germany or Slovakia, had lower scores in certain indicators/criteria.

Educational issues
The rubrics are intended as a means of measuring 'progressive self-assessment' of 'professional growth' (Islam & Stamp, 2020;Parmigiani et al., 2022).Understanding this intended use of the rubrics suggests several pragmatic considerations.First, given that research (Killick, 2020) shows that education for GC occurs best across courses in a progressive and integrated fashion, the rubrics are best used at selected time points throughout the teacher education programme in an ongoing manner.Given the poor reliability of the rubrics when used with students with beginner language skills in English, the rubrics should not be used in their current forms with these students.A possible solution to this issue would be translation of the rubrics into other languages.Second, given that experiential learning is a key factor in the development of GC, the rubric could be used prior to, during, and after a focussed GC activity (e.g.service learning, signature projects, and international projects).Importantly, upon return from these intensive experiential learning activities, growth through additional reflection over time by the pre-service teacher with the crucial support of the teacher educator is also likely, so the rubrics could be used once again after a suitable developmental period (Gisolo & Stanlick, 2021).By using the rubrics in this repeated manner, not only are they useful in providing evidence of GC growth in students, but they are also useful in determining which experiences (e.g.international practicum) or courses foster the greatest GC growth in students collectively.Finally, although primarily intended for pre-service teacher selfassessment, the rubrics may also be used by teacher educators to help guide student focus and development toward GC (Parmigiani et al., 2022).Given that there is sparse evidence about the specific, course-based active pedagogies that best promote GC in teacher candidates, these rubrics provide a new and reliable way to track teacher candidate progress, informing the need to modify and enhance teacher candidate experiences and preparation as indicated by the rubrics, with the goal of maximising courses' impact on GC development.

Conclusion
Based on our findings, we can state that this study, conducted in 10 different countries, revealed that the set of rubrics has a high level of internal consistency and reliability, so it can be used within teacher education programmes as a self-assessment instrument to monitor the development of GC of pre-service teachers.
As stated in the discussion paragraph, Area A revealed high consistency whilst Areas B and C showed some overlaps.As a result, we recommend that Area B and C scales can be administered as presented, but the results should split into two subscales for each area in order to focus better on specific facets of GC development.For these reasons, we suggest that all teacher educators who use this set of rubrics pay attention to these aspects so that they can observe and supervise effectively the development of GC among their preservice teachers.In particular, we recommend using the rubrics in specific contexts such as a pre-and post-test before and after an international placement/internship.In addition, the rubrics can be administered, as a formative assessment method, during an academic course or a workshop focused on intercultural/international issues.The rubrics can also represent a self-reflection exercise after simulations based on real-life/real-world situations.
In addition to the affordances of the current study, we would like to highlight some limitations.First, the set of rubrics was administered only in institutions located in Western countries.Although the sample size of participants is high, it will be necessary, in the future, to administer the rubrics in other areas of the world.The second limitation is represented by the unique version of the rubrics in English.In the future, it will be important to consider the opportunity to translate the rubrics into several languages to allow a larger number of teacher educators and pre-service teachers to use the rubrics in an easily accessible way.
Ultimately, linked to the last limitation, an upcoming study should focus on the differences connected to the variable 'country where the pre-service teacher is attending the teacher education programme.'It would be interesting to investigate levels of global competence and reasons for the differences across countries so that we can support the professional growth of pre-service teachers and enhance their GC awareness and development on an international scale.
students with special needs in the education system and the connection between academia and field experience in teacher training.Aviva was a pedagogical advisor and lecturer on learning disabilities and inclusive education.She has a Ph.D. in special education from the Hebrew University of Jerusalem.Her thesis was about the resilience of children with learning disabilities in inclusive classes.
Kate Ferguson Patrick has 15 years' experience teaching in primary settings in the UK and Australia before becoming an academic at University of Newcastle, Australia in 2001.She is a senior lecturer and primary curriculum and pedagogy specialist with experience teaching in primary Mathematics, HSIE, integrated curriculum and pedagogy including Cooperative Learning.She has published extensively in cooperative learning and global education and is currently working internationally with colleagues on a Global competence in Teacher education project.She leads the Global Education Research and Teaching team at UoN. Alona Forkosh Baruch is associate professor and head of the International Office at Levinsky-Wingate Academic College.She served as head of ICT (Information and Communication Technologies) in Academia and head of the Authority for Empowering Teaching and Learning.She has wide experience in international comparative research and overall research on topics that intersect between innovation and technology.She studies ICT implementation as a lever for paradigm change and innovation in the schooling system and in higher education and teacher education, new literacies in the information era, social media and education, and aspects of curriculum in the information era and special education in the digital era.She is a member of national and international committees of research, development, policy and implementation of technology in education.
Eileen Heddy serves as the Director of the Office of Support for Teacher Education Programmes and Global Student Teaching at The College of New Jersey.Her research interests include international field experiences, action research, civic engagement and culturally responsive pedagogy.She holds a master's degree in history from Rutgers University and is a doctoral candidate at the University of South Carolina.
Maria Antonietta Impedovo is Associate Professor at ADEF Laboratory, Aix-Marseille University, France.She teaches at the School of Education at Aix-Marseille University, France.She held her Ph.D. about Educational technology in Italy, followed by a postdoc in Switzerland.Her main research interests are identity, teacher agency, and teacher professional development in a sociocultural psychology perspective.Since 2010, she has been actively participating in multiple national and international formative and research projects like the capacity building "Blended learning for teacher educators between Asia and Europe".She has written scientific papers in Italian, Spanish, French, and English.
Marcea Ingersoll is an Associate Professor in the School of Education and a member of the Centre for Interdisciplinary Research on Narrative at St. Thomas University.Building on her experiences as an international teacher and teacher-educator, her scholarly work is situated at the curricular crossroads of narrative, teacher education, and international schooling.
Mellita Jones is an Associate Professor of education and the international community engagement coordinator for the National School of Education at Australian Catholic University in Ballarat, Victoria.She has extensive experience in leading Australian pre-service teacher immersion programs in Solomon Islands since 2009and Kiribati (2016-2018)).She has also worked with teachers in Brunei Darussalam and written UNESCO's teacher education program for the Pacific region.Her research is concerned with the purpose of teacher education for securing a socially just and equitable world and is thus linked to global citizenship education concerned with social justice.Mellita has won a number of awards for teaching excellence and community engagement.She currently serves on the Executive committees for the Australian Teacher Education Association (ATEA) and the World Federation of Associations of Teacher Education (WFATE).
Yael Kimhi is a senior lecturer and currently the Acting Rector of the Levinsky-Wingate Academic College (formerly the Levinsky College of Education).She taught special education both at the Levinsky College of Education and at Bar Ilan University.Her research focuses on teacher training, inclusive education, and best practice concerning diverse populations (ASD, giftedness, and children with special needs).Prior to her academic positions, she was a special education inspector and a referent inspector for ASD at the Ministry of Education in the Central District of Israel.She was a former leading national ASD counselor in Israel, leading both segregated and inclusive models for pupils with special needs and ASD.
Mόnica Lourenço is a researcher and teacher educator at the University of Aveiro (Portugal).She is co-convenor of the Special Interest Group on Teacher Education of the Academic Network on Global Education and Learning (ANGEL -https://angel-network.net/).Her main research interests include teacher education, global citizenship, internationalization of the curriculum, plurilingualism, and education for diversity.
Suzanne Macqueen is a Senior Lecturer in the School of Education at the University of Newcastle, Australia.Originally a Primary teacher, she now teaches courses in the areas of social studies curriculum, quality teaching and professional preparation including classroom management.Her PhD examined the topic of non-traditional student experiences in higher education, using narrative inquiry and Bourdieu's thinking tools of habitus, capital and field.Suzanne is a member of the Global Education Research and Teaching (GERT) team which focuses on the promotion of and research into teaching for the development of cultural competence and global citizenship.
Valentina Pennazio is currently an associate professor of Special Education at University of Genoa (Italy).Her main research interests are focused on: teacher education, special education and technologies for inclusive education An award-winning teacher, Laura Sokal has published over 75 peer-reviewed research articles and 4 edited books on the psycho-social needs of today's students.Her SSHRC-funded research program includes studies ranging from adjustment in pre-school children to mental health issues in post-secondary students to burnout in Canadian teachers.Laura's extensive community service includes service on boards and committees concerned with social justice and creating meaningful and fair opportunities for under-served children, youth, and adults.Aside from working in schools in Canada and teaching in Australia, Turkey, Germany, and Nicaragua, Laura has served as the University of Winnipeg's Associate Dean of Education, and as a Child Life Therapist at Children's hospital.In her current role as Professor of Education at the University of Winnipeg, she enjoys learning with and from her students.
Renata Timkova is an Associate Professor at the Department of British and American Studies of the Pavol Jozef Šafárik University, Slovakia.Her main research interests are focused on linguistics, second language acquisition, Slovak language as a foreign language, student teacher trainingspeaking and listening skills, language interference and academic English.As a member of board of experts she is responsible for the delivery, development and quality assurance of the study programme Teaching of English Language and Literature (joint degree studies) at her home university.She is the author and co-author of several scientific studies, articles and university textbooks Sina Westa works currently as a research associate and project coordinator for the project "QUALITEACH" at the Universität Erfurt, Erfurt School of Education.As a research fellow she has obtained her PhD in 2017 within the Marie Curie Network for Initial Training "Universities in the Knowledge Economy" at the University of Ljubljana, Faculty of Education, Slovenia.She holds a MSc in Applied Social and Community Research from the University of Brighton, UK; and a 1 st State-Examination for Primary School Teaching, Ludwig-Maximillians-University Munich, Germany.Her research interests lie in the area of academic values, values in education, internationalization of higher education, teacher education, teaching profession, and policy studies.Sina Westa's research focuses on qualitative research from an international and interdisciplinary perspective.

Free additional comments
This box may be used to write additional qualitative comments.

Caption
The levels are structured as follows: not applicable -I'm not involved in this criterion.emerging -I show a low willingness to act in relation to the criterion.developing -I show a willingness to act in relation to the criterion but I tend to give up and not to deal with it thoroughly.achieving -I thoroughly act in relation to the criterion.extending -I thoroughly act in relation to the criterion, I extend and practice it independently.
Wikan is a professor in Geography.Her research interests are Global competence, Global citizenship, intercultural competence in teacher education Area C. Global competence -m able to reflect deeply on the ways that I think about myself C1b I'm able to reflect deeply on the ways that I think about the curriculum design and the teaching m able to interact and cooperate with colleagues, students, parents, etc. from diverse backgrounds C2b I'm able to interact and cooperate with colleagues, students, parents, etc. from diverse linguistic backgrounds C3 managing complex learning environments C3a I'm able to observe the features of several learning environments and critically analyse diverse school contexts and systems C3b I'm able to create effective learning environments and manage classes with students from diverse backgrounds C3c I'm able to adapt my teaching strategies to several educational situations C4 intercultural teaching C4a I'm able to design instruction that matches my students' developmental needs C4b I'm able to critically examine the curriculum to determine whether it reinforces negative cultural stereotypes C4c I'm able to create learning environments where everybody can develop plural multifaceted learning, considering different points of view C4d I'm able to design a learning environment that embraces cultural diversity C4e I'm able to use experiences and perspectives of diverse students as conduits for teaching more effectively C5 international practice C5a I'm able to practice in international school contexts C5b I'm able to transfer into the school system of origin that I observed during the internshipm able to adopt interactive and cooperative strategies with students from diverse backgrounds C6b I'm able to carry out inquiry-based models of teaching to enable students from diverse backgrounds to actively work on ideas in order to construct knowledge, solve problems, and develop their own understanding of the content C6c I'm able to support students from diverse backgrounds in working together on communitybased authentic projects and real-world experiences C6d I'm able to develop global learning through discussions about news events occurring around the globe and to connect them to classroom subjects C7 interactive assessment strategies I'm able to design and implement formative assessment methods to inform instruction with students from diverse backgrounds (self & peer assessment, portfolios, etc.)

Table 1 .
Participants and institutions.

Table 2 .
Demographic and school/educational variables.

Table 3 .
Reliability Coefficients for English Levels.

Table 5 .
Frequentist individual item reliability statistics.

Table 6 .
Factor loadings of EFA without a forced solution.
Figure 1.Factorial components' plot in the rotated space.

Table 7 .
Factor loadings of EFA with forced 3-factor solution.The remaining indicators/criteria of Area B are included into Factor 2 and, lastly, Factor 3 is composed of all indicators/criteria of Area A.

Table 8 .
Results summary of CFA models.

Table 9 .
CFA results and decision matrix.

Table 10 .
Differences between groups.