Reflections on Addressing Educational Inequalities Through the Co-Creation of a Rubric for Assessing Children’s Plurilingual and Intercultural Competence

Abstract

Recognising linguistic diversity as a person’s characteristic is arguably central to their multilingual identity and is important as an equity issue. Different indicators suggest that students with migrant backgrounds, whose linguistic diversity is often not reflected in European education systems, tend to underperform compared to their peers without migrant backgrounds. There is a dire need, therefore, to alleviate the educational inequalities that negatively affect some of the most plurilingual students in European school systems. This can be carried out by revisiting assessment tools. Developing assessments to make children’s full linguistic and cultural repertoire visible, and what they can do with it, is one way that potential inequalities in school systems and assessment practices can be addressed so that cultural and linguistic responsiveness of assessments and practices can be improved. This paper explores the concept of discontinuities or mismatches between the assessment of plurilingual children’s linguistic practices in one primary school in Catalonia and their actual linguistic realities, including heritage languages. It asks: (1) What are the children’s linguistic profiles? (2) What mismatches and/or educational inequalities do they experience? and (3) How does the co-creation and use of a rubric assessing plurilingual and intercultural competence attempt to mitigate these mismatches and inequalities? Mismatches are identified using a context- and participant-relevant reflection tool, based on 18 reflective questions related to aspects of social justice. Results highlight that mismatches exist between children’s plurilingual and intercultural knowledge and skills compared to the school, education system, curriculum, and wider regional and European policy. These mismatches highlight two plurilingual visions for language education. The paper highlights how language assessment tools and practices can be made more culturally and linguistically fair for plurilingual children with migration backgrounds.

1. Introduction

Within the field of education, it is important to consider and recognise diversity as part of learners’ multilingual identity (Forbes et al., 2021). Failing to do so, as is currently the case in many European contexts (Moore & Bernaus, 2021), can result in inequalities for plurilingual1 children’s academic attainment levels. In Catalonia (Spain), where this study is focused, many plurilingual children are expected to become increasingly plurilingual and intercultural with (potentially additional) school languages (i.e., Catalan, Spanish, and English) forming part of their repertoire at primary school level. However, children may already have an extensive linguistic repertoire beyond the objective of the European Union’s (EU) language policy for every EU citizen to master two languages in addition to their “mother tongue”2 (European Parliament, 2025). Therefore, the plurilingualism seen in the education system in Catalonia does not appear to cover or align with all students’ linguistic backgrounds.

Such misalignment can cause potential educational inequalities and, therefore, can be considered a social justice (SJ) issue, defined as the equitable sharing of social power and benefits within a society (Osborn, 2006). Within the education context, SJ can include aspects such as curricular elements or instructional choices, such as vehicular languages used. Within this context, misalignment can be detrimental for students with migration backgrounds, who, according to different indicators, are underperforming compared to their non-migration background peers (European Commission/EACEA/Eurydice, 2019). Indeed, many children from migration backgrounds (understood as first-generation migrants born in a country that their parents were not or children who migrated with their families) demonstrate “emergent competence in the language(s) of instruction […] and are already highly disadvantaged in the education system” (Erling & Moore, 2021, p. 530; echoing Gerszon et al., 2020). Identifying mismatches that may lead to inequalities is an especially relevant issue for schools and policy makers, considering that classes are becoming more “super-diverse” (Vertovec, 2007) than ever. ‘One-size fits all’ assessment approaches (Karavas & Mitsikopoulou, 2018), accounting for single school languages, may not be representative of many plurilingual children’s repertoires, which include different combinations of societal, school, and heritage languages.

Cultural mismatch theory (Stephens et al., 2012a, 2012b) echoed this idea of discontinuities/misalignments. Addressing mismatches or discontinuities between plurilingual children’s homes and schools is a necessity if equitable education is to be a norm. Therefore, one challenge of the school system is to recognise children’s linguistic and cultural diversity as involving school-taught languages and related cultures as well as children’s heritage languages (HLs) and cultures3.

One way to do this is by revisiting assessment tools as they can give visibility to some competences over others. The fact that single school-taught language(s) are formally assessed in many schools in Europe, without explicitly considering children’s non-school languages or related cultures, needs to be revisited if school systems are to address linguistic diversity as an equity issue. Developing assessments to make visible children’s full linguistic and cultural repertoire, and what they can do with it, is one way that this mismatch can be addressed. This scenario highlights the need for more just assessment tools and practices such as those proposed by culturally responsive education (CRE). CRE focuses on discontinuities between school, home, and community and the relationship with children’s achievement (Gay, 2018). Also pertinent to plurilingual children is linguistic sensitive education (LSE), which recognises that students come from different linguistic backgrounds and may have varying levels of proficiency in the language of instruction (Bleichenbacher et al., 2023). According to CRE theory, if schools and teachers reflect and draw on children’s cultural and language strengths, students’ academic achievement will increase (Gay, 2018).

Thus, identifying and addressing mismatches or discontinuities can contribute to the process of developing culturally and linguistically responsive assessment practices and tools. However, to our knowledge, mismatches have not been explicitly addressed in studies on plurilingual children with migration backgrounds. The goal of this study is to identify which mismatches/inequalities exist for plurilingual children, related to assessment tools and practices, in one primary school in Catalonia (Spain) using reflective questions related to different aspects of SJ. Reflective questions were formulated by researchers especially for this purpose and applied to an assessment rubric co-created by parents/carers, children, and teachers, designed to assess children’s plurilingual and intercultural competence. The study also seeks to reflect on how the rubric co-creation process, and final rubric version, attempted to mitigate the mismatches identified using the reflection questions.

2. Theoretical Framework

2.1. Social Justice and Language Education Assessment

SJ can be defined as the equitable sharing of social power and benefits within a society (Osborn, 2006). In language education, “this would include the curricular elements as well as the instructional choices implemented to aid in that endeavour” (Randolph & Johnson, 2017, p. 100). However, aspects of SJ in the area of assessment can also consider tools and practices.

Ortega (2020) highlights that an SJ perspective underscores the need for “a commitment […] to promote human values of equality and justice” highlighted in psychology by Vasquez (2012, p. 337). This has given rise to “a philosophy, an approach, and actions that embody treating all people with fairness, respect, dignity, and generosity” in education (Nieto & Bode, 2012, p. 12). In the field of second language acquisition (SLA), an SJ lens reveals the role that systemic inequities play in any kind of language learning (Piller, 2016) and an explicit goal of research is to support equitable multilingualism (Ortega, 2020).

Studies on language assessment for SJ include various aspects that contribute to what may be considered fair or just assessment tools, practices, and policies at any given time. Studies have focused on a wide variety of aspects, including democratising assessment (Palmer et al., 2019; Randolph & Johnson, 2017), teachers’ practices (Randolph & Johnson, 2017; Tavares, 2023), monolingual mindsets (Cleave, 2020; Erling & Moore, 2021), the tests themselves (Bagio Furtoso et al., 2023; Gardner et al., 2009), approaches to assessments (Dendrinos, 2019; Gorter & Cenoz, 2017), education systems (Cleave, 2020; Erling & Moore, 2021), and accessibility issues (Cleave, 2020; Commission of Europe, 2018b), amongst others. Indeed, according to McArthur (2015), assessment for SJ can be understood as an ongoing commitment to problematising issues of justice and assessment rather than pursuing final solutions. The diverse research focus reflects this ongoing problematisation, and we now outline the most relevant studies related to plurilingual children with migration backgrounds and the context of this study.

Palmer et al. (2019) and Randolph and Johnson (2017) highlight the need to democratise language assessment and take a critical view, building on the work of Shohamy (2001), Roever and McNamara (2006), and Inbar (2008), amongst others. This is because, according to Bagio Furtoso et al. (2023), “assessment exerts great power in people’s lives, and therefore, language assessment should help diminish social inequalities, not reinforce them” (p. 162).

Gardner et al. (2009) reviewed educational assessments for school-age students. Issues raised included: (1) access to assessment opportunities; (2) misuse of aggregate results of summative assessments by the media and politicians consolidating unfair/inaccurate stereotypes; (3) the proposal that summative assessments without feedback do little to address the impact of SJ; (4) pedagogically linked classroom approaches, such as assessment for learning, peer/self-assessment; (5) ensuring that content of examinations is appropriate (non-oppressive, meaningful, accessible to all, and non-offensive); (6) a tension between what a student can learn and what their school may be capable of assessing.

A study, by Cleave (2020), focused on how a more equitable, inclusive, and coherent education system (in England) could achieve better outcomes for plurilingual learners. Key findings included: (1) facilitating learner voice as engagement with multilingual learners and families to create inclusive education systems, owned by their diverse, multilingual communities; (2) having teacher/practitioner communities who value multilingualism in the classroom as a learning resource and asset; (3) diverse and shared leadership that represents the classrooms and communities; (4) asset-based approaches as systems and programmes that build on the assets of multilingual learners that can help shift schools away from a monolingual mindset; and (5) learners having equitable access to the language of instruction as well as multiple language proficiency. The report emphasised the importance of the school language mastery and proposed official recognition/awards for bilingualism and bi/pluriliteracy because, according to the author, fluency in other languages opens opportunities for academic success, prepares learners for 21st century careers, and allows them to participate fully in an increasingly globalised and interconnected world (Cleave, 2020). While Cleave’s report highlights important aspects that could improve plurilingual children’s achievements in language assessments, in the context of the current study, many children are already participating in a globalised and interconnected world in their own class and communities, exposed to a wide range of languages and cultures: it is not something they are going into but rather something they already inhabit in their daily interactions with each other. This highlights what can be considered as intercultural communicative competence (ICC) which many children in the context of this study engage in. ICC “refers to human beings who communicate with each other but do not share the same linguistic repertoire or cultural environment” (Infoscipedia, 2024). This is important because both intercultural competence and plurilingual competence form part of what children are being assessed for, in the case of the present study, using the co-created rubric.

Linked to the assessment of ICC, Bagio Furtoso et al. (2023) explored the underlying SJ agenda by analysing the assessment of an ICC exam task (the Celpe-Bras exam). Using the premise that language and culture are inseparable, they showed that producing texts (output) in a given language is a movement of mobilising and articulating cultural background knowledge with the information presented in the source texts (input). They propose that the exam, based on ICC, can promote SJ because ICC can raise issues of “socially constructed and dialogically negotiated identity markers” (Bagio Furtoso et al., 2023, p. 162), a pillar of SJ language education. Given that the children in the context of this study are linked through a variety of languages to different cultures, it is conceivable that they mobilise and articulate cultural background knowledge through their various languages.

In the field of teaching English as a foreign language (TESOL), which includes assessment, Tavares (2023) highlighted the presence of coloniality, social injustice, inequalities, and monolingualism to name a few, echoing previous authors (e.g., Kubota, 2019; Pennycook & Makoni, 2019; Phan & Barnawi, 2022). Tavares (2023) suggests that TESOL can be transformed by focusing on students’ linguistic repertoires; adopting decolonial pedagogies; deconstructing teacher education and educators; engaging in regional knowledge in pedagogy creation; and reimagining TESOL through different perspectives. These aspects are relevant to the present study because the historical dimension helps to highlight the elevated position and role of English as an FL in the Catalan curriculum and, in addition, the elevated position that Spanish has as a societally dominant language over Catalan in some areas of Catalonia. Both languages need to be recognised as dominant if equitable multilingualism (Ortega, 2020) is to take place by local educators and policy makers regarding all plurilingual children in Catalonia, not just children from migration backgrounds. A historical perspective also helps to explain how a European vision of plurilingualism, understood as a collection of languages, fits a neoliberal agenda that arguably extends colonial principles: dominance of some languages which requires maintenance of fixed national identities. These identities are more easily achieved if maintenance of single languages (and achievement measured through single-language assessments) is the only legitimate representation of plurilingual repertoires available to plurilingual children.

In a language learning classroom more generally, Randolph and Johnson (2017) explored opportunities for SJ instruction. Focusing on lesson planning, course design, and professional development they offered an overview of how traditional assessment practices may be reimagined. The authors suggested that, rather than teachers creating assessments that focus on cultural knowledge, they could incorporate assessments where students were given choices and participated through self-assessment (Randolph & Johnson, 2017). This is relevant because, in the context of the current study, learners do not have a choice as to whether they can take plurilingual assessments that assess multiple languages simultaneously or which languages they want to assess. Furthermore, they are not used to taking part in self-assessment related to their linguistic repertoire and/or what they can do with it.

In relation to HL, but not focused on assessment specifically, Ortega (2020) argued that heritage language development (HLD) is an SJ issue and proposed that HL speakers and their languages must be understood as connected to minoritised communities and to the experience of inequitable multilingualism. This is for several reasons: “their access to mature HL speakers and rich HL input is impacted numerically negatively” (Ortega, 2020, p. 38) and their multilingual learning is inequitable in that it is “rooted in systemic marginalization” (Ortega, 2019, p. 27). This is relevant to this current study because many children’s HLs are not seen even when plurilingual competence is considered in the curriculum. Their HLs can be considered a minority language while Catalan can also be considered as one.

We now turn to SJ and plurilingual education in Europe because it is the broader context where the current study takes place. Erling and Moore (2021) link plurilingual education to critical work on neoliberalism, language, and education. In neoliberal ideology, language education is viewed in terms of standardised curricula with assessment and learning objectives to produce workers that meet the demands of changing workplaces (Erling & Moore, 2021) and supports competitiveness, echoing Heller and Duchêne (2016). Language is also regarded as “capital—for individuals, companies, nations which in turn links to discourses of linguistic instrumentalism” (Erling & Moore, 2021, p. 524), a concept first coined by Kubota (2011). Linguistic instrumentalism emphasises the importance of learning particular languages, such as English, and varieties of them for individual successes as learners and workers (Kubota, 2011). In this line, Flores (2013) suggests that “plurilingualism may unwittingly be used as a tool of neoliberal governance that reinforces rather than challenges current relations of power” (p. 509), offering a critical review of the emergence and circulation of the European ideal of plurilingualism. This can remind educators to question institutional discourses and policies at a local/regional level and can partly explain why languages, such as English, retain such a strong presence as the preferred foreign language in state school curriculums, such as in Catalonia. The European ideal of plurilingualism is arguably supported by single-language mindsets, with English as an instrumental language for work in the EU, Catalan for work in the region, and Spanish for work in both the region and nation. There is an absence of official discourse or practices related to achievements in other languages/cultures that children may already know. This brings us onto how plurilingual children’s linguistic and related intercultural skills are currently assessed in our context.

2.2. Assessment of Children’s Plurilingual and Intercultural Competence in Relation to SJ

SJ is relevant for language education, not only in assessing single languages but also when assessing children’s plurilingual and intercultural competence. There are a several studies focusing on multilingual testing and assessment in different social and educational contexts (for overviews see Dendrinos, 2019; Gorter & Cenoz, 2017; Melo-Pfeifer & Ollivier, 2023). These studies highlight considerations regarding assessment of children’s plurilingual and intercultural competence in relation to current language competence assessments, testing practices, and equity issues.

Several authors have highlighted that language competence assessment and testing practices remain monolingual (Dendrinos, 2013; Shohamy, 2009). The absence of multilingual approaches in assessment and testing (Dendrinos, 2019; Gorter & Cenoz, 2017) is due to the belief that making the multilingual turn in the field of testing and assessment is more challenging than it was to introduce it in teaching practices. Tests and exams are predominantly designed to measure competences and skills that learners have developed in single languages and “language teachers and testers do not know how to assess language skills or content knowledge using languages in combination” (Dendrinos, 2019, p. 3).

Plurilinguals develop specific meta- and cross-language competencies which appear underrepresented in current assessments. These can include calling “flexibly upon different parts of this competence to achieve effective communication with a particular interlocutor e.g., partners may switch from one language or dialect to another; or a person may call upon the knowledge of a number of languages to make sense of a text” (Council of Europe et al., 2001, p. 4). Similarly, cross-language interactions occur in the mind of a plurilingual learner/user “to draw upon resources from all their languages, to make use of analogies between languages (such as in the case of cognate relationships), to draw cross-lingual inferences, transfer knowledge and translate” (Hofer & Jessner, 2019, p. 6). These skills and knowledge form part of determining the performance of the plurilingual’s multi(lingual) system (Herdina & Jessner, 2002). However, meta- and cross-language competences are not generally measured in European schools. Hofer and Jessner (2019) suggest that assessment of these multi(lingual) competences does not need to replace (all) extant monolingual paradigms but can complement and improve existing practices.

Linked to this point is the fact that different strategies for plurilingual/multilingual competence assessment are conceivably an equity issue because certain assessment strategies for plurilingual children may favour one plurilingual repertoire or practice over another. As highlighted by Melo-Pfeifer and Ollivier (2023), building on Shohamy’s (2011) continuum, some strategies may be focused on several languages being used but differentiated/separated. On the other end, some “tests are based on the approach in which a mixture of languages and open borders among them is recognised, accepted, and encouraged” (Shohamy, 2011, p. 427), which is, according to Shohamy (2011), more far-reaching, accepting, and legitimising. Single-language tests, on the other hand, which are added together to represent a child’s linguistic repertoire may offer a very different experience for (and vision of) learners who commonly practice shifting between two or more languages. Children who are more familiar with mobilising their plurilingual repertoire in flexible ways may not perform as well in language tests that assess singular language competence compared to plurilingual assessments that accept translanguaging practices. According to Ortega (2020), “insisting on the separation of languages can denigrate language mixing and codeswitching, which are natural bilingual practices that many HL children and youth experience in their homes and communities” (p. 28).

While separation of languages (that is being performed by the child) can be a useful skill to develop because it allows for sustained spoken or written maintenance/performance in one language, the use of multiple languages to communicate simultaneously also draws on important knowledge and skills because it allows for valuable and necessary plurilingual, linguistic mediation practices. Within an SJ lens, these practices can be normalised through assessment strategies and tools that education systems adopt.

To aid in this normalisation of the concept of (plurilingual) mediation, the CEFR highlights competence referring to a student’s multiple languages used to communicate simultaneously. This would include, for example, translating between written texts, interpreting for two people who do not share the same language or simplifying/modifying a concept in another language. These are examples of mediating a text, mediating communication, and mediating concepts, respectively (Commission of Europe, 2018a). Mediation appears in the primary Catalan curriculum in the foreign languages subject as specific competences 8 and 9, although it is more salient in the secondary school curriculum (Generalitat de Catalunya, 2022) and it is restricted to taught school languages. Such competences translate into English as follows:

-

Competence 8. Mediate between different languages in predictable situations, using strategies and knowledge to process and transmit basic and simple information, in order to facilitate communication.

-

Competence 9. Reflect in a guided manner on language and recognise and use personal linguistic repertoires through processes of comprehension and production of oral, written, and multimodal texts, using appropriate basic terminology, to begin developing linguistic awareness and improving skills in the application of these processes.

Currently, assessments do not exist for children’s plurilingual mediation skills in Catalonia which is why assessments that incorporate meta- and cross-linguistic competences, the construct encompassed in the CEFR’s plurilingual mediation, are arguably crucial in the potential development of diversifying language tests and practices for plurilingual children.

We turn now to review other critiques of assessments relevant to plurilingual children, such as the ‘one-size fits all’ approach adopted by international tests (Karavas & Mitsikopoulou, 2018) and the need to include local content, norms, and values into teaching materials and tests (Canagarajah, 2006; Dendrinos, 2013, 2015, 2019).

Furthermore, quality and equity are also issues raised by Borghetti and Barrett (2023), who highlight the difference that rubrics use can make in terms of transparency. They address “the need to make learners aware of the assessment modes […] where they fully understand the purposes and uses of the assessment” (Borghetti & Barrett, 2023, p. 3). Transparency can be achieved through learner involvement in assessment practices, such as discussing scoring systems, reading rubrics in advance, or participating in assessment tool development (Borghetti & Barrett, 2023).

Another issue is related to accessibility and behavioural appropriateness, which are addressed in the European Language Portfolio (ELP) Checklists for Young Learners (Commission of Europe, 2018b). Accessibility considers terminology, the communicative behaviour in the descriptors, the teachers as descriptor-mediators, and the purpose of self-assessment tools. Behavioural appropriateness includes whether descriptors adequately reflect the behavioural options available to learners, understood through the notion of domains (e.g., social, tourism, work, and study), which can be problematic for children if the domains are beyond their capacity or reflect tasks that they do not normally perform.

After highlighting the aspects relating to SJ, language assessments and plurilingual children in our context, we now turn to the several theories that can support teachers/researchers in understanding and working towards mitigating/overcoming potential injustices related to cultures and languages in relation to assessments.

2.3. Culturally Responsive and Linguistically Sensitive Assessment Practices

The theory of cultural difference was created by scholars from the 1970s and 1980s because they were concerned with “the disparities in academic achievement between mainstream students and students who are marginalised” (Gay, 2018, p. xii) and as a pushback against the cultural deprivation paradigm which contributed to blaming the victims for structural exclusion (Ryan, 1971). Culturally responsive education (CRE) theory postulates that the discontinuities between the school culture and the home and community cultures are an important factor in low academic achievement. Consequently, according to Gay (2018) “the academic achievement of these students will increase if schools and teachers reflect and draw on their cultural and language strengths” (p. xii).

While CRE is rooted in studies pertaining to “students of colour” (Gay, 2018), a second generation of cultural difference theorists created a theory of teaching that gives guidance to educators who are trying to improve the academic achievement of students from diverse racial, cultural, linguistic, and social-class groups. The theory is relevant to plurilingual children with migration backgrounds who may or may not fit the European ideal of plurilingualism which is reflected in school or curriculums, e.g., their plurilingual repertoire may not include languages that would fulfil a neoliberal agenda to support mobility and competitiveness within the EU, nor may it include the vehicular school language (through which achievement is predominantly formally measured). Therefore, they may face a disadvantage or inequity due to their differing plurilingual repertoires.

The idea of “discontinuities”, from CRE theory, can also be understood as a mismatch, a notion from cultural mismatch theory (Stephens et al., 2012a). This theory proposes that “when the culture of an academic institution differs significantly from a student’s family or home culture, the student experiences conflict or tension that can impact their academic performance” (Fink, 2023).

Because this current study is concerned with mismatches pertaining to linguistic and cultural aspects, we incorporate the construct of language-sensitive education (LSE) to complement CRE. LSE requires educators to be aware of the potential language barriers for students, making appropriate accommodations and adjustments to support their learning. This can include “providing bilingual or multilingual resources, using inclusive language that avoids stereotypes and bias, and acknowledging and respecting the cultural backgrounds and perspectives of all students” (Bleichenbacher et al., 2023, p. 2, citing ChatGPT 8 March 2023).

In sum, culturally and linguistically responsive assessment tools and practices would conceivably seek to reflect and draw on all children’s full cultural and linguistic strengths so that these are made visible and assessed accordingly. In addition, assessment practices would conceivably focus on overcoming the discontinuities or mismatches that occur between, on the one hand, assessment tools and practices that reflect the linguistic repertoire of plurilingual children and what they can do with it, and, on the other hand, language assessments already in place and offered through the educational system, via schools. In order to make assessment of children’s plurilingual and intercultural competence more culturally and linguistically sensitive through the identification of mismatches, three research questions were asked:

(1)

What are the linguistic profiles of the primary education students participating in the study?

(2)

What mismatches and/or inequalities exist for plurilingual children in the participating Catalan state school (understood through the lens of SJ)?

(3)

How does the assessment rubric co-creation process and/or product attempt to mitigate these inequalities?

3. Methodology

3.1. Approach

This study uses a qualitative interpretive approach. In order to identify possible mismatches/inequalities the first step was to create a reflection tool to guide in the identification of these mismatches, specifically related to the context of the study. The Culturally Responsive Classroom Assessment Inventory (Evan, 2021), based on Stembridge’s (2019) Culturally Responsive Classroom Assessment Framework, served as a model for this as it offers several questions as a reflection tool to improve the cultural responsiveness of classroom assessments. It showed that responsive assessment practices can be developed through the use of reflection tools based on questions grouped into themes. The steps for creating questions for our own tool were to first review the literature on assessment practices and SJ, particularly in Europe (which formed part of the theoretical framework) and then to select relevant aspects/findings related to SJ found in these studies that could serve as reflection prompts that were posed as questions and were deemed as relevant to our own specific context. This relevance criteria pertained to whether the aspects, such as transparency or learner voice, were: (a) applicable to regional context but linked to national or European policy; (b) applicable to the Catalan curriculum and assessment tools and practices carried out via the school; (c) related to assessing children’s language-related and intercultural competences which is what the assessment rubric aimed to evaluate; (d) applicable to the “super-diverse” (Vertovec, 2007) linguistic and cultural repertoire of children in the study; (e) considerate of Catalan as a minority language spoken in Barcelona; and (f) applicable in considering English as a foreign language within a wider agenda of how plurilingualism is promoted within the EU. From this selection-for-relevance process, 18 reflection questions were created (see Appendix A). The questions were then grouped together in 5 themes (see Appendix A), namely (1) children’s linguistic repertoire, knowledge, and skills; (2) assessment practices; (3) accessibility for learners; (4) the content of descriptors; and (5) linguistic and cultural diversity and power relations. Only the first 3 themes are reflected and discussed in this paper due to word limitations, with the fourth theme being addressed as a separate, forthcoming article by Segura and Knight.

In order to understand the data sources and the results, we also outline the process/methodology of the original study, which the reflection questions were applied to. The original study used a Participatory Action Research (PAR) approach, incorporating focus groups and pedagogical activities to facilitate co-working between teachers, parents/carers, and children in 5th and 6th grade in primary school to create a rubric for children’s plurilingual and intercultural competence. PAR was used to capture children’s full linguistic repertoire which would not be possible without the families’ and children’s perspective. Working with teachers, families, and children as partners, rather than participants, meant that (joint) rubric development was more democratic and reliable, involving users in decision-making processes and including children’s use of non-school taught languages as part of what was being assessed. Partners contributed to the rubric co-creation process in two ways: (1) by sharing linguistic and cultural realities from different perspectives to ensure alignment with the rubric content, and (2) by providing feedback on the different rubric iterations’ format and usability. A total of 8 partner sessions took place with different partner groups. Each focus group session was facilitated and guided (where necessary) by the researcher-partners. Following the PAR approach, conceptualised as a cyclical process, all partners were involved and analysis of the feedback from partners was carried out through the focus group documents, after each session. This allowed us to design materials for upcoming sessions and to gather feedback on the various improved versions of the rubric as it evolved. Further detailed description of each PAR stage can be read in the materials shared by Segura and Knight (2024) in the Open Science Framework repository.

3.2. Partners

Participants in the original study were considered as partners and included 17 teachers, 9 parents, and 40 children from 5th and 6th years in a primary school in Barcelona, Catalonia (Spain).

Regarding the children-partners, it was deemed that the optimal grade students to take part in the project were those in 5th and 6th in primary, which correspond to the last two grades in the upper cycle in primary education in the Catalan education system. The rubric was designed to be used in the last year of primary education, namely 6th grade, because while children are assessed annually through yearly report cards, the final year of primary education is when three external, official, single-language exams are taken in English, Spanish, and Catalan. Therefore, this group of children was considered fit for the co-creation and piloting process.

Important to this study is the linguistically and culturally diverse profile of the children. Many children have a diverse linguistic and cultural repertoire because most of them have parents/carers who are first generation migrants, so the children have grown up or were born in Catalonia. Alternatively, many children have recently migrated with their parents/carers. Many children are emergent speakers of school languages if these do not form part of their HLs. Many also belong to low-income families, but not all. Some children have Catalan as an HL along with Spanish and/no other HLs. Further details about the linguistic profile of child-partners form part of Section 4 for RQ1.

3.3. Data Collection

Data sources for this study consist of the final rubric (co-created by all partners), various pedagogical activity/focus group documents for partners (created by researchers), and the partner responses to these documents (written comments and edits to the rubric descriptors and voting indicators collected during the process of rubric creation and final use. These were completed during each of the 8 focus group sessions with the partners. Written comments and voting indicators were transcribed in Excel documents for data analysis.

3.4. Data Analysis

To answer RQ1, focusing on the linguistic profiles of the student participants, data was gathered through the first pilot of the rubric with children which required them to indicate the languages they were able to speak and/or understand. This information was turned into numerical data and descriptive statistics (via counts) were reported.

RQ2 inquired about the mismatches and/or inequalities identified between home/community and school system for the plurilingual student participants. To answer this question, the researcher-partners used the 18 reflective questions (see Appendix A) as a tool to revisit and reflect on the data sources from the focus/group alongside our knowledge of the PAR process. RQ3 was asked concurrently with RQ2 while researchers revisited the data sources and discussed, using the reflective questions, how the co-creation process (focus groups and activities with each partner group) and product (rubric) attempted to mitigate the mismatches identified in RQ2.

3.5. Ethical Considerations

Information sheets and consent forms for all participants in the PAR study were developed, modifying Universitat Internacional de Catalunya’s (UIC) ethics document templates. Consent forms were in Catalan for teachers, families, and children. However, the project objectives were transmitted to families not familiar with Catalan in the various HLs through the parents’ WhatsApp phone application, and participation was explicitly made voluntarily. Project aims were also explained orally in Spanish at the first parent/carers’ focus group, the main language of their oral discussion. Ethics approval was granted by UIC’s Research Ethics Committee.

4. Results and Discussion

4.1. Linguistic Profile of the Children

To answer RQ1, we present the number of languages spoken/understood by 5th and 6th year children, which were self-reported in their rubric self-evaluation (Figure 1), as well as the languages they spoke/understood.

Table 1. Number of languages spoken/understood by 5th and 6th year children (self-reported).

No of Languages	5th Year	6th Year	No of Children That Reported Speaking N Languages
1 language	0	0	0
2 languages	0	1	1
3 languages	3	5	8
4 languages	6	7	13
5 languages	2	10	12
6 languages	1	1	2
7 languages	3	0	3
8 languages	0	1	1

shows that all children were plurilingual, with most of them speaking or understanding more than three languages in both groups. The most common linguistic profile of children, numerically, represented the ability to speak or understand four and five languages. Notably, the number of languages spoken/understood by children exceeds the objective of the EU’s language policy for every EU citizen to master two languages in addition to their mother tongue (European Parliament, 2025) which can be considered an HL.

Table 2. Languages spoken/understood by 5th and 6th year children (self-reported).

Language	No of Children That Reported Speaking/Understanding the Language
	5th Year	6th Year	Total
Catalan	15	25	40
Spanish	15	25	40
English	15	24	39
French	2	10	14
Italian	6	6	12
Arabic	1	6	7
Hindi	5	0	5
Urdú	4	1	5
Chinese	0	3	3
Punjabi	2	1	3
Japanese	0	2	2
Russian	1	1	2
Tagalog	1	1	2
Bengali	1	0	1
Galician	1	0	1
German	1	0	1
Guarani	0	1	1
Ilocano	0	1	1
Korean	0	1	1
Polish	0	1	1
Portuguese	0	1	1

presents the summary of the results of the languages that the children self-reported being able to speak and/or understand. Fourteen different languages were spoken or understood in the 5th year and seventeen different languages in the 6th year. The most common languages were Catalan, Spanish, and English. Nine languages were European languages and twelve were non-European. Both the results from Table 1 and Table 2 demonstrate the linguistically “super-diverse” (Vertovec, 2007) nature of the two classes which is representative of the school more generally. The linguistically diverse nature of the school also implies that the classes are culturally diverse.

The information displayed in Table 1 and Table 2 was generated by children’s completion of the front page of the rubric as shown in Figure 1. The completion of this page, and what it represents, allowed children to visually represent all of the singular languages in their repertoire, including school languages and HLs. This was performed by ticking and totalling them in the star. This highlights the first mismatch, namely, the mismatch between children’s repertoires and the instrumental languages (Kubota, 2011) of the curriculum (Catalan, Spanish, and English in the case of Catalonia) which we will describe and interpret as a result relating to RQ2 and RQ3 on mismatches and how they are mitigated.

4.2. Reflective Questions: Identifying and Mitigating Mismatches

We now present the results of RQ2, related to identified mismatches, and RQ3, related to how mismatches were overcome/mitigated by involving all partners in the co-creation process, or by the utilisation of the rubric as a co-assessment tool. We report the results by responding to the reflective questions grouped into the first three themes outlined in our methodology, namely: (1) children’s linguistic repertoires, knowledge, and skills, (2) assessment practices, and (3) accessibility for learners. Results for RQ2 and RQ3 are presented concurrently4.

(1)

Children’s linguistic repertoires, knowledge, and skills

As seen in Table 1 and Table 2, the first mismatch related to the discrepancy between children’s repertoires and the instrumental languages (Kubota, 2011) of the curriculum. The rubric allowed children to tick and list all the languages that they spoke and/or understood, not just school languages but also their HLs (Figure 1). This gave visibility to children’s non-school languages or related cultures, which addressed linguistic diversity in a more equitable way. The linguistic profile of the children participating in the study showed that they were all highly plurilingual, speaking an average of four to five languages, and even eight. Among all the children in the 5th year, 14 different languages were spoken or understood, with that rising to 17 languages for the 6th year students. The list of languages was generated through consulting children and families in different focus group sessions and therefore different perspectives (Tavares, 2023) were central. The systematic marginalisation (Ortega, 2019) that occurs with HL speakers of non-school-taught languages could be overcome, albeit briefly. The languages represented in the rubric, by children, therefore reflect a decolonialised vision of language assessment, highlighted by Tavares (2023) and others, because of the great diversity of linguistic repertoires that can be represented, with no dominance or hierarchy within the repertoire.

The second mismatch relates to the European ideal of plurilingualism, both in terms of numbers of languages reached by many children and the inclusion of non-European languages. In the case of these children, their repertoire exceeds the EU recommendation of knowing three languages. These findings corroborate the need to reflect on current linguistic assessment practices that in this context relate to single languages, and school-taught languages, regardless of the plurilingual realities of schoolchildren. Consequently, this completion meant that the school, teachers, and children could reflect on the value of multilingualism and facilitated learner voice, and they were offered official recognition using an asset-based approach as highlighted by Cleave (2020). It also facilitated participation and engagement of learners in self-assessment (needs highlighted by Gardner et al., 2009).

The third mismatch in this section pertains to singular languages that are assessed in current language assessments in Catalan schools which pertain to Catalan, Spanish, or English. These languages may form all or part of children’s linguistic repertoire (or not at all). Current assessments at primary school level do not assess plurilingual mediation practices (considered as strengths) that plurilingual children carry out. In the co-produced rubric, linguistic strengths were represented by the presence of plurilingual mediation descriptors (from the CEFR), and partners could assess how strong children were in relation to each descriptor using a 4-level Likert scale with smiley faces (Figure 2). Thus, the rubric made visible and represented primary aged children’s ability to carry out mediation activities and strategies plurilingually (Figure 2, Section 1 about “General mediation”) that reflected children’s meta- and cross-linguistic competencies. In sum, the rubric allowed children to reflect the meta- and cross-language competences (highlighted by Herdina & Jessner, 2002) that they could carry out, because the descriptors for plurilingual mediation in the CEFR had been jointly chosen by the children, the parents, and the schoolteachers. Although mediation appears in the curriculum (implying plurilingual mediation) it is naturally focused on school-taught languages which this rubric overcomes because it does not specify the languages that the children mediate in.

Thus, these mismatches were mitigated both in the rubric creation process and in the rubric itself as an assessment tool. In the process of creating the rubric, children used Catalan, Spanish, English, and any other shared HL(s) to orally discuss, make decisions, and give feedback so they were reflecting and drawing on their cultural and linguistic strengths.

These results highlight a clear lack of visibility of children’s HLs (and what they can do with them) in assessments and curricula documents. This is conceivably because there is a lack of visibility of non-European languages within the EU language policy documents on plurilingualism and assessments itself. There is no mention of HLs in the curriculum which appears to reflect a systematic marginalisation (Ortega, 2019) of non-European HLs in language assessments and practices: an erasure in the discourse around many plurilingual children’s repertoire. In addition, children’s ICC is not currently assessed, which has the potential to be seen as an achievement of plurilingual children in “super-diverse” (Vertovec, 2007) contexts. If HLs are omitted from plurilingual and intercultural assessments, we propose that it risks not supporting their full plurilingual identity and/or recognising achievement in relation to their potentially highly developed ICC. Assessing children’s ICC would be a way of highlighting many children’s achievement in this area, which the focus on single-language assessments cannot. Lack of visibility of HLs other than school languages in assessments can be somewhat mitigated by highlighting the number of languages that children speak or understand, demonstrating their (potentially high) levels of plurilingual and intercultural competence (self-, parent-, and teacher-reported), and including cross-language competence compared to single-language tests. This in turn harnesses children’s plurilingual identity and sheds light on achievement in these areas, which currently remains unseen for many plurilingual children.

The results also highlight that academic achievement (in the form of single-language tests which dominate many education systems in Europe, including Catalonia) matches some children’s cultural and language strengths but possibly not all. Language competence assessment and testing practices are school-language- and single-language-based. In addition, assessing plurilingual mediation—with school-taught or non-school-taught languages—is not a current practice. We propose that a child who has strengths in school-taught languages may do better academically compared to a child who understands/speaks eight languages or a child whose linguistic repertoire is not as strong in the taught school languages. In addition, a child who is used to maintaining linguistic separation between languages may achieve more than a child who is adept at shuttling between a range of languages frequently, according to speakers and purposes. Having the relevant concepts (i.e., plurilingual mediation) in the CEFR Companion Volume (Commission of Europe, 2018a) can support the assessment of meta- and cross-language functional competences as part of their assessment of plurilingual competences. However, until assessment tools reflect these meta- and cross-language competences (such as through acknowledging language mixing), what children can do with their full repertoire cannot be made visible and therefore recognised as an achievement.

(2)

Assessment practices

In the current Catalan education system, academic achievement in the 6th year of primary school is determined by three summative tests in Catalan, Spanish, and English. As such, state-level and school-specific tests are designed to measure the competences and skills that learners have developed in single languages (in line with the expected learning objectives set in the Catalan primary school curriculum). These languages are emphasised for individual successes (Kubota, 2011). Two challenges that lead to education inequalities/mismatches arise from this system, and attempts were made to mitigate them in the rubric co-creation project.

The first challenge for teachers is that they cannot evaluate all children’s repertoires to the fullest because no assessment tool exists to do so, and teachers do not have the language skills to assess children’s HLs beyond the three main school languages, a need highlighted by Dendrinos (2019). By including children (and parents/carers) in the co-creation of assessment tools, this mismatch of knowledge could be somewhat mitigated, as it gave them the chance to propose assessment items/descriptors that included non-school-taught languages (otherwise unknown by teachers). By using the co-created rubric, achievement was co-evaluated subjectively by children, parents, and teachers, assessing the language practices that were familiar to the child to present an inside (teacher assessment) and outside of school (parent assessment) picture of the child’s (inter)cultural strengths and full linguistic repertoire (child assessment). Thus, when using the rubric, all partners (teachers, students, and parents) assessed the child, which allowed for triangulation/comparison of all the assessments, overcoming teachers’ limitations to assess children’s strengths in non-school languages and out-of-school activities. A “Checklist page” (see Figure 3) at the end of the rubric, in which children could tick examples of their life events and activities, norms, and values, was also incorporated at the end of the rubric. These examples came from children themselves during focus group activities and their incorporation supported a more personalised and holistic representation and helped to mitigate a ‘one-size-fits-all’ approach (Karavas & Mitsikopoulou, 2018).

The second challenge relates to teachers’ assessment of children’s languages as being part of a repertoire, instead of separate languages. Currently, language teachers (based on our experience of working as partners in the original project) do not know how to assess language skills using languages in combination. Working with descriptors related to plurilingual mediation can support the acknowledgment that, for instance, translanguaging is a common practice by plurilingual speakers (Ortega, 2020). Following this, all the descriptors included in the rubric involved the children performing the actions in the descriptors using at least two languages, such as transmitting information, summarising, or explaining texts originally in one language using another target one. Therefore, the rubric accounted for simultaneous language use, instead of single language use, while not specifying any specific language. Teachers were also not familiar with being descriptor-mediators (Commission of Europe, 2018b) because self-assessment was not a common school practice and assessment of plurilingual mediation is a relatively new competence they need to assess. Teachers need to fully understand it before being able to convey it to children. The creation process, where teachers were involved in choosing and modifying relevant descriptors and giving examples to children, supported their role as descriptor-mediators.

Following these reflections, it is important to note that, although linguistic instrumentalism (Kubota, 2011) is evident in the rubric being in Catalan, the main schooling language by law in state schools, there is no mention in the rubric of this language being the focus of assessment over other school languages (Spanish and English) or children’s other HLs.

This allowed for a more democratic (Shohamy, 2001) plurilingual assessment to be carried out because assessors could include any (and any number of) languages in the assessment, depending on each child’s repertoire.

Finally, in relation to assessment practices, current assessments do not involve children or families in assessment activities, such as discussing terminology or developing assessment tools. Summative assessment, the most common way of assessing academic achievement in our context, has the main purpose of objective assessment of students’ learning. However, it is normally teacher assessment only, leading to learners not being familiar with self-assessment or working with rubrics. Similarly, parents are merely recipients of teacher reports, but not normally asked to assess their children’s competences. In these cases, the purposes of assessment are not usually explained to the learners nor the parents, who also do not take part in the development of assessment tools, which typically relies solely on teachers or policy makers.

In the rubric creation project, to guarantee a more democratic (Shohamy, 2001) and transparent assessment (Borghetti & Barrett, 2023) process, both children and parents were involved in the development of the assessment rubric, which they later used to co-assess, providing different perspectives (Tavares, 2023). The purpose of the rubric was explained to children and parents before they were involved in the rubric co-creation (Figure 4) through introductory activities targeting key concepts and PowerPoint slides with images. Then, children and parents worked on the descriptors included in the rubric, by adding, eliminating, or modifying them, as well as giving feedback about the layout and format of the rubric to guarantee usability. Including families and children’s expression of their full linguistic repertoire in school assessments can therefore help address issues of visibility and representation in assessment practices and tools.

(3)

Accessibility for learners

The language used in assessment tools (i.e., teacher assessment rubrics) is generally rather technical, as they are often used by trained teachers. These assessment rubrics need to be adapted if they are to become self-assessment tools for young children in primary school, like in the case of the present study, or their parents whose language/s is/are not the school’s vehicular language.

In order to make the terminology accessible while maintaining the meaning, partners’ input in the rubric co-creation process was essential. Through the focus group activities, the (accessibility of) language of each descriptor and its communicative behaviour in the rubric was checked and discussed with all partners, following the same procedure detailed next. The terminology was assessed at two different stages. First, children were asked to review the written descriptors, make changes, and highlight the words they did not understand (steps 3 and 4 in Figure 5). Second, after piloting the rubric, they were asked to assess language comprehensibility (see Figure 6, items C, D, and E) via a voting system and written comments. In both cases, teachers (and researchers) became descriptor-mediators (Commission of Europe, 2018b) by accompanying children’s discussions, during which they clarified concepts and gave examples, when necessary. The same procedure was followed with parents and teachers in different focus group sessions.

Involving the children (and all partners) in text editing processes ensured accessibility of terminology, which was especially important in the context of the present study: super-diverse, (Vertovec, 2007) classes of children speaking up to five or six schooling and HLs. But, more importantly, it was important because children (and their parents/carers) showed different proficiency levels in Catalan (as corroborated by the schoolteachers). Hence, the Catalan was simplified (by children, teachers, and researchers) to make the rubric’s descriptors simpler/clearer to make it more accessible in terms of terminology used, as highlighted by the Commission of Europe (2018b).

5. Conclusions

This study aimed to explore what mismatches and/or inequalities exist for plurilingual children when assessing plurilingual and intercultural competence in one Catalan state school (understood through the lens of SJ) and how a rubric co-creation process and product, aimed at assessing plurilingual and intercultural competence, attempted to mitigate these mismatches/inequalities. The identification of mismatches was carried out through the creation and use of reflective questions based on aspects of SJ and assessment deemed relevant to a specific sociolinguistic context.

Several areas of potential mismatches were identified relating to (1) children’s linguistic repertoires, knowledge, and skills; (2) assessment practices; (3) accessibility for learners; (4) the content of the descriptors; and (5) linguistic and cultural diversity and power relations. Three of these areas were reflected on in the current paper so that cultural and linguistic responsiveness of classroom assessments can be improved.

The paper also showed how the co-creation of an assessment rubric with families and children can mitigate some of these mismatches/inequalities by making children’s full repertoire visible, incorporating plurilingual mediation into assessments, and involving children and families in co-assessment practices. These actions “acknowledge and respect […] cultural backgrounds and perspectives of all students” (Bleichenbacher et al., 2023, p. 2) which is a central aim of language-sensitive education. The selection and use of reflective questions can conceivably guide the creation of future assessment tools and practices in being culturally and linguistically responsive, supporting reflection on key SJ aspects and drawing on children’s full cultural and linguistic strengths so that these are made visible and encompass a variety of plurilingual communication practices, including simultaneous use of two or more languages.

The rubric co-creation showed how all stakeholders could reflect upon and act towards greater equity in society as highlighted by Randolph and Johnson (2017) through reimagining (Tavares, 2023) what fairer language assessing should include, when focused on children’s plurilingual realities. A main conclusion is that current assessment tools and practices in the context of this study reflect not a monolingual mindset (Cleave, 2020) but rather a narrow plurilingual mindset, involving single, separate languages, which is Eurocentric and non-inclusive of the whole linguistic and cultural repertoire of some children. This suggests that there are competing visions of which plurilingualisms are acceptable: one vision focused on European languages and European identity and another, open to all languages and the possibility for representing plurilingual identities without borders. These differing visions may ultimately inform the creation of future plurilingual assessment tools and practices.

Future research and assessment development could focus on overcoming mismatches that occur between, on the one hand, assessment tools and practices that reflect the linguistic repertoire of plurilingual children and what they can do with these languages (e.g., plurilingual mediation) and, on the other hand, language assessments already in place and offered through the educational system, via schools. Failing to do so may lead to inequalities in achievements between children with different languages and linguistic practices.

Finally, this study offers an example of how reflection questions can support the identification of mismatches, but authors are mindful of the need for an ongoing commitment to problematising issues of justice (McArthur, 2015) because the aspects reflected on are only one way of attempting to mitigate some mismatches relating to assessments in Europe. Despite this, we hope that the project shared, and the reflective questions addressed, can lead to further reflection and action on assessment and SJ so that more diverse assessments and approaches for current and future plurilingual children, particularly with migration backgrounds, can take place.