Teacher assessment literacy in culturally and linguistically diverse classrooms: A Norwegian case study

This study investigated inclusive teacher assessment practices in culturally and linguistically diverse Norwegian lower secondary schools as well as tensions in and the potential of teacher assessment literacy. Case study data gathered through interviews with 21 teachers in five schools were analysed to examine the schools ’ assessment policies, teacher assessment practices, adaptions and challenges teachers face when assessing diverse student groups. We found that teacher assessment literacy development benefitted from collective and shared whole-school practices and that, while teachers had multiple strategies for assessment and adapting those to their students ’ needs, these were not specifically targeted at cultural diversity.


Introduction
As classrooms are becoming more culturally complex and linguistically diverse, teachers need to develop assessment practices and strategies that further all students' learning.We argue that such practices must be inclusive and responsive to students' culturally and socially based knowledge and understanding to facilitate education for all.This principle constitutes a key component of the Nordic education model, in which inclusion and adaption are fundamental (Blossing et al., 2014).Previous research indicates that teachers find the implementation of inclusive education challenging, particularly accommodating all students in diverse classrooms and giving them equal opportunities to demonstrate their competence in assessment situations (Florian & Beaton, 2018;Herzog-Punzenberger et al., 2020;Ysenbaert et al., 2020).Nevertheless, scant research exists that specifies what inclusive assessment in culturally and linguistically diverse (CLD) classrooms entails and the implications of inclusive assessment for teacher assessment literacy.
Inclusion in this paper is understood as meeting the social and academic needs of all students.Therefore, it relates to the multitude of differences in students' backgrounds and abilities (i.e.diversity) inherent in the classroom (Göransson & Nilholm, 2014).Inclusion entails not seeing student differences as deficits but instead emphasising equity by building on their strengths, by focussing on giving them the possibility to participate, build on their differences and learn to the best of their ability.However, teachers' understanding of inclusion sometimes works against their understanding of adaption by creating a tension between wanting to realise inclusion by treating all students the same (equality) and wanting to treat students differently by adapting their teaching to offer equal opportunities (equity) (Nortvedt & Wiese, 2020).Consequently, to achieve equity, adaptions and accommodations must be made to address the different skills, abilities and ways of thinking represented in each student group (Nilholm, 2021).However, when treating students differently teachers must pay attention to and avoid perpetuating and cementing socially and culturally based inequalities (Milner, 2019).
This paper presents an analysis of the practices, strategies and challenges reported by Norwegian case study participants 1 when reflecting on teaching and assessing students in CLD classrooms, focussing particularly on students with a migration backgroundthat is, students who were born outside the country and migrated themselves or whose parents have migrated (Organisation for Economic Co-operation and Development [OECD], 2015).Migrants arrive with a myriad of social and cultural backgrounds, social statuses and educational and life trajectories (Vertovec, 2007), and they master the Norwegian language to varying degrees.Thus, migrant students are not a uniform group but rather a diverse group with multiple characteristics, needs and challenges (OECD, 2015).Moreover, increased migration has caused classrooms to become even more culturally complex and linguistically diverse, highlighting the importance of further promoting inclusion in Norwegian classrooms.
Inclusive education is an overarching principle in Norway.Approximately 97% of Norwegian students attend their local schools, so classrooms reflect inherently diverse cultural and parental backgrounds.Currently, 19% of students in compulsory education in Norway have a migrant background (NDET, 2021), and about 6% attend Norwegian as a second language (NSL) classes (Bufdir, 2020).Hence, standardised solutions cannot be uniformly applied when seeking to provide all students with equal opportunities for participation in assessment situations.Instead, teachers must adapt or differentiate assessments to accommodate individual students' needs.
This paper adds to the knowledge on what inclusive assessment in CLD classrooms entails and highlights the related assessment literacy skills and knowledge that teachers need to fairly assess students with diverse backgrounds.Using data from teacher interviews in five case schools, this paper specifically focusses on teacher assessment literacy in assessing migrant students by addressing the following research questions: 1) Which school assessment policies and guidelines do teachers report, and how do they influence their assessment practices?2) What assessment practices do teachers in Norwegian lower secondary schools use to assess students in CLD classrooms? 3) What are the tensions and potentials teachers report related to assessing migrant students, and how may these factors necessitate changes in teachers' assessment strategies and practices for adapting assessment formats and situations?Drawing on the findings from the above questions, the overarching aim of this paper is to discuss how teachers can achieve inclusive assessment in CLD classrooms as well as the implications for the development of teacher assessment literacy.

Teacher assessment literacy
Assessment, teaching and learning activities are three aspects of classroom practice that offer students opportunities to learn (Black & Wiliam, 2018).While students sometimes act as assessors in peer-and self-assessments, most assessments are conducted by teachers, who direct students on how to attain learning objectives.Here, classroom assessment is understood as "formal and informal procedures that teachers employ in an effort to make accurate inferences about what their students know and can do" (Popham, 2009, p. 6).When designing and performing these assessments, teachers draw on their assessment literacy, which is a composite, complex skill that encompasses teachers' "capabilities to plan and implement quality assessment tasks, to interpret evidence and outcomes appropriate to the assessment purpose and type, and to engage students themselves as active participants in [the] assessment of their own learning" (Looney et al., 2018, p. 443). DeLuca et al. (2019) emphasised that the "contemporary conception of assessment literacy … recognizes that teachers make assessment decisions based on an interplay of technical knowledge and skills as well as social and contextual elements" (p.3).This contemporary understanding extends beyond the technical (Looney et al., 2018;Pastore & Andrade, 2019), which "requires looking beyond teachers' knowledge in assessment and … investigating teacher approaches to assessments in relation to their classroom teaching and learning contexts" (DeLuca et al., 2019, p. 2).
Hence, assessment-literate teachers (DeLuca, 2019;Pastore & Andrade, 2019;Xu & Brown, 2016) must know multiple ways to represent knowledge (for both students and teachers) and understand different assessment formats and the forms of learning they allow insights into to construct valid, reliable assessments that consider the contextual nature of knowledge and learning in diverse classrooms, where student expressions and ways of representing knowledge may vary (Popham, 2009).Moreover, assessment-literate teachers must know how to involve students in assessments, make accommodations so students have the opportunity to participate, adapt assessments to meet students' needs and provide clearly communicated feedback that brings each student's learning forward towards the next steps.Thus, assessment literacy is closely linked to teacher professionalism, teachers' self-perceptions as assessors and their assessment identities (Coombs et al., 2020;Looney et al., 2018;Xu & Brown, 2016).

Teacher assessment practices
Teacher assessment practices encompass both teachers' visible actions and activities that occur in the classroom and the (practical) theories, values and beliefs that influence them.These influences are often tacit and refer to, for example, teachers' theories of learning, experience, expectations and norms (Black & Wiliam, 2018;Ysenbaert et al., 2020).Contextual conditions, such as school assessment policies and collective strategies, also influence teachers' assessment practices (DeLuca et al., 2019;Looney et al., 2018).
Assessment is a social activity embedded in routine classroom practices and actions, and is informed by students' and teachers' social and cultural contexts.Thus, assessment is contextually dependent and varies between school systems, individual schools and even individual teachers (Looney et al., 2018).Consequently, teacher assessment practices, such as designing, planning, performing and grading assessments, should be understood as dynamic, negotiated within each specific classroom context where teacher actions meet student learning (Deluca et al., 2019;Pastore & Andrade, 2019).Furthermore, teachers' assessment practices are influenced by their beliefs about the ways students can best demonstrate their knowledge and by the ideals of what constitutes learning in their subjects (Heritage, 2016;Nortvedt & Wiese, 2020;Ysenbaert et al., 2020).Thus, teachers' choices and views of assessment and their following assessment practices originate from the teachers' own perceptions or theories of learning (Black & Wiliam, 2018;Ysenbaert et al., 2020).In other words, they originate in teachers' understanding and perceptions of who their students are and how they think, learn and make meaning.Teachers' theories of learning also affect how they view alignment between teaching strategies and assessment formats and whether they regard them as separate or integrated (Ysenbaert et al., 2020).
To fulfil the aims of assessment for learning (AfL; Black & Wiliam, 2018) and ensure coherence and validity in assessments, the forms of assessment must mirror the teachers' teaching practices and curriculum goals.Validity here is understood as fairness in the interpretation and use of assessment data and the avoidance of biased inferences (Gipps & Stobart, 2009;Popham, 2009), and it is central for achieving equity.Moreover, formative assessment in the form of AfL necessitates a dialogic orientation to teaching and assessment, such as employing assessment strategies and practices that promote dialogues with students, to allow teachers to access the full breadth of their thinking and understanding (Black & Wiliam, 2018;Heritage, 2016).

Perceptions of classroom expectations and participation in CLD classrooms
Schools and classrooms are situated within specific cultural and social discourses.Thus, the expectations, values and forms of knowledge required of students and teachers differ within and between countries and even between schools (Herzog-Punzenberger et al., 2020).Likewise, students' expectations regarding what constitutes knowledge and learning and how to participate and interact in the classroom may differ.For example, students who have primarily been expected to listen to teachers and reproduce content will have different expectations than those who have been expected to contribute to classroom discourse through critical thinking and problem-solving (Cumming & Van der Kleij, 2016;Hodge & Cobb, 2016;Nortvedt et al., 2020).In other words, students may have prior experiences with either teacher-centred or E. Wiese and G.A. Nortvedt student-centred classrooms or with formative-or exam-oriented classrooms (Xu & Brown, 2016), which may be similar to or different from the dominant discourse in their receiving schools.These circumstances will influence their perceptions of assessment outcomes or emphases, that is, whether they focus on merits, ranking and accountability or on advancing individual students' learning (Herzog-Punzenberger et al., 2020).
In CLD classrooms, the heterogeneity in students' cultural backgrounds and linguistic competence levels place additional demands on teachers' assessment literacy, repertoire and skills for aligning, adapting and differentiating teaching and assessment, and particularly on ensuring the assessments are equitable and fair for all students (Nortvedt et al., 2020;Xu & Brown, 2016).Students' understandings and ways of ascribing meaning are culturally based and linguistically founded.Hence, in CLD classrooms, teachers must not only be attuned to students' individual ways of learning but also consider that students' ways of making meaning and thinking are informed by their "cultural script" (Cumming & Van der Kleij, 2016).This is evidenced not only in students' cultural and linguistic differences but also in their lack of "familiarity with particular task formats and contexts" (Cowie, 2012, p. 687).Cowie (2012, p. 687) argued that students' cultural values, beliefs, experiences and ways of communicating may influence their willingness and ability to engage in assessment situations.

Achieving equity through inclusive assessment in CLD classrooms
Inclusive assessment requires that all students have the opportunity to participate in assessment activities based on who they are, considering their frames of reference, knowledge and skills (Florian & Beaton, 2018).Consequently, teachers must adapt or differentiate assessments and make accommodations to address the different skills, abilities and ways of thinking of each student group.Thus, they must acknowledge how students' cultural backgrounds impact the way they learn, act and participate and understand how this affects their performance in assessment situations.This will ensure that students can demonstrate their thinking and lessen the impact that culturally based expectations may have on student learning.
Hence, equity and validity (fairness) are essential to realising inclusion in CLD classrooms.Gipps and Stobart (2009) stated that "equity does not imply equality of outcome and does not presume identical experiences for all" (p.106).As such, equity requires practices and actions that consider student differences, which demands that teaching and assessment practices be adapted to accommodate students' needs.Equity, therefore, is realised when students are treated differently and according to their own needs.A central tenet for achieving fair and equitable assessment in CLD classrooms is ensuring multicultural validity.Multicultural validity is achieved when students' cultural understanding and knowledge are considered in the design, performance and grading of assessments (Cowie, 2012;Stobart, 2005).Thus, multicultural validity can only be achieved by using multiple assessment formats that allow students to show varied competences across formats, thereby allowing them to demonstrate varying forms of understanding or to use alternative representations to exhibit knowledge and mastery, such as by replacing individual tests with problem-based and more collective or creative assessments (Herzog-Punzenberger et al., 2020).

Research context
Compulsory education in Norway includes primary and lower secondary school, grades 1-10.Norwegian schools are governed by a national curriculum.AfL and adapted teaching are key principles mandated in the national Education Act and in the regulations for the 1998 Education Act (Lovdata, 2006). 2 According to the act, students should understand how and what knowledge or skill is assessed, and they should receive feedback that supports their learning.The principle of adapted teaching ( §1-3) relates to all teaching and assessment activities, meaning all activities should be adapted to accommodate the needs and abilities of each student.
The national curriculum identifies overarching goals for student learning and competency goals for all subjects in compulsory and upper secondary education.While the national curriculum states what students should learn, teachers are free to choose how and when to address these competency goals and consequently to design their own assessments and choose the method of assessment.Thus, nationally, teachers are given considerable trust and autonomy in teaching and assessing students.For instance, teachers are responsible for developing assessment criteria and grading, and they set most of the students' grades.While AfL should be implemented continuously throughout all grade levels, students do not receive grades until grade 8.According to the 1998 Education Act, schools are only required to assess students with grades biannually, at the end of each semester.Accordingly, some schools practise "grade free assessments" that focus on feedback and dialogue.
Migrant students who speak a home language other than Norwegian and who enter lower secondary school in the second half of a school year can be exempted from grading with their parents' permission (Education Act, 1998;§3-21).Moreover, students in lower and upper secondary schools who are entitled to NSL classes can be exempted from grading while enrolled in the NSL course, according to the Education Act §2-8 or 3-12.Only students who have not sufficiently mastered the Norwegian language attend NSL classes (for a more in-depth explanation, see Thomassen & Munthe, 2020).

Methods
This paper presents outcomes from a case study in which five schools' experiences with assessment in CLD classrooms were examined.This study is part of the larger Aiding Culturally Responsive Assessment in Schools (ACRAS) project investigating the assessment of migrant students.For the current paper, we drew on teacher interviews to discuss teacher assessment literacy, emphasising inclusive and adapted assessment in CLD classrooms.The interviews allowed us to obtain a description of the teachers' practices but also, through the process of interviewing and the nature of the interviews, allowed teachers to reflect on and voice practices that were tacit and often taken for granted, thereby making the practices visible for the teachers as well as for us.

Sample
The sample consisted of 21 teachers representing five lower secondary schools in eastern Norway (Oslo and surrounding suburban area).This purposive sample was selected to ensure variety in terms of school size, school type (public/private 3 ), the proportion of migrant students and the assessment practices addressed.Volunteer teacher 2 The wording of the regulation for schools following the Independent

Schools
Act (Lovdata 2006 https://lovdata.no/dokument/SF/forskrift/2006-06-23-724/KAPITTEL_5#KAPITTEL_5) is similar to that of the regulation in the Education Act.Therefore, we only refer to the regulation for the Education Act in the following.
3 Most Norwegian students in compulsory education attend public schools.
Private schools mainly use alternative pedagogies (e.g.Steiner or Montessori).
Consequently, we included one private school in our sample to potentially widen teachers' assessment perspectives.
E. Wiese and G.A. Nortvedt participants, who were sought from each school, were interviewed after providing informed consent.Teachers could choose to be interviewed individually, in pairs or in small groups and were given opportunities to share and reflect on their experiences.Interviews typically lasted 45-60 min.The teachers represented various subjects (e.g.social and religious education, mathematics, sciences and languages), ensuring an array of assessment practices and experiences to discuss.The teachers interviewed also varied in gender, length of teaching experience, experiences teaching abroad, cultural and linguistic backgrounds and experience in teaching in CLD classrooms.Each teacher typically taught three subjects (e.g.mathematics, natural science and physical education) and typically taught 8th-10th grade students.As our study was focussed on 8th grade (13-year-old) 4 students, the participating teachers were asked to highlight their experiences with these particular students (Table 1).

Interview procedures
Semi-structured interviews were conducted following the ACRAS project interview guide, which included a national context and allowed for teachers' reflections on their experiences with adapted assessment and assessment in diverse classrooms.Interviews were conducted during school hours at the participants' schools from January to March 2018.The two authors, together with two colleagues (who each participated at two schools) and University of Oslo master's-level teacher education students, conducted all of the interviews.We used the same interview guide for all interviews; however, teachers in pair and group interviews were also invited to build on each other's reflections.The interview questions related to school policies, teachers' assessment strategies, their approach to adapting their teaching and assessments to accommodate the perceived needs of migrant students and the challenges to ensuring inclusion for these students (in assessment).In most cases, three interviewers conducted each session: one of the authors acted as the main interviewer, and the other two interviewers took notes to support the analysis, with the roles shifting toward the end of most interviews to allow the note-takers to ask follow-up questions.All interviews were audio recorded and transcribed verbatim by the master's students following agreed-upon transcription practices.

Analytical procedures
The interview transcripts were analysed using thematic analysis (Robson & McCartan, 2016;Xu & Zammit, 2020).Excerpts were categorised manually using an Excel file to gain a better overview.The two authors together performed several rounds of inductive analysis to identify themes, such as teachers' perceptions of the impact of diversity on assessments, concepts of fairness and validity, school policies and support structures.The deductive analysis performed by the first author in the second phase of the analysis produced additional categories, such as strategies teachers used to adapt their teaching and assessment to migrant students, teachers' understanding of diversity and the related challenges they experienced in the classroom, teachers' perceptions of using migrant students' background in the classroom and teachers' perceptions of AfL versus grading.The two authors discussed and agreed upon the themes and codes to ensure transparency and reliability in coding.
Next, the data were scrutinised to identify patterns that could describe aspects of teacher assessment literacy and how they affected teachers' assessment of migrant students.While the analysis predominantly focussed on individual teachers' experiences and practices, the case study format allowed us to see how collective practices manifested within each school, enabling a comparison across schools related to collective practices and perceptions of assessment, such as within-school variations in assessment practices and perceptions of assessment, teaching and adaption.
As this was a case study involving only five schools, some practices may be specific to certain schools, thus making them recognisable.To ensure anonymity of both the schools and individual teachers represented, interviewees are referred to as "all", "some" or "most", except when being quoted.Moreover, the five case study schools are referred to as schools A-E, and in that context, teacher A01 refers to teacher 1 in school A.

Limitations
Across the schools, the teachers mainly addressed feedback rather than grading.We used assessment, not grading, as a term in the interview questions, which may have contributed to teachers' focus on feedback.This paper builds on a case study, and as such, the data represent a contextual situated response, limiting replicability.Moreover, the teachers and schools were self-selected.By interviewing selfselected teachers who were likely familiar with more practices related to teaching in CLD classrooms or more favourably inclined to teach migrant students, the sample had a selection bias and was not representative of all teachers in Norwegian lower secondary schools.Instead, we selected schools that made it possible to identify a range of teaching practices aimed at including migrant students in assessments.The data, therefore, yielded rich information on teacher assessment literacy in CLD classrooms suitable for the current study.As noted by Thomassen and Munthe (2020), Norwegian teachers are uncomfortable addressing ethnicity because it is considered culturally inappropriate.Hence, linguistic awareness was more present in the research data.However, we argue that this may also serve as a signifier for describing cultural diversity.

Results
To answer the research questions and discuss the influences on teacher assessment literacy in CLD classrooms, we present findings related to 1) school assessment guidelines and policies, which create the contextual conditions that influence teacher actions and practices in the classroom; 2) teacher assessment practices; and 3) the ways teachers adapt assessments to accommodate students' cultural and linguistic backgrounds.

School assessment guidelines and policies
The teachers reported that their schools had not established formal assessment guidelines.However, they revealed some tacit but clear expectations that acted as guidelines or even policies within the schools during the interviews, including guidelines related to grade-free assessment or collective assessment practices, such as using a template for providing feedback to students.These guidelines were often initiated by school leaders and, to varying degrees across schools, shared and implemented by teachers.These guidelines or "policies" were not strictly directed toward migrant students.Rather, all of the schools and teachers maintained policies about adaptions directed toward all students, including migrant students.In addition, all the schools conveyed expectations about providing oral feedback to students.In some schools, teachers were merely expected to take time to talk to students about their performance and progress, while other schools established structures to dedicate time and space for providing oral feedback.For example, after term exams, school E teachers were exempted from teaching for one school day and, instead, held one-hour, one-on-one sessions with five students 5 to discuss and grade their term papers.Some of these "guidelines" may be considered inclusive because they were directed toward ensuring diversity and meeting the needs of all students, such as adapting assessment situations and policies regarding oral feedback.Schools also implemented policies that guided and promoted teacher cooperation, such as providing time for meetings to discuss and support teachers' assessment practices and interpretations of curriculum goals.The successful implementation of these practices seemed to be linked to their value being recognised by all teachers, demonstrated by structures that supported the collective implementation of initiatives and by the continuous support and facilitation of these practices by school leadership.

Teachers' classroom assessment formats and strategies
All of the interviewed teachers agreed that the aim of any assessment situation is to allow students to "show what they know", emphasising that their chosen assessment formats had to reflect this aim.When asked about which assessment formats they used, generally, they answered "written and oral assessment formats" 6 and mostly discussed formal assessment situations with summative aspects that included grading.However, when asked about their adaptions and teaching strategies, they mentioned a number of other assessment formats.Some reported moving away from using only tests as the basis for their assessments and grading and introducing alternative assessment formats instead, such as "literary talks", "subject discussions" (fagsamtale) and films or screencasts students created, which allow students to express more composite forms of knowledge.The introduction of one-to-one computers in several case study schools seemingly facilitated the expansion of the assessment formats the teachers could use and aided adaptions to teaching and assessments.Thus, the interviews revealed that the teachers used a variety of assessment formats, both oral and written, some formal and others informal, performed as a group, in pairs and individually.Formal assessments, such as tests, were used for both grading and for formative purposes.Informal assessments with, for example, process-oriented formats were primarily used formatively to support student learning and did not necessarily impact students' grades, indicating that formative and summative assessment are integrated in some classrooms but not in others, where they are, instead, seen as separate activities.
The participants described collaborating with colleagues to design tests and develop assessment criteria that enabled them to discuss various topicssuch as what signified low, medium and high achievementto ensure grades were assigned consistently across classes and teachers within the school.School A teachers also reported using tests and criteria developed by the local authority to ensure validity.All teachers used formal tests, especially at the end of the teaching period, which formed the basis for student grades.Some teachers preferred not to give students traditional tests or believed that such exams did not offer students the best way to display their knowledge, but they reported feeling that "you need to do it anyway" (A04) to ensure that school leaders, parents and students were satisfied.
Across schools, teachers sometimes disagreed on how to formulate test questions.In these instances, they would adapt the questions as appropriate for their own students.They also often disagreed on the degree to which certain test formats and questions allowed students to demonstrate reflection, critical thinking and problem-solving, which were considered key competencies across all subjects by all participants.Therefore, many advocated for open tasks that required students to reflect on written tests.However, they all indicated 8th grade students, regardless of background, struggled with reflection, critical thinking and problem-solving.Participant C02 discussed the dilemma of assessing reflections as they relate to students' thought processes: They [students] think that showing reflection is just saying a bit more about things, but in reality, it is about showing that you have understood and can put things into context.… They must sometimes feel like they are in deep waters and not know exactly what they want to say and let their heads work.Yes, I make criteria and goals, but if they experience it as a rubric for crossing off, then you take away their thought processes.
Teachers across schools preferred oral assessments for assessing reflections, such as discussions during which students explained content and reflected on the topic in their own words.All teachers highlighted the importance of modelling reflection in oral assessment situations, which they viewed as particularly helpful for migrant students because they often needed linguistic support to fully demonstrate their understanding.Furthermore, to ensure that students understood how to present their reflections, teachers modelled how texts should be composed, for instance, strategies for argumentative writing.This was done dialogically so students could suggest arguments or formulations and help shape the text.The teachers explained that formal written language is significantly different from the language used for oral expression, but many claimed that presenting reflections orally helped develop the ability to produce formal written reflections.

Feedback and grading
Teachers in Norway are legally obliged to grade students at least twice a year.However, the interviewed teachers identified potential negative effects of grading, particularly for migrant students.Participant D02.2 stated that grading can affect the trust between student and teacher because students take grades personally.Some teachers mentioned trying grade-free assessments for which students were only given (written) feedback.School B had established collective assessment criteria and templates for providing feedback to students, to which teachers mostly adhered.In schools C and D, leaders actively promoted grade-free assessment but allowed individual teachers decide whether to implement the practice.Moreover, teachers shared related challenges, such as students and their parents inquiring about grades because they were considered a more trustworthy (valid) and understandable measure of student learning and academic performance.
According to all of the teachers, feedback and AfL practices were more connected to their teaching practices and to student learning than grading.AfL is a national assessment policy included in the Education Act.By virtue of the language in the act, teachers are obliged to provide feedback to individual students.All teachers stressed their preference for providing feedback orally rather than in writing but said they often did not have the time or opportunity to do so.Many reported that they could more easily teach students to reflect if they could provide oral feedback to students and discuss their responses to open questions on written tests. 5They were selected by teachers to ensure that all students received this feedback during the three years of lower secondary school. 6Written tests were paper based and digital; oral tests included physical and digital presentations.

Adapting teaching and assessment to classroom diversity
Overall, the interviewed teachers seemed to have positive attitudes toward the contribution of student diversity to their teaching and assessment.Using students' backgrounds and directing teachers' teaching and assessment practices toward students' cultural orientations and linguistic capabilities are central to inclusive teaching and assessment (Milner, 2019;Nortvedt et al., 2020).All teacher participants seemed aware of this but described mixed experiences with implementation in their own teaching.Only one had formal training in teaching in diverse classrooms or teaching NSL.Therefore, the practices that the teachers applied in the classroom were primarily based on their own experiences, theories of learning and knowledge of adaption and formative assessment.All of the teachers recounted efforts to adapt their teaching and assessment practices to their migrant students, describing multiple ways to adapt assessment formats.These adaptions were based on various factors, with language skills needed for written exams highlighted by most teachers.Teachers saw communication (oral) as a key to learning and to gaining insight into student learning, especially for migrant students.They emphasised dialogue and mapping students' knowledge as the starting point for the adaption process.
A contested subject was the organisation, need for and use of remedial teaching for students with migrant backgrounds.Indeed, the teachers and schools differed on whether and how often students should be removed from class for remedial or extra language instruction.This practice was viewed as an obstacle to learning because those students miss the opportunity to follow and learn from the lesson in their regular class.However, teachers viewed it less negatively if the remedial class covered the same content as the regular class.
Notably, some teachers shared that the quality of instruction in the regular classroom suffered and became less rich when special needs and migrant students were not present to represent their perspectives.For these teachers, more diversity offered more perspectives in the classroom dialogue.Participants E01.1 and E01.3 observed, "We often think it is very useful for the students to stay in class, especially when we are discussing literature etc."They explained that classroom discussions benefitted students who were able to participate and those who only listened to the discussions: "Some of the students that are out in smaller groups are … non-native speakers, and they would really benefit from hearing the discussions in class because they [discussions] are about life".
All teachers described adapting assessment situations or formats (e.g.tests or presentations) to accommodate individual students or student groups based on their individual needs.Sometimes these adaptions and accommodations were made for the whole class because they were considered to benefit all students, not only those who needed them, and sometimes teachers made adaptions to address individual students' cultural or linguistic backgrounds.
The teachers stated that adaptions, such as offering linguistic support, were imperative for offering their students the best possible opportunities to participate in teaching and assessment activities.The interviewees identified two strategies for implementing whole-class adaptions to ensure that all students could demonstrate their understanding of the subject matter, at least in part: 1) supporting students' linguistic and conceptual understanding and 2) differentiating the content, level and form of representation.
The teachers in the five schools indicated that migrant students' conceptual understanding, not their overall language skills, posed the main obstacle to their learning and academic success.Conceptual understanding refers to the ability to both understand and apply academic concepts, which are often subject specific and contextual in nature, such as when the concept has multiple uses or culture-specific connotations or is similar to everyday concepts.The teachers noted the challenge was knowing how and when to use the concepts correctly.The primary obstacle to promoting students' ability to demonstrate their knowledge and learning, according to the teachers, was determining the extent of support needed by individual students.The teachers stressed that when students do not master everyday concepts, such as "compare", they cannot apply them in a subject-specific context, and they may underachieve or underrepresent their knowledge as a result.However, some teachers observed that many native-speaking students, particularly lowincome (low-SES) students, experience similar challenges.Thus, they saw an overall need for linguistic support, although they often described their schools as lacking in the shared strategies needed to investigate and address such needs.Accordingly, some teachers encouraged all their students to bring notes, concept sheets or "mind maps" to assessment situations.Sometimes teachers also encouraged their students to use digital dictionaries to search for concepts in their first languages to scaffold conceptual learning: … I always give them a "mind-map" that I have made beforehand, and they can fill them in to give the categories … [and] "concept maps" [so] that they always get a handout with the most important concepts, [and I give] them a PC so they can search for the concepts because they are often weak in conceptual understanding in their mother tongue.(D03) The teachers reported differentiating the content, level and form of representation by varying the difficulty level and types of tasks on assessments.For instance, tests may include both reproductive and reflective tasks, asking students to first define certain concepts and then to answer related questions that required them to explain their thought processes.Some teachers would also adapt the response format, such as by allowing students to respond orally to written assignments, a technique they said was frequently used.Other teachers allowed students to use drawings to represent content, which were evaluated according to how well they represented the topic and the degree to which they demonstrated understanding, not on their artistic merit.

Adapting teaching and assessment to students' cultural backgrounds
Teachers in all five schools expressed concerns about including students' cultural and religious backgrounds into their teaching due to their own lack of experience and their awareness of student vulnerability.Teachers with extensive experience in diverse classrooms reported that more diverse learning environments allowed them to include students' backgrounds into their teaching more easily than more homogenous classrooms.
Religion and some social science teachers proposed students' prior learning and knowledge about their home or other cultures could be a primary source of knowledge on important curriculum topics.They described efforts to accommodate students by enabling them to draw on or exhibit their knowledge in class, efforts that were sometimes hampered by students' unwillingness or inability to express their views and cultural backgrounds in class.They saw this challenge as related to representation and to migrant students' concern about being unjustly stereotyped.The teachers emphasised that students generally wanted to avoid being seen as different or being a "token" representative, preferring to be considered "ordinary Norwegian youth".
Teachers clarified that both judgement and awareness were necessary when deciding whether to introduce or include students' cultural backgrounds and religion into classroom situations.For example, sometimes teachers experienced that students were mistakenly seen as representing cultures, religions or linguistic groups, such as assuming students from the Middle East were Muslim.Teachers explained these decisions were difficult and were often made based on the relationship with and knowledge of the individual student.Some teachers, however, solved the conundrum of student representation by asking students in advance if they felt able, or were willing to, contribute.
If they are occupied with their own religion and think it is fun to tell others about it, then they are allowed.But not everyone is, and some think it is unpleasant when other students ask … either because they feel it is private or because they perceive themselves as different E. Wiese and G.A. Nortvedt when others ask if they know something which might be exciting for them.… You have to know them a little first.Not everyone is an expert in their own religion … but again, some students think it is alright to contribute and think it is fun to sort of be experts.(C02) Conversely, the mathematics teachers across the schools claimed not needing to consider students' cultural knowledge and backgrounds because they claimed that their subject is universal.Differences between students were mainly attributed to the teaching in their home countries (Nortvedt & Wiese, 2020).However, at the same time, the mathematics teachers still demonstrated awareness of students' linguistic differences and need for adaption.

Adapting teaching and assessment to students' linguistic abilities
The teachers stated that migrant students' linguistic competences varied, particularly because this composite group included some students who had grown up in Norway and others who had only recently arrived.Some students spoke Norwegian fluently, and teachers often described them as "ordinary students" who needed no adaptions outside the ordinary.Meanwhile, recent arrivals to Norway struggled with both understanding the language of instruction during lessons and with expressing their subject-related prior knowledge.Language was seen as both a bridge and a barrier to student knowledge and learning, but not necessarily by the same teachers or in the same situations.Some teachers believed language difficulties could be bridged, such as when a teacher speaks certain migrant students' home language and could act as a translator so the students could take tests in their home language, or when teaching material existed in multiple languages so students could refer to those in their native language.Additionally, Google Translate and other translation tools were seen as a way to bridge migrant students' learning.The teachers indicated that, as another adaption, they often allowed students who needed accommodations to take written tests orally because students new to the Norwegian language often spend too much time trying to form readable sentences when they are required to complete written tests.In addition, they reported helping students with prompts.Alternatively, students completed the written test and afterwards expanded on their answers orally.A third strategy the teachers applied was assisting students in understanding test questions by reading them aloud and rephrasing or explaining the content.Participant A04 described doing this: I help the student [in written tests in math] and explain what needs to be done because I think it is better that he knows what to do with a little help than not being able to understand [at all].
The teachers also expressed strong positive beliefs about the possibilities ICT tools offered for language supportsuch as those that orally reproduce written content by, for example, reading textbook content out loudwhich promote non-native-speaking students' language learning and understanding and facilitate adaption to students' linguistic needs.The teachers emphasised that such tools can be useful for teaching and assessments.
We now have 'smartbooks' and soundthat is, textbooks that allow them to have something read aloud and have some resources … that come with PCs for every student.… In 8th grade, for example, they have [name of textbook] that can be read out to them, and they can hear at the same time as reading, and I think that is a great resource for minority students and students who have weak language skills.(D01.1)

Analysis: what impacts teachers' possibilities for achieving inclusive assessment in CLD classrooms?
Table 2 illustrates the complexity teachers face in realising inclusive assessment in CLD classrooms, identifying possibilities for realising inclusive assessment in these classrooms at three levels (school, teacher and classroom/student) and the associated tensions derived from our findings.The depicted tensions can threaten teachers' ability to realise inclusive assessment, or, as we argue, if dealt with professionally, these tensions and potentials can aid in further developing teacher assessment literacy in CLD classrooms.
As mentioned, inclusive assessment entails designing and performing assessments that give all students the possibility to participate, while considering who they are as well as their cultural and linguistic frames of reference, knowledge and skills (Florian & Beaton, 2018).We observed that some structures at the school and teacher level were more likely to promote inclusive assessment practices among teachers.At the school level, shared and collectively understood practices informed by an agreed-upon overarching pedagogy facilitated the development of inclusive assessment practices by requiring teachers to continuously co-develop, argue for and describe their practices.The teachers reported that sharing similar ideals and practices with other teachers within the same school gave them opportunities to implement shared practices to better accommodate the assessment of migrant students.For example, shared criteria and adaptions agreed upon by faculty teaching the same class can help students better understand the assessment demands.Moreover, familiarising students with a variety of assessment formats and criteria in multiple subjects can help them develop key competences, such as reflection and critical thinking across subjects.The teachers in the schools in our study that had established structures and collective practices more often saw opportunities to align their teaching and assessment, potentially facilitating more equity and fairness than teachers in the other schools.Such structures included allocating time for teachers in the same discipline or for staff across the school to collectively develop assessment forms, criteria and templates that support the alignment of teaching and assessment practices, thereby also supporting teachers' assessment literacy development.Shared school structures enhance the potential for cooperation and agreement between teachers, yet as our study confirmed, some teachers may perceive such guidelines as threatening their individual autonomy, potentially conflicting with their perceptions of what constitutes good teaching and assessment.Furthermore, while some teachers reported working closely with colleagues to design teaching and assessment situations, the responsibility for adaptions often fell to the individual teacher.Thus, these adaptions did not, as intended, work as collective practices agreed upon and enacted by the whole school.Instead, they represented ad hoc approaches taken in response to perceived difficulties students face "here and how", potentially posing a threat to realising equity.Moreover, the extensive emphasis on formal assessment by some teachers and schools in the form of end-of-period tests and grading may weaken the alignment between teaching and assessment.In sum, we found that building structures and collectively agreed-upon practices are necessary for supporting the enactment and development of assessment literacy in CLD classrooms.
Across the schools, the teachers' reflections revealed that realising alignment and ensuring that students are familiar with the assessment formats are particularly challenging with migrant students, as they may not be familiar with the often culturally laden ways the assessments are performed.Still, we argue that the shared emphasis on critical thinking and reflection as key competences across all subjects may allow for alignment both within and across subjects and teachers within the school.However, we also found that teachers' perceived (lack of) time was a constraint to realising this alignment.Indeed, the possibility of achieving inclusive, culturally valid and equitable assessments was strengthened by allocating time in teachers' schedules to collaborate on (aligning) teaching, assessments and criteria, to share experiences and co-reflect on student issues and to provide students with oral feedback.However, when teachers did not have such opportunities, collective practices were implemented less often, and teaching and assessment became more individualised, thereby also reducing opportunities to further develop their assessment literacy.

Discussion
The aim of this study was to add to the knowledge on what inclusive assessment in diverse classrooms entails and to highlight the related assessment literacy that teachers need in order to fairly assess students with migrant backgrounds in increasingly CLD classrooms.Assessment is a social activity embedded in the context of the school and the classroom (Xu & Brown, 2016).Both assessment and teaching are complex, composite skills that become even more so in the context of diversity.In CLD classrooms, teaching and assessment practices need to allow for diversity in students' individual "scripts", knowledge and experiences (Cumming & Van der Kleij, 2016).This places great demands on teachers' knowledge and assessment literacy as they strive to realise inclusive assessment.As noted, the Norwegian context of adapted education, inclusive classrooms and AfL is fruitful for investigating inclusive assessment and for providing examples of how inclusive assessment can be implemented internationally.
Initially, we suggested that inclusive assessment (Florian & Beaton, 2018) entails adaption and catering for participation in assessments (Cowie, 2013) to give all students, regardless of their background and linguistic ability, the opportunity to learn and make their knowledge visible.Furthermore, we highlighted the need for alignment (Gipps & Stobart, 2009;Ysenbaert et al., 2020) between teaching and assessment to ensure inclusive assessment in CLD classrooms.Thus, the ability to dialogically establish and uphold norms for student participation is a key component of inclusive assessment in CLD classrooms.In particular, these norms can establish a pattern for classroom dialogue in which a multitude of students' voices can be heard (Black & Wiliam, 2018;Cumming & Van der Kleij, 2016;Hodge & Cobb, 2016), thereby also better positioning teachers to formatively assess student knowledge and understanding, as highlighted by some of the teachers in our sample.Thus, teachers' beliefs about students' competency and how they understand and interpret students' often culturally and linguistically situated expressions influence assessment validity and whether it can be achieved and thereby either promote inclusion and equity or impede student learning and development.Therefore, teachers must understand not only how to achieve equitable assessments but also how diversity manifests in individual student actions and identities (Cowie, 2013;Cumming & Van der Kleij, 2016;Gipps & Stobart, 2009).
The participant teachers reported that some migrant students had different expectations of what learning, participation and knowledge entails that affected how they demonstrated knowledge in the classroom.In CLD classrooms, teachers must therefore ensure that the inferences they draw through assessments actually represent their students' achievements (Gipps & Stobart, 2009;Xu & Brown, 2016) and that they are not the result of, for example, unfair access to the curriculum that (unwittingly) prioritises some student groups over others (Nortvedt & Wiese, 2020).This can happen when the ways in which language skills, cultural biases or cultural ways of understanding and representing knowledge that influence students' possibilities for (fair) participation in teaching and assessment situations are not considered (Cowie, 2013).However, several teachers in our sample emphasised that this situation could be mitigated by informing and familiarising students with expectations related to formal and informal assessment formats and ways to present content in assessments (e.g.reflections and critical thinking).All of the participants expressed awareness of and insights on how students developed critical thinking over time during lower secondary school.Such awareness allowed the teachers to incorporate a developmental trait into their assessment practices and better enabled them to realise inclusion.
Moreover, as exemplified in the analysis above and suggested by Ysenbaert et al. (2020), collectively agreed-upon and enacted practices at the school level can enable collective discussions and shared understandings of assessment and assessment practices, thus supporting teachers in performing inclusive assessments.These discussions can facilitate the alignment of teaching and assessment, allowing teachers to draw on and exchange diverse understandings and knowledge within the teacher community.This includes intercultural competences and an understanding of the knowledge of culturally and linguistically diverse students.Thus, these discussions can support teachers in developing their ability to adapt assessments to accommodate migrant students' needs.In our study, only one teacher reported receiving training specifically on cultural and linguistic diversity in the classroom, in NSL instruction or in teaching and assessing migrant students.This lack of training could result in a tendency toward homogenisation rather than focussing on the complexity and heterogeneity in expression exhibited by culturally and linguistically diverse students.This was illustrated by teachers who described migrant students who had mastered the Norwegian language well enough to follow lessons and participate in regular assessments as "ordinary students".Indeed, this focus on linguistic homogeneity, which can threaten the validity of assessments and obstruct possibilities for creating inclusive learning and assessments, was particularly obvious in our sample.This might be a consequence of the culturally embedded practice of not mentioning ethnicity because it is considered inappropriate, leading teachers to focus instead on students' individual identities and needs (Thomassen & Munthe, 2020).
Our findings suggest that the two principles of AfL and adaption, which are implemented in the Norwegian Education Act (1998), often act together, and where they are used according to the intentions of the law, they strengthen one another and hold the possibility to work towards more inclusive teaching and assessment practices by giving teachers and schools autonomy and space to develop and implement locally based curricula and practices specifically designed to meet their particular students' needs.However, this autonomy places stringent demands on teacher assessment literacy, demanding considerable knowledge and professionalism on the part of individual teachers in developing suitable forms of assessment and adaptions that are inclusive and culturally responsive and directed at encompassing the breadth of student diversity in the school.
As underlined by Xu and Brown (2016), to be culturally responsive, E. Wiese and G.A. Nortvedt assessment literacy in CLD classrooms needs to be conceptualised in the plural form, as assessment literacies, implying that teacher assessment literacy relates to both general and contextualised assessment practices.
Our study revealed that assessment literacy (Deluca et al., 2019;Pastore & Andrade, 2019;Xu & Brown, 2016) in contemporary classrooms needs to encompass inclusion-related competencies (Cowie, 2013;Florian & Beaton, 2018), such as knowledge related to realising equity and (multicultural) validity in assessments (Gipps & Stobart, 2009).As we have found in this study, in line with Abdulrahim and Orosco (2019), teachers often have strategies related to multilingualism in the classroom, yet they are less aware of the impact cultural differences and diversity have on student learning and participation.To accommodate students' linguistic skills, the teachers highlighted strategies and practices to ensure that both grading and formative assessments are valid and equitable, including adapting the assessments (e.g.giving students concept maps and allowing the use of dictionaries and digital resources) and making accommodations (e.g.allowing oral instead of written tests).Furthermore, the teachers emphasised a dialogic orientation toward teaching and assessment to enhance assessment validity, for instance, by preferring oral feedback to written feedback to ensure student understanding.Even so, in our sample the teachers expressed insecurities and described a gap in their competency related to including students' cultural background in assessments.For example, they had concerns about how to avoid unjust stereotyping and "tokenisation" of migrant students; how to avoid unfair, unjust and inequitable assessment practices and inferences; how to adapt assessments fairly and equitably in CLD classrooms; and how to increase participation and accommodate differences in students' culturally and linguistically based identity and expressions.Yet, while several teachers were aware of these dilemmas, our study indicated that they still lack strategies for including students' cultural backgrounds in assessment situations.Thus, further development of (inter)cultural awareness and studies related to understanding how students' "cultural scripts" can be used in assessment situations to realise inclusive assessment and accommodate the diversity in CLD classrooms are still needed.

Conclusion
Building on and expanding the literature on assessment literacy (Deluca et al., 2019;Pastore & Andrade, 2019;Xu & Brown, 2016), we argue that to achieve inclusive assessment of migrant students and meet the diversity present in CLD classrooms, teachers must develop a dynamic assessment literacy that includes a broad repertoire of inclusive teaching and assessment formats that can be adapted to accommodate the breadth of student learning and mastery in the classroom as well as their cultural and linguistic backgrounds.Overall, our findings indicate that teacher assessment literacy in diverse classrooms is developed at the intersection of equity and validity, that is, by providing assessments that are equitable, valid and fair, considering the students' cultural and linguistic backgrounds and are adapted to their individual needs.For teachers and schools, this implies catering to and developing initiatives, strategies and practices that ensure inclusive assessment for all.Furthermore, our findings indicate that both initial and further teacher education should focus not only on technical aspects of adaption and assessment but also on how teachers can address cultural aspects of participation to achieve valid and equitable assessments in CLD classrooms.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Table 1
Teachers and case schools.
4 13-year-old students were selected because they attend secondary school in all four ACRAS countries.E.Wiese and G.A. Nortvedt

Table 2
Factors influencing teacher assessment literacy in diverse classrooms.