Teacher Beliefs and Practices of Language Assessment in the Context of English as a Lingua Franca (ELF): Insights from a CPD Course

Abstract

English as a lingua franca (ELF) has challenged English language teaching and learning and has recently impacted teacher education as well as teacher beliefs and practices. Nevertheless, there are very few studies exploring teachers’ beliefs and practices towards ELF-aware assessment. Within the framework of an online ERASMUS+ course, the current study investigated the perceptions of English language teachers towards assessment within two multilingual contexts, Norway and Italy. The data were collected through the assessment course activities and final assignments. The findings showed that most teachers were in agreement with balancing ELF-aware teaching and assessment but struggled to address formative and summative assessment mandates in their multilingual contexts. The evolving classroom context that characterizes multilingual settings demands teachers to develop flexibility and innovation as part of their language assessment literacy trajectories. This study has several implications for language teachers, teacher trainers, and future researchers.

1. Introduction

The ‘multilingual turn’ in education (Council of Europe 2007; King 2017; Lüdi 2021), emphasizes, among other things, the value of multilingual spaces. The Council of Europe spearheaded the differentiation of plurilingualism and multilingualism supporting that multilingualism acknowledges the multiple languages that function in a society or in a learner’s environment. Accordingly, the European Center for Modern Languages (ECML) was commissioned to develop actions in support of multilingual classrooms, notably the plurilingual competence framework (Dendrinos 2019). Plurilingualism recognizes and supports the whole person’s linguistic (Bakhtin 1981; Gumperz 1964; Gumperz 1982; Cook 1991; Cook 1992) and experiential landscapes (Busch 2015; Bahry 2021; Blommaert 2009), taking into account the totality of linguistic knowledge (Wandruszka 1979). It also sheds light on how multiple languages are used and interact with each other “to create something unique” (Piccardo et al. 2021, p. 1). Finally, plurilingualism acknowledges the individual’s experiences developing over time and spaces (Blommaert 2009). In this study we will focus on multilingual aspects of classroom interaction, in the area of inquiry known as ‘English as a Lingua Franca’.

English as a Lingua Franca (ELF) is “a prime example of multilingual communication” (Leung and Jenkins 2020, p. 27) in that it exists in places where multiple languages are used and reflects plurilingualism of an individual’s own idiolect (Otheguy et al. 2015) or holistic cartography, which cannot be separated from affect or one’s emotional response (Kramsch 2009).

ELF was first conceptualized as a lingua franca of interlocutors who do not speak English as their first language (Seidlhofer 2011). It uses language forms and word choice that are divergent, context dependent, and non-fixed (Canagarajah 2006; Leung 2013; Seidlhofer 2004). As a mediation practice, speakers do not restrict themselves to one language or to a standard of the language (Harding and McNamara 2017; Leung and Jenkins 2020). The baseline, if you can even create one, is not based on an adult native speaker from either language A or language B, but on an “effective multilingual who is able both to adjust the languages in their repertoire and to translanguage as appropriate in the moment” (Leung and Jenkins 2020, p. 32). ELF is concerned with how English is used outside the classroom and focuses on users, not learners (Newbold 2015).

The primary role of the speakers’ other languages has come to the forefront (Leung and Jenkins 2020), renaming English as a multilingual franca and pointing out that English is just a language choice among many (Jenkins 2015). In educational contexts, most teachers in the outer Kachruvian circles are ELF speakers themselves, supported by local contexts and social media which are widely ELF, with inner-circle NS (native speaker) English in classroom settings taught through textbooks and native recordings, which does not reflect the whole of students’ or indeed teachers’ experience with English in real life. Finally, ELF is primarily oral and interactional (Kouvdou and Tsagari 2018) characterized by communicative strategies reflecting the multilingual communication skills of its users (Leung 2013). On this basis two ELF corpora, e.g., VOICE (VOICE 2021) and ELFA (ELFA 2008), have been created which are exclusively made up of spoken data.

On the other hand, in language testing and assessment, standard English has been the dominant variety of English taught in the classroom and the benchmark by which English is typically assessed and tested (Harding and McNamara 2017; Jenkins 2020). Therefore, deviations from inner circle norms are viewed as (fossilized) errors based on error analysis (Corder 1967) and interlanguage studies (Selinker 1972), even though in linguistics, they are naturally occurring phenomena and an essential part of language learning and development (Newbold 2018). In ELF, these forms are an unfixed characterization of ELF (Seidlhofer 2011) that usually do not interfere with understanding of meaning (Newbold 2015). Errors would only be those elements which interfere with successful communication (Kouvdou and Tsagari 2018; Newbold 2018) while intelligibility should be viewed as more important than correctness (Seidlhofer 2004).

ELF-aware assessment practice should, therefore, include communicative (Elder and Davies 2006), purposeful tasks (Brown 2014; Seed 2020) that are authentic in order to avoid negative washback (Harding and McNamara 2017; Tsagari and Cheng 2017). Such assessment focuses on an interactional component (Brown 2014; Jenkins and Leung 2013; Seidlhofer 2004) that awards hedging and self-repair (Mauranen 2003). Candidates are also awarded for negotiating meaning (Jenkins et al. 2011), showing an ability to adjust to interlocutor’s style (Seidlhofer 2011), and demonstrate proficiency in pragmatics over grammar (Canagarajah 2006) including adaptability, sociolinguistic skills, style shifting, interpersonal communication, and conversation management. Discourse strategies such as patterns of organization and cohesive and transitional devices are also taken into account (Canagarajah 2006; Leung 2013). It is also argued, that for as long as non-native speaker (NNS) accents do not impede understanding they should be included in listening comprehension assessment (Newbold 2015). Finally, interlocutors may have different levels of proficiency, reflecting authentic interaction while scoring should weigh “the upper limit of languaging potential shared by participants” (Newbold 2018) and be awarded jointly (Chalhoub-Deville 2019).

Assessment of strategic competence and communicative effectiveness has been promoted in multilingual, and ELF assessment fields (Council of Europe 2020; Elder and Harding 2008; Gandini 2020; Harding and McNamara 2017; Leung and Jenkins 2020). Various multilingual testing and assessment projects are grounded in language contact theories (Fiorenza and Diego-Hernandez 2020; Gandini 2020; Schissel et al. 2019a; Sifakis et al. 2020) especially in mediation, oral interaction (Kouvdou and Tsagari 2018) and translanguaging (Dendrinos 2019; Leung and Jenkins 2020).

Researchers lament that the field of language testing and assessment has been slower to reflect the findings of ELF researchers in the field of multilingualism (Elder and Davies 2006; Gorter and Cenoz 2017; Harding and McNamara 2017; Jenkins 2020; Newbold 2015; Shohamy 2011), which has undermined multilingual practices (Chalhoub-Deville 2019; Flognfeldt et al. 2020; Saville 2019; Schissel et al. 2019b). As universities across Europe increase their offers of English medium instruction (EMI) with ELF as the medium of communication, “it is now urgent to engage with assessing ELF” (Newbold 2015, p. 32).

Motivated by the calls above, the current study seeks to explore the perceptions and practices of English language teachers towards assessment in multilingual contexts in a comparative study undertaken in Norway and Italy aiming at contributing to the body of research in the field. The study aimed at addressing the following research questions:

• RQ1: How do teachers understand ELF-aware assessment?
• RQ2: Do teachers use ELF-aware assessment in their classes?
• RQ3: If so, what methods and approaches do they use? If not, why not?
• RQ4: What kind of contextual needs do teachers address when they practice ELF-aware assessment?

The goal was to understand whether English language teachers are equipped with sufficient awareness and competence that allows them to organize their language assessment in a way that caters for the needs of multilingual learners. The insights to be gathered in this study aim at informing teacher education about areas that need a more careful consideration in the educational development of multilingual language students.

The following section will present the two educational contexts of Norway and Italy where the study has been undertaken and will report on the multilingual assessment landscapes therewithin.

1.1. Norway

The total population in Norway is 5.4 million of which 800,000 (14.8%) are migrants. The total number of students who are migrants or born to migrant parents in schools are 86,689 or 19.5% of the population in elementary schools, 33,690 or 17.6% of the population in lower secondary, and 50,694 or 20.5% of the population in upper secondary schools (Statistics Norway 2021). Another 10,026 or 19% of the students in preschools, aged 1–5, receive intensive Norwegian lessons. Foreign languages are taught for 6 years beginning in lower secondary. There are two official languages: Norwegian and Sami, and two written standards: Nynorsk and Bokmål. The right to an education in sign language is also guaranteed. There is a wide variety of dialects, all of which are welcome and widely used in the general society.

No grades are awarded in the primary grades 1–7 in Norway. English is taught for 11 years. English education is compulsory from first grade through 11th (first year at upper secondary). Teachers have a great deal of autonomy in how teaching is organized as it is considered part of a teacher’s professional judgement (Karseth and Sivesind 2010). Student achievement is based on local interpretation and strong alignment with the national curriculum aims, which are loosely based on CEFR levels (Council of Europe 2001). National exams are taken in the autumn, in grades 5, 8, and 9 and cover math, reading, and English. In upper secondary school, there are 5–6 exams, of which 3–4 are externally-set written exams and 2–3 are locally-set oral or practical examinations. Teachers are encouraged to use the results of the exams to inform their learning plans. There are no high stakes tests for university admissions.

Multilingualism in and around English is a feature in the latest curriculum guidelines (Norwegian Directorate for Education and Training 2011). Teachers should be able to plan, conduct, and assess teaching in multicultural learning environments. Students must be able to identify similarities and differences between English and their own mother tongue by 10th grade and use these differences to help with their language learning. Students are expected to be able to comment on their own work and learning, understand words and expressions from a variety of types of English, and use “central patterns for pronunciation”, whatever those central patterns may be. Digital media is designated as a resource for authentic language use and cultural context while languages are recognized as a resource in schools and society. Language diversity is promoted in the development of individual identity while at the same time common societal values are taught.

Assessment became a national priority for professional development in 2006, when it was found that assessment literacy of teachers and teacher educators were poor (Ministry of Education 2006–2007) and teachers needed to improve their formative assessment practices (Looney et al. 2008; OECD 2011; Tveit 2014). On the basis of this, a national project, the Improved Assessment Practices, was implemented, placing emphasis on assessment for learning (AfL) and introducing new policies for supporting student learning processes. Recent research has also shown that overall, teachers have somewhat low levels of language assessment literacy and report low confidence levels in using formative assessment practices, e.g., self-assessment, peer-assessment, working with the CEFR, etc. (Tsagari 2021; Bøhn and Tsagari 2021; Vogt and Tsagari 2014). Finally, teachers often lack the training and competence to implement multilingual assessment practices (Flognfeldt et al. 2020), and sufficient practice in multicultural settings (Thomassen and Munthe 2020).

1.2. Italy

Italy has 60 million inhabitants, 5.4 million of which are non-Italian. The country is a fast-growing multilingual country because of its language minorities, bilingual and trilingual regions, dialects spoken all over the country, and its widening multilingual migrant population. In the last decade, 36.5% of students with a non-Italian citizenship (NIC) are at primary schools. Romania, Albania, Morocco, and China are the largest minority groups (MIUR 2015; ISMU Foundation 2021). The growth of the multilingual population in Italian primary schools, and pre-primary schools, has posed new challenges to teachers (MIUR 2018; Candelier 2003; Lopriore 2018a; MIUR 2014).

Schools start with Early Childhood education (3 to 6), followed by Primary Education (6 to 11), then by middle school (11 to 14). At the end of middle school, students have a final ‘school based’ exam. Students can then choose a 5 year-high school with ‘national’ exams. Foreign Languages (FLs) are primarily English, taught for 13 years, and one of the other European languages, e.g., French, German, Spanish. Recently there has been a growing number of other languages offered in Italian schools: Chinese, Russian and Arabic, taught for 8 years. The CEFR (2001) is used by the Ministry of Education to indicate FL school exit levels (Middle school: A2, secondary school: B2). Private FL certification is used in secondary schools. At university level, English is taught in all courses while several European FLs and Chinese, Arabic and Russian are offered in most FL degree courses.

The Italian Ministry of Education has developed specific guidelines to sustain multilingual language education and recommended that teachers developed specific teaching skills for teaching Italian, the language of schooling, to migrant learners (MIUR 2015). As a consequence, several initiatives were taken to organize training courses at local school level to face the challenges of the integration of migrant students in the Italian classrooms.

In Italy, teachers learned about assessment mostly as a theory and in terms of subject specific (e.g., foreign languages vs History or Maths) traditions. This model was used in the national teacher education courses run between 1999 and 2009 (Gallina 2021; Quartapelle 2013; Lopriore 2018b; Lopriore 2021). At primary level no grades are awarded. Descriptive judgements of learners’ achievement of each subject area objectives have been used since the 2017 ministerial decree on evaluation (MIUR, DM 62/2017) where evaluation is highlighted as a way to promote learners’ identity development as well as learner’s self-assessment. Foreign language teachers have learnt about assessment through coursebooks, international certifications, the use of the CEFR, the Ministry national guidelines (MIUR 2012, 2017, 2018) and individual school evaluation plans. Recently, the National Institute for the Evaluation of the Education and Training System (INVALSI), under the supervision of the Ministry of Education, promoted the improvement of educational levels through the development of assessment tools to measure students’ learning outcomes and the quality assessment of schools. National standardized tests for grades 2, 5, 8, 10 and 13 are one of the three strands of INVALSI institutional commitment in all the tested subjects (Italian, Mathematics, English Reading and Listening). INVALSI offers training seminars for teachers involved in the yearly test administration and recently devoted studies on the assessment of multilingual Italian school population (Leggi et al. 2020). Italian teachers, as well as students’ parents, at first quite critical of the INVALSI tests, are now modifying their responses thanks to informative workshops (https://invalsi-areaprove.cineca.it/index.php?get=static&pag=home_tf, accessed on 22 February 2022).

2. Materials and Methods

2.1. The Research Context

The data of the current study was collected within the context of an Erasmus+ project titled ‘English as a Lingua Franca Practices for Inclusive Multilingual Classrooms (ENRICH 2018–2021). The project consisted of a network of researchers from Greece, Italy, Norway, Portugal and Turkey. The purpose of the project was to help English language teachers integrate languages of international communication in multilingual classrooms to help their learners, particularly learners from migrant backgrounds, and develop skills crucial for social inclusion and employability in the current globalized world. ENRICH designed and offered a professional development course (CPD) which consisted of 30 online modules based on multi-level, cross-country exploration of teachers’ and learners’ needs in multilingual classrooms. Each module included a video lecture, a range of activities, and other useful multimodal materials and resources. At the end of the piloting phase, participants designed, implemented and reflected on an ELF-aware lesson plan for their multilingual classrooms as part of their final assignment (see enrichproject.eu/lessons/180-5-final-assignment, accessed on 22 February 2022).

The language assessment module was to be taken at the end of the CPD course and so teachers had already considered the implications and various other aspects of ELF. The assessment module focused on ways in which issues related to ELF can be integrated in assessment practices. In other words, emphasis was placed on how language assessment in the classroom can be culturally and linguistically sensitive and flexible by adopting an ELF-aware perspective. To this end, different assessment orientations were described, e.g., summative and formative assessment. This section included four activities in total (see also Appendix A):

Activity 1: a reflection task raising awareness over aspects which can potentially influence ELF-aware assessment.

Activity 2: a reflection task on what informs teachers with regard to their own assessment practices, test washback, and how tests influence learning.

Activity 3: a reflection task on how textbooks treat ELF-aware assessment practices.

Activity 4: a reflection task on how a set of accommodation strategies based on oral performance could be used as a part of an ELF-aware teaching and assessment ecosystem of learning.

2.2. Participants

English language teachers from elementary, lower and upper secondary levels and adult education participated in the study: nine teachers from Norway and twenty-one teachers from Italy. Teachers were recruited through calls for participation via social media (FB, Instagram, etc.), professional teachers’ organizations and the researchers’ personal networks. For teachers’ experience see

Table 1. Teachers’ experience.

Level	Countries	N
Pre-primary (3–5 years old)	Italy	1
	Norway	-
Primary (6–10 years old)	Italy	6
	Norway	-
Middle school (11–13 years old)	Italy	7
	Norway	3
High school (14–19 years old)	Italy	3
	Norway	3
Vocational/Technical school (14–19 years old)	Italy	2
	Norway	1
University (20 years old)	Italy	2
	Norway	-
Preparatory Adult Education (Adult students)	Italy	-
	Norway	2

.

At the beginning of the course, all participants were informed of the aims of the project and signed a consent form to permit the current researchers to use their data anonymously.

2.3. Analysis

Teachers reflected on assessment in the ELF classroom through the assessment activities (see Appendix A). All teachers’ responses (quantitative and qualitative) to the four activities of the assessment course and the final assignments were collected, processed and analyzed by the researchers.

The closed items of some of the activities were presented to participants in either 5- or 6-point Likert scales. The activity items were informed by the suggestions made in the literature and the Needs Analysis conducted at preliminary stages of the ENRICH project (Lopriore 2021). The statistical analysis was based on descriptive statistics including frequencies and percentages.

Teachers’ reflective comments to the assessment activities and final assignments were analyzed using thematic analysis (Braun and Clarke 2012; Hsieh and Shannon 2005). The analysis was carried out in three stages. In the first stage, the transcripts were read through to get an overall impression of the material. In the second stage, reading the transcripts allowed themes to emerge and preliminary categories were created using in vivo coding (Miles et al. 2014). In the third stage, comparisons were made across informants from Norway and Italy to arrive at an overall description of the construct of ELF-aware assessment. In this phase, some of the in vivo codes were replaced by researcher-generated codes to achieve a more coherent description of the themes identified. The analysis of the data was undertaken separately by the researchers to enhance cross-verification of data by way of establishing investigator triangulation.

The participants’ identities have been anonymized, and they will be referred to as “N01”, “I01”, etc. (which are code names for “Norwegian Teacher 1”, “Italian Teacher 1”, and so on). When ‘A’ was added to the teachers’ code names this indicates that extracts come from the final assignments, e.g., ‘NA01’ or ‘IA01’. All teachers will be referred to with the pronouns ‘she’, ‘her’ and ‘hers’ to avoid identification.

3. Results

3.1. Teachers’ Views on Assessment

Teachers were first asked to rate a number of statements about their views on assessment (see Activity 1, Appendix A) on the basis of a 6-point scale (see

Table 2. Teachers’ views on assessment.

Activity 1, Statements	Countries	1 = Strongly Disagree	2 = Disagree	3 = Neither Agree Nor Disagree	4 = Agree	5 = Strongly Agree
Pronunciation 1. Learners should sound like native speakers when speaking.	Norway	8	0	1	0	0
	Italy	15	4	2	0	0
Tests and Grades 2. Tests and grades are important to evaluate success in the classroom.	Norway	0	5	2	2	0
	Italy	6	8	4	3	0
Feedback 3. Written/oral feedback is an important part of the learning process.	Norway	0	0	0	2	7
	Italy	0	1	0	4	16
Self-assessment 4. My learners use self-assessment in language tasks.	Norway	0	0	2	7	0
	Italy	0	3	5	8	5
Communication 5. The primary goal of speaking tasks is communication.	Norway	0	0	0	2	5
	Italy	0	0	0	9	12

) and then provide reasons to explain their views.

3.1.1. Pronunciation

Teachers reported strongly that they did not think that learners should sound like native speakers (see teachers’ responses for Statement 1, Table 2). However, three teachers were neutral. We believe that neutrality “might” hide a reflective process that is taking place, possibly leading to change of attitudes or positioning that needs time to be fully elaborated, thus allowing teachers to express a clear positioning, i.e., agree or disagree. Changes like the ones brought in by ELF might be profound and neutrality might mean (positive) hesitation and the need to gain more experience and confidence.

In their comments and assignments, teachers further reflected on native versus non-native speaker pronunciation. They clearly consider native speaker pronunciation ‘not a priority’. In the following instance, teacher NA05 cautions that though students do not need to sound like a native speaker, they must still be aware of their own pronunciation as well as features that can interfere with communication.

… the intention is not to sound like a native speaker but to be aware of the issue of how features of their own accent, shaped by their mother tongues, could cause difficulty when speaking English.

(NA05)

Teachers regard pronunciation as an important aspect of their practice. For example, teacher I04 underlines the importance of being able to communicate but adds an important reference to the context requirements and to the need to implement the native-speaker standards adopted within learning institutions.

It depends on the situation: if the learner attends high school/university and has to take an exam, standard pronunciation may be important, otherwise what’s important is the ability to communicate and interact.

(I04)

However, the above attitudes can create conflict for teachers as to how to handle the issue of Standard English and language ownership. Is English British, or American, or Indian?

The most effective thing that I have learned is that English belongs to everyone, not only to native speakers.

(IA12)

[…] the majority of the students reported that […], they were exposed to varieties of English and made them realize that they belong to a global community of English speakers, where their NNS [non-native speaker] variant is not any less inferior to that of native speakers of English. This improved their self-esteem, and confidence in speaking English, and mitigated their fear for negative evaluation.

(NA05)

Both Italian and Norwegian teachers reported a shift in their perspectives. In addition to reflecting on their own shift in assessing pronunciation away from a native-speaker centric orientation, teachers viewed their students through the lens of how well students work within ELF concepts. In the sample lessons, we noted that teachers stressed students’ exposure to varieties of English from across the globe. Furthermore, teachers discussed the concept that English belongs to everyone, including themselves. The validity of an assessment will thereby be enhanced as the learners exhibit improved self-esteem and confidence in their oral abilities and increased oral production when communicating in ELF.

3.1.2. Tests and Grades

Teachers are somewhat divided and tend to position themselves differently on whether tests and grades are important means to evaluate student success in the classroom. A total of nineteen teachers either disagreed or strongly disagreed that tests and grades are important, whilst six were neutral and five believed tests and grades were important (see Table 2). Through their answers to the Activity 1 as well as in their assignments, teachers’ comments explain the wide range of responses. These show that teachers have several different challenges to address: local, national and high-stakes tests versus the importance of Assessment for Learning (Black and Wiliam 2018). It seems that the principles of ELF conflict with the long-held beliefs of testing standards (Widdowson 2021).

The following comment reflects the Norwegian national professional development incentive for teachers to develop their confidence of Assessment for Learning (Norwegian Directorate for Education and Training 2011).

Good and thorough feedback either in writing or as a video or audio response to students’ work is much more efficient and motivates them to improve, not just chase the numbers. But it is time consuming.

(NA03)

While feedback can be time consuming for teachers (see next subsection), the support from policy makers has made an impact on teaching practices. Learner-centered assessment, in which the student not only learns but is capable of contributing to knowledge (Bremner 2021) is an essential element of ELF-aware assessment. It positions learners in the assessment process of their own communicative capability using all of their linguistic resources.

Teachers also revealed an awareness of the value and the need of using tests to measure students’ learning, even though communicative competence is often overlooked in tests. The following teacher also highlights her students’ exposure and interaction with English outside the classroom and that learners come into class with their own language repertoire, including a wide range of skills perhaps picked up outside of the classroom (Sundqvist 2009; Sundqvist and Sylvén 2016).

I think that success may be evaluated through tests and grades, but that isn’t always true: in my experience you can have a child who loves attending English lessons and feels involved and tries his best to collaborate during the lessons, even if they don’t reach high marks at the moment. Sometimes there are children who don’t like to study, or to do homework, but, outside school, in a real-life situation I have seen them, they show great competence, no fear to communicate and unexpected creativity to interact.

(I04)

The following teacher is aware that tests and grades are only some of the ways to measure learners’ progress and there are many ways to test learners such as the methods teachers use to teach, particularly in multilingual/multicultural contexts.

In today’s diverse classrooms, it is as important for teachers to have different ways of assessing students as they do with teaching them. In addition to state and national standardized tests, teachers should use a variety of research-based methods throughout the learning process to provide students chances to prove their learning in multilingual classrooms.

(I01)

In spite of the absence of long-standing pre- or in-service teacher education courses in language testing and assessment in Italy and Norway, teachers have been using their classroom experience to reflect upon current assessment practices and to devise new forms of assessment that would represent their students’ learning experience and cater for the needs of their multilingual classrooms (Flognfeldt et al. 2020; Lopriore 2021).

3.1.3. Feedback

Almost all of the teachers agreed, or strongly agreed, that written and oral feedback is an important part of the learning process (see Table 2). Written and oral feedback is part of the formative assessment process which is also central to ELF assessment practices.

Formative assessment is important in ELF as the focus is on communication within multilingual spaces. As teacher NA03 points out in her assignment, the focus is for students to concentrate on how they can improve their communication skills. In mediation practices, it is the interlocutors who know if the communication has succeeded or not.

Formative feedback is given as written feedback (on their video/audio/presentation/hand-in); no grading. I want the students to focus on their own learning and how they can improve, not reduce all their learning and work towards a grade, an arbitrary number.

(NA03)

IO4 below, raises two important issues, that of age-related feedback types and learners’ responses, and that of ways of sustaining and enhancing feedback for all students.

It depends on the age of the learner: when children are young, they need to be motivated by the activities, to feel engaged, the feedback reach them unconsciously through repetitions, chants, songs, role play; in grade 4 and 5 of primary school some of them start being aware of the learning process and they ask for feedback in a more official way: when I ask for volunteers to say the alphabet (for instance, could be the months, or the days of the week, or other topics), my pupils pay a lot of attention to the suggestions I give their friends and this helps awareness. As for the written feedback, it is usually a due act, except for the times when the children themselves ask me to have a test in order to show me how confident they are in a new topic.

(I04)

The teacher’s observations are based on her primary teaching experience. During primary school education, learners evolve very rapidly, and they grow in their experiences both emotionally and cognitively. Learners at this level are capable of reflecting upon their own learning experiences, of remembering them and of building on them. They learn to observe what happens within the classroom as well as the processes enacted by teachers to engage them (Enever 2011). During the last phase of the primary cycle and during the ‘tests’, they are aware of the ‘format’ and of the rationale of using a test, so they rely upon their previous experiences to face it and they ask to be tested because they need feedback (also in Lopriore 2015).

Formative feedback by both teachers and peers was discussed in the lesson plans of the teachers, assessing both classwork and the final presentations. Teachers wrote that feedback would be provided to students through teacher, peer- and self-assessment, as well as through task achievement on communicative activities.

In both activities, emphasis is placed on communication through interaction with peers that values the learner-user’s own personal experiences. This provides an opportunity for the teacher to facilitate interactions and provide feedback to the groups and in plenum.

(NA08)

The teacher will give feedback to the whole class after each phase of the lesson and will give individual feedback on the final presentation.

(IA02)

Teachers tried to find a balance between free communication and working with language form (fluency vs accuracy), some only corrected when it interfered with meaning. Working with fluency and communication are both important elements in ELF-aware assessment.

I correct only if it is necessary, so only when it disturbs understanding.

(NA04)

Students carried out 3 different assessment activities where they were required to reflect on mistakes, results and give personal opinions too.

(IA14)

3.1.4. Self-Assessment

The majority of teachers also reported using self-assessment in their classrooms (see Table 2), whilst seven teachers are neutral and three reported not using self-assessment. Again, the neutrality may reflect the process which teachers are undergoing in thinking about the importance of tests and grades as opposed to formative methods such as self-assessment.

Through their comments, we found out that teachers highlighted the importance of self-assessment and how it is implemented in their multilingual classrooms, as required in the Norwegian curriculum. A common practice for pupils was to submit exit notes to the teacher on the way out of the classroom (also in Flognfeldt et al. 2020). However, this does not always focus on the student’s own work and learning, but on the enjoyment of the lesson, which often turns into an assessment of the teacher’s performance as noted by the Norwegian teacher below.

In the latter part of the lesson, I would try a kind of feedback by simply asking the learners if this lesson was something they wanted to do again sometime. The learners could only answer yes or no. This is a way to find out what the learners think about the lesson at the time being. However, it is not reliable feedback on how the lesson went. To write yes or no on a note, does not say anything about why they would like to do this again or why not. The feedback could show that some learners felt they learnt something, but it could also be that someone wrote yes because they liked the teacher or no if they disliked them. There is really a lot of bias to consider.

(NA02)

In their assignments, teachers also point out the importance of the curriculum influence on self-assessment. The focus must be on how an individual performs in his or her own language repertoire and be able to develop as needed. Indeed, the Norwegian curriculum (Utdanningsdirektorat 2020) states that students must be able to reflect on their own work and learning and apply their existing linguistic knowledge to their language learning. As the student reflects on their own learning processes, they also reflect on their own language repertoire and experiences, which is fundamental for ELF and for multilingual practices. Governmental support and mandates are an important part of the implementation of such practices.

Being aware of one’s own limitations and strengths is an important part of being able to meta-think about one’s own learning process. This is increasingly focused on part of the Norwegian general curriculum, so I think it must be adhered to.

(NA03)

The Italian teacher’s comment below highlights the importance of using diverse modes for implementing classroom-based assessment. Through the proposed techniques, learners in multilingual contexts can adopt a different perspective and strengthen the evaluation of their performance from different angles. Teachers also refer to the cognitive and metacognitive dimension elicited by self-assessment activities. This shows how working within a reflective perspective has triggered a deeper understanding of their learners’ developmental process.

I absolutely use this activity with my multilingual pupils as soon as they can afford it (for first graders it’s too soon); it’s a very useful part of the learning process. Sometimes we do it all together, sometimes in pairs or small groups, sometimes I give them the key answers and they check individually. I think this is a moment of metalinguistic and metacognitive awareness.

(I04)

Some Italian teachers have been using the European Language Portfolio (ELP) (Council of Europe 2011) which has a very strong focus on self-assessment in their multilingual classes. The Italian Ministry of Education (MIUR) has strongly supported the Council of Europe language policies, by adopting the CEFR competence levels as exit language levels in the school curriculum, and by encouraging teachers’ use of the ELP and the ELP self-assessment grids.

Yes, I use pre-prepared self-assessment grids which they use in their group discussions or pair work. I also use the ELP, still a very pedagogically valid tool in the language classroom.

(I10)

Teachers also offered general descriptions of peer-assessment in their assignments. Different contexts included engaging in a communicative task, checking answers in a teacher-led discussion, or feedback while working in groups.

I wanted to engage them more and have them produce more language by practicing conversations and debates with their peers.

(NA09)

Several teachers reported a shift towards a more student-centered classroom in multilingual contexts, which resulted in a lot of peer support between students with minimal interjection from the teacher.

The lesson is student-centered: learners have a central role, and the planned tasks are based on peer cooperative work. During the activities students are requested to communicate more than the teacher who acts as a facilitator. Peer tutoring: more expert students help less expert ones.

(IA01)

3.1.5. Communication

All teachers agreed or strongly agreed that communication should be considered as the primary goal of communicative classroom activities (see Table 2). In other words, the focus of assessment of communicative activities should be on how successful communication is between two interlocutors.

In their comments, teachers stressed that the primary focus of ELF assessment should be on communication, e.g.,

Getting a message through is more important than producing an utterance with no mistakes

(I02)

Teachers further explored the concept of communication by working with their students in the final lesson plan presented in their assignments. Observing language in action changes the focus of assessment to the ability to conduct a communicative activity. Teacher NA03 observed that students who refused to speak in class before were suddenly speaking because the focus of the class had shifted. Focusing on communication rather than pronunciation changes not only what is assessed, but how learners perform. In this case, it removed the stress of error correction so that students could communicate for the first time in class. As teacher IA12 pointed out, ELF lessons focus on fostering communication rather than pronunciation.

Some students that had not spoken much in class all year were suddenly speaking in their own English—where they previously had totally refused to do so. We had a good and long conversation as a class about how the subject of English is about communicating, and not so much about the language itself.

(NA03)

The most effective thing that I have learned is that English belongs to everyone, not only to native speakers and that I, as a teacher, must foster communication, and avoid stressing pronunciation that much.

(IA12)

3.2. Teachers’ Assessment Practices

In Activity 2 (Appendix A), teachers were asked what informed their assessment practices the most and were asked to select their choices from a list (see

Table 3. Teachers’ Assessment Practices.

	Norwegian	Italian	Total	Percentage
a. International high stakes exams	0	1	1	2.5%
b. National tests	1	4	5	12.5%
c. Local tests	0	3	3	7.5%
d. Classroom assessment: demonstrating progress throughout a course	2	9	11	27.5%
e. Curriculum aims	2	4	6	15%
f. A high standard of British English: training my students to a very high standard, as close as possible to perfect British English	0	0	0	0%
g. A high standard of native English: training my students to a very high standard, accepting both British and North American forms	0	1	1	2.5%
h. A high proficiency of a non-native variety: accepting non-native variances and encouraging their use	0	3	3	7.5%
i. Communicative ability: demonstrating communicative competences through interactive activities	2	8	10	25%

).

Only two Norwegian and eleven Italian teachers responded to this activity. Teachers were asked to mark their top three choices, resulting in a total of 40 responses. The most important assessment practice seems to be classroom assessment (27.5%), while communicative ability slightly outweighed testing (25%), which is still a very important factor with 22.5% of international high stakes, national or local tests. Teachers have a variety of testing practices to use and respond to, according to the local needs. With regard to pronunciation, 7.5% of responses preferred non-native variances, while only one teacher preferred a native speaker model. While British English no longer stands alone, one teacher did choose a native variety of English as the best to learn, while three chose non-native accents. While we expected support for non-native varieties (the assessment module was placed towards the end of the course), it seems that there is still some preference for inner-circle varieties.

Classroom assessment was reported as the most important context in which they operated followed by curriculum aims. We expected classroom assessment to be high in importance in the context of good ELF practices, as well as demonstrating communicative ability. Norwegian teachers marked curriculum aims as one of their top priorities, as was expected (Mølstad and Karseth 2016). In the Italian context, national and local tests were also important, as the students need to complete a course at a given level in the CEFR, such as B1 or B2 (MIUR 2012, 2018; Lopriore 2021).

3.3. Textbook and Learning Materials

Teachers were asked to review their textbooks or learning materials and consider whether native speaker norms were used to measure oral skills, if formative or alternative assessment practices were included, and whether communication was the goal (see Activity 3, Appendix A). On that basis they were also asked if they felt they could adapt an assessment strategy or technique or create an ELF-aware one.

Teacher NA05 noted the impact that textbooks have on the classroom, that the content is based on inner circle norms, which seems to impact ideology and identity and thereby influence assessment performance.

Making the lesson ELF aware was not an easy task when suitable materials are not available because teachers and school authorities resort to the textbooks and materials that are based on the inner circle norms. The consequences of this decision affect teaching considerably in terms of ideology and identity.

(NA05)

Textbooks used by our study participants seem to lack ELF-aware assessment tasks. This places an extra burden on teachers who, in both contexts, report creating their own authentic materials, especially spoken recordings.

For me, freeing myself from the textbook assignments that are so easily available, but not very ELF aware, also took some effort.

(NA09)

There aren’t any specific formative assessment practices. In some chapters of the literature course, at the end of most of the topics there are some grids and maps to complete. If students complete and share them in plenary, I think that those materials can be useful for formative assessment.

(I14)

Teacher I14 refers to literature coursebooks and reflects on the paucity of assessment activities in this field where, differently from language courses, evaluation is mostly linked to either knowledge of authors or of stylistic aspects. These teachers’ comments reveal how they are already moving ahead of what coursebooks are offering and suggest ways of introducing forms of alternative assessment in their multilingual classes. The Italian teachers said that the models used in textbooks are usually the native speakers, English or American, and that the main aim is “communication, and that forms of self-assessment are sometimes provided”. (I07)

The only part in the textbook that fits the idea of an ELF assessment is a kind of self-assessment with a list of abilities they should have acquired during the unit so that they can say if the target has been reached or not. My colleagues and I usually create by ourselves the spoken tests for a more formative assessment, and we try to create as many authentic tasks as the situation offers.

(I04)

If the teacher is using ELF-aware practices, then there is a mismatch between the teaching and the assessment activities in textbooks. Some teachers also stressed that their textbooks had authentic texts that were only in standard English and did not feature different models of English. Some textbooks featured assessment items that focused on discrete responses, while other textbooks featured assessment tasks that entailed self-assessment, peer work, and task-based activities. The quality of these activities may not necessarily be in line with ELF practices. It is hard for teachers to predict if textbook assessments are ELF-aware, as the concept of ELF-aware assessment had not been discussed before in this course. Still, it was during the course that teachers started to develop their own ELF-aware assessments particularly in the final assignment. As they got involved in a reflective ELF-aware perspective and gradually became acquainted with this approach, it was inevitable for them to look for feasible ways to inform their assessment practices.

3.4. Observation Form and Final Assignment

In the fourth and final assessment activity of the assessment course, teachers were presented with several accommodation techniques used by highly successful multilingual speakers (Kouvdou and Tsagari 2018) as alternatives to assess communicative competence in oral exchanges (see Activity 4, Appendix A). Presented in a table, teachers were asked to reflect on the possibility of using the grid as a type of self-assessment or alternative assessment.

I can absolutely use this form to assess my students’ oral skills, especially for the formative assessment. I do not wish to change anything because all the strategies are useful to obtain communication.

(N04)

I can adapt the observation form through explanation, turn-taking, topic control, feedback, and conversational pair work.

(N05)

Teachers had several ideas for how to adapt the framework in their courses, e.g., use it for self- and peer-assessment, as well as teacher-led assessment. They all seemed at ease with using rubrics for assessment.

The observation form includes items that reflect the communicative competence model of Bachman and Palmer (Linguistic Competence, Socio-linguistic competence, strategic competence and discourse competence) with an overt ELF perspective which allows for mediation and translanguaging strategies for achieving communicative understanding and co-construction of meaning. I would add more communicative functions and skills so that I could also be assessing their performance skills as well as their accommodation strategies, perhaps not all in the same lesson or day, but it would be useful to have a grid with both communicative skills, functions, and strategies.

(I10)

Teacher I10 referring to Bachman and Palmer’s (1996) discourse competence model, considered the many facets required for a better assessment of oral interaction. Other teachers considered the strategies that highly-skilled multilinguals use, as presented in the course:

This form is a good way to observe negotiation and communicative skills of students in oral interaction. I do not think it is applicable in over-populated classes. I would use this form for small speaking groups or peer-assessment activities.

(I13)

Some teachers pointed out that some of the criteria are not suitable for younger learners or larger classes:

I could use this observation form to assess my students’ oral skills perhaps leaving out “adapt grammar for interlocutor” and “adapt vocabulary for interlocutor” because my students are aged 11 to 14 and they often learn things for the first time at this stage. The original descriptors will be customized in order to fit the designed activity better.

(IA01)

I think this form could be adapted for self-evaluation or maybe even peer review. One of the changes that would be needed would be to replace the linguistic terms with more accessible terminology such as “dialogue partner/discussion group instead of “interlocutor” for example.

(N09)

Teachers developed specific grids to accommodate their lesson. References were also made to CEFR descriptors and to self-assessment grids adapting them to the format of their own lesson plan. Not everybody devised a complete grid. They just mentioned they were going to use the grid in formative assessment, e.g., “Students are evaluated through a grid previously shared with them” in “Group work” or in “Effective final tasks” (Italian group).

With regard to the final teachers’ assignments, detailed analysis showed that the teachers in both groups used the assignment as an opportunity to evaluate the use of ELF-aware practices in their classrooms and report the change in student and their own perceptions towards language and language use.

In the Italian context, several teachers recently received training in formative and summative assessment practices through INVALSI, and implemented summative ELF-aware assessments in their lessons. The Norwegian teachers do not need to address summative requirements, but the widespread use of formative assessments in their lesson plans reflect the national implementation of assessment for learning practices.

4. Summary and Discussion

The current study investigated the perceptions of a group of Norwegian and Italian English language teachers towards ELF-aware assessment within the framework of an online ERASMUS+ CPD course. Within this context, we first explored how teachers understand ELF-aware assessment (RQ1). The results showed that, first of all, ELF awareness in both contexts was more widely spread than we expected. Their classrooms are portrayed as multilingual spaces and teachers are functioning in spaces that are already anchored in English as a lingua franca. Teachers demonstrate that they are practicing a variety of assessments, particularly self-assessment within a formative framework.

Many teachers also stressed that welcoming a variety of pronunciations was a new concept but were willing to do so and experiment with their students. Even though the majority opted for a preference for localized pronunciation forms, one teacher preferred an inner-circle English accent to outer or expanding circle accents while another conducted a lesson that focused on intelligibility of accents, both in comprehension and production. Teachers from both countries reported limited representation of authentic multilingual voices (sound recordings) in textbooks and therefore, they created their own authentic tasks and authentic sound recordings. Authentic resources and language use was the core of almost all the teachers’ lessons, confirming teachers’ understanding of the importance of authentic resources (Harding and McNamara 2017).

However, in answering RQ2 and RQ3 (use and methods/approaches of ELF-aware assessment), we found out that there was variation in practices and principles regarding the use of alternative methods, feedback provision, the role of students in the assessment process, the tests’ design, and the use of rubrics and specifications. Practices varied in terms of who administered the assessment, if it was used by the teachers, by peers, or in self-assessment. Some teachers felt that they should employ all types of assessments, while others were comfortable involving students too. Most teachers restrained from correcting form unless it interfered with meaning, confirming their understanding that intelligibility should be viewed as more important than correctness (Seidlhofer 2004). Exactly what constitutes meaningful interference and how correction is implemented merits further study. Teachers also shared that assessment should cover a variety of criteria, not only “mistakes” but also successful completion of a communicative task, assessing using reflection, “personal opinions”, “results”, and analysis. These elements are different kinds of formative assessments, but there is no systematic use of one kind of assessment over another.

With regard to the RQ4 (contextual needs teachers need to address when practicing ELF-aware assessment), the results showed that there was a general agreement amongst teachers of both groups that assessment activities in ELF environments should be communicative, focusing on oral communication (Ayden and Karakas 2021), mediation practices (Seidlhofer 2004), and student-centered learning. Teachers reported overwhelmingly that it was important to assess communicative ability. Some teachers also viewed student-centered learning and prioritizing practicing oral skills as a new approach. Overall, there was a shift in teachers’ understanding that self-assessment is an important element of ELF-aware assessment. All the lesson plans included a focus on students dialoguing or working in groups on communicative tasks.

However, even though the rubric of skills of highly effective multilingual speakers was presented to the course participants as a short, simplified list of skills that teachers readily engaged with when reflecting on ELF assessment, context seemed to play an important role. For example, the Norwegian group tended to use more general assessments with the students with exit tickets, checking answers in pairs, or reflecting on the lesson, being used. On the other hand, in Italy, three teachers adopted their own rubrics for assessment where teachers are accustomed to working with rubrics and “I-can” statements due to CEFR being widely implemented in textbooks and becoming a regular part of teacher practice. One teacher (IA01) used plurilingual and mediation rubrics from the Companion Volume (Council of Europe 2020) for her ELF-aware assessment. The CEFR and the Companion Volume are not widely used in Norway, as teachers instead focus on local curriculum requirements and would require widespread training and implementation in textbooks.

Nevertheless, in both contexts, the use of self-assessment was prominent, and this changes the relationship of teachers and students too. ELF-aware assessment seems to require students to take an active role in the learning and assessment which is in contrast with more traditional systems where summative orientations towards assessment are practiced.

Teachers understand that ELF is centered on communication with a specific focus on conversational turn taking. For some, it is in its broadest sense that conversation is happening independently between two learners. They recognized specific elements, such as turn taking, feedback, mediation and translanguaging, to help maintain effective communication. One teacher went as far as connecting assessment to Bachman and Palmer’s (1982) communicative competence model.

The assessment of discourse is centered on formative assessment within groups, assessed by teachers, learners, or self. The decision to allow peer- or self-assessment of language production has to do with teacher comfort in allowing students to partake in assessment, even if it is classroom formative assessment, depends on teacher comfort with the democratization of classroom roles. Teachers had different approaches to their assessments in practice. Some left conversation as open and unrestrained as possible, while others, particularly in the Italian group, devised grading rubrics to be used in self-assessment and summative assessment at the end of the lesson. One teacher devised a plan to use ELF as the topic of the oral exam, differing from assessing the oral exam in an ELF way. ELF is a new concept for some leading teachers to explore the concept with their students. Support should include how to implement assessment of ELF student practices.

In conclusion, ELF-aware assessment of the teachers in this study focuses on formative assessment, mostly as alternative assessment: performance-oriented assessment based on mutual comprehensibility and successful communication. Teachers teach in two paradigms: balancing teaching successful communication and meeting testing/assessment mandates. They are aware that they have to teach their students to communicate and at the same time teach the textbook or meet curriculum aims, and, in the case of Italy, prepare students for their standardized tests. Teachers are aware of the conflict between supporting their students at being competent users of the language, versus teaching to pass the test, taking them through certification exams successfully.

Is there space for ELF in assessment? Not as it currently stands. Assessment should reflect the realities of our multilingual classes. Rather than require students to become native speakers, ELF assessment focuses on helping students to become competent users of the language so that pronunciation, word form and lexical choice of ML speakers are as valid as those of native speakers. The evolving classroom context that characterizes multilingual settings demands teachers to be flexible and innovative.

The development of ELF-aware assessment, as seen in the current study, is an evolving topic which in some ways is being implemented by trial and error in the classroom. Even though it is not possible to generalize to a larger population, it is our belief, on the basis of the results from the current study, that the field of language teacher education and practice can move forward in a reflective practice between researchers, educators, and learners. However, further research focusing on student experiences and reflections is needed.

5. Conclusions

On the basis of the results of current study, it is evident that teachers need support in developing both ELF teaching practices and ELF assessment. In the two multilingual contexts investigated, training and guidance at the national level in language assessment is needed, particularly in issues of assessment in ELF contexts and indeed in the use of the Companion Volume where new concepts and practices for most teachers in multilingual contexts are presented. Familiarity with both ELF and ELF terminology and how to assess ELF communication is also needed. Therefore, teacher training development courses in language assessment aiming to raise teachers’ language assessment literacy (Taylor 2013; Tsagari 2020) need to adopt expanded notions of accommodating the assessment needs of multilingual students and their teachers (Tsagari 2022).

The field of research has also put learners at the center of activity. For example, there has been a surprisingly large number of recent studies on learners’ perceptions of ELF (Guerra and Bayyurt 2019), on topics such as attitudes towards accents, positive self-perceptions, evolving identities, and beliefs (Lopriore 2019). However, even though there are studies on student perceptions and experiences with self- and other types of formative assessment, studies of self-assessment within ELF are sparse. Further research is also needed on the type of feedback that teachers give in ELF-aware assessment, how and what teachers focus on in their feedback, and how this is conducive to language learning. Publishers and textbook writers should also investigate recent research findings in the field of ELF and assessment in multilingual contexts and propose more ELF-aware activities and forms of assessment. Additionally, future research should investigate current publishers’ choices as it regards ELF-aware practices and assessment. Such research findings would provide useful information to support individual countries’ language policies so that ELF is included in the curriculum.

One of the limitations of the study was that teachers had a restricted time to implement their lesson plans, a requirement of the CPD course they attended. In both countries, these pressures were further exacerbated by the COVID pandemic and sudden switch to online teaching. This placed extraordinary pressure on the teachers, which might have resulted in lower completion rates of the course. In Italy, digital schooling started at the beginning of the course making it impossible for teachers to implement their lesson plans. In Norway, the government cancelled exams in the spring of 2020 due to concerns that they would not be able to deliver fair exams under lockdown and schools went to full or partial digital teaching in March until June. All teachers had restrictions on how they could carry out their lessons: shortened lessons, distancing between desks that inhibited talking, inability to carry out assessments, and most importantly, the time pressures that were exacerbated by the pandemic to get the basic curriculum delivered. These events had an effect on all assessment practices, including in our study.

Furthermore, this study only examined the beliefs and practices of teachers who had taken part in the ENRICH course, and therefore classified it as a small-scale study and, as such, cannot be generalized to the general teaching population. A larger, in-service sample of teachers’ beliefs and practices is needed in order to illuminate differences in teachers’ beliefs and practices before and after the course.

However, despite the limitations, the participating teachers were already transitioning to work with ELF-aware assessment practices, although they do not yet have the labels for these practices. Perhaps that is why our intervention (implementation of the ENRICH course) landed softly. Teachers were open to trying out new assessment ideas and many of their practices (as seen in their assignments in particular) already agreed with our assessment suggestions. In the Norwegian and Italian contexts, teachers have been working with multilingual populations and making adjustments to their teaching. It seems that participants’ engagement in the CPD course has created opportunities for learner-centered approaches to assessment and prompted participants’ previous competences and experiences in multilingual classrooms. Therefore, it is safe to conclude that perhaps this is a moment, with the effect of high migration on the educational systems, that teachers are ahead of the researchers and have a lot to share from their experiences.