Utbildning på engelska vid svensk universitet: Stödja undervisning, förbättra lärande
The amount of teaching and learning at universities that occurs in second language (L2) English continues to increase rapidly. This situation leads to classrooms where students and teachers have varying levels of English proficiency. L2 listening and speaking skills impact the quality of education and learning that occurs. As such, students and teachers in English medium education face a number of challenges. This one-year sabbatical will allow for in-depth investigations of existing data generated from L2 English lectures and related measurements of student learning. Data include lecture recordings, transcriptions, teacher interviews, student surveys, and student reports collected after lectures. Importantly, the data set provides insights in to the relation between the teacher’s spoken L2 English output and the students’ receptive listening ability and learning. All data was collected from 2019-2020 and has generated initial findings; however, several possible research avenues have yet to be explored in this unique data set, including lecture strategies adopted by L2 English-users, affects of various aspects of teachers’ spoken output (e.g., accent, rate of speech), and the influence that student first language (L1) may have on L2 English lecture comprehension in the Swedish university context. This project will allow for exploration into these topics and will generate a series of publications in high-level scholarly journals.
Final report
This report covers activities from 1 September 2024 to 31 August 2025.
The amount of teaching and learning at universities that occurs in second language (L2) English continues to increase rapidly. English medium instruction (EMI) situations like these often lead to classrooms where students and teachers have varying levels of English proficiency. L2 listening and speaking skills impact the quality of education and learning that occurs. As such, students and teachers in English medium education face a number of challenges. This one-year sabbatical allowed for in-depth investigations of data generated from L2 English lectures and related measurements of student learning, along with interviews with university faculty who teach in L2 English. Data included lecture recordings, transcriptions, teacher interviews, student surveys, and student reports collected after lectures. Importantly, the data set provided insights in to the relation between the teacher’s spoken L2 English output and the students’ receptive listening ability and learning. The project allowed for exploration topics such as teacher keyword use and lexical choices along with the influences that teachers’ linguistic backgrounds can have on education.
Data included recordings and transcripts of EMI lectures, student surveys, post-lecture student reports, and EMI teacher interviews. To implement the various analytical procedures, specific tools were used with different types of data. ELAN software was used to investigate questions related to teachers’ rate of speech during different teaching sessions, while AntWordProfiler was used to study teachers’ vocabulary choices while lecturing. These analyses constituted quantitative aspects of the project. In addition, teacher speech in lectures was studied qualitatively using Sinclair and Coulthard’s (1975) framework for understanding and categorizing classroom discourse. These factors were investigated both from a comparative perspectives (i.e., comparing the practices of multiple teachers) and from a longitudinal perspective that monitored the evolution of teaching practices made by the same EMI teacher over a series of classes. Various existing word lists (e.g., Academic Word List (Coxhead, 2000), and Spoken Academic Word List (Dang, et al., 2017) were used to explore the frequency and accessibility of teachers’ lexical choices. Teacher interviews were also analyzed qualitatively, probing questions about linguistic and academic backgrounds as well as teacher views on student preparedness for studying content and general academic work in L2 English.
The first goal was to explore differences between EMI teachers in terms of rate of speech, vocabulary usage, and keyword framing during lectures. Previous work (Siegel, 2022) informed this analysis by indicating two lectures that students had either relative success or particular challenges at understanding; specifically, where students were either able to report high levels of teacher-selected keywords or where students struggled to do so. A deeper, comparative investigation of those two lectures was therefore conducted. This involved ELAN software to analyze rate of speech at evenly-distributed points throughout the lectures. Vocabulary analysis included use of vocabulary lists and lecture transcripts analyzed using AntWordProfiler, which gives data output on the frequency of vocabulary coverage. In other words, this analysis demonstrated how accessible the teachers’ speech was in terms of vocabulary choices and keyword frequency. Framing of keywords (i.e., how teachers presented keywords during EMI) was qualitatively investigated by identifying utterances that contained keywords and employing Sinclair and Coulthard’s (1975) framework to categorize utterances into, for example, informatives, asides, and comments. A sample of the analysis of keywords appearing in various classroom speech acts shows that informatives appeared over 80 times in one lecture but over 50 times in another. Likewise, metastatements were used more than 20 times in one lecture but only two times in another. Findings like these suggest that different teachers uses these speech acts to varying degrees, meaning students have different access to the information.
A second goal was to monitor the evolution of EMI teachers’ output in a longitudinal fashion. This perspective is based on second language acquisition theories that posit teachers use modified spoken output to accommodate listeners in the beginning and that such scaffolding can be gradually retracted. Such accommodation may be related to accent familiarity, rate of speech, vocabulary use, and repetition of keywords in teacher speech. To investigate these aspects, three lectures taught in succession by the same EMI teacher were in focus. Analysis of lecture recordings and transcripts given at three points during a course monitored for any observable changes in teacher speech that might align with second language acquisition and accommodation theories. The study employed ELAN software to monitor rate of speech and AntWordProfiler software to compare the teacher’s word choice with word lists of common usage (e.g., General Service Word List, Academic Word List, etc.). Tables 1 and 2 below provide examples of speech/articulation rate and vocabulary choice, respectively.
Table 1
Speech rate analysis
Syllable/ min Speech Rate Total pause time Articulation Rate in sps
Lecture 1 Averages 177 2.95 21 4.56
Lecture 2 Averages 188 3.14 21.15 4.6
Lecture 3 Averages 210 3.5 16.85 4.87
Table 2
Vocabulary coverage analysis
Lecture 1 Lecture 2 Lecture 3
GSL 1K 86.2% 86.5% 85.8%
GSL 2K 3.2% 3.1% 4.6%
AWL 4.5% 4.1% 4.3%
Off list 6.1% 6.3% 5.3%
Total 100% 100% 100%
Findings from this study suggest some unconscious evolution of language use over the course of a series of EMI lectures, with the teacher increasing speech and articulation rate as the lectures progressed, thereby suggesting more confidence in accent familiarity and listener accommodation as well as comfort level with the teaching style and/or subject matter. Vocabulary coverage generally remained consistent and did not evolve, with approximately 90% of vocabulary coming from the GSL1K and GSL2K in all three lectures, with between 3.1% and 4.6% of words appearing on the AWL. However, keyword appearance did evolve, with a clear reduction in keyword appearance, likely as the lectures shifted from more general to more specific disciplinary knowledge.
The third goal for the project was to explore diversity among EMI teachers and understand how their linguistic and educational backgrounds may impact their views on EMI students as well as their teaching practices. This qualitative angle on the topic involved analysis of a series of interviews with EMI teachers, some Swedish and some expatriate teachers who work in Sweden. Previous research on EMI teachers has generally viewed them as a homogeneous group rather than appreciated the potential heterogeneity that they embody. The study focused on how teacher background factors (e.g., linguistic, academic) can influence their self-perceptions as EMI teachers and how they perceive EMI student preparedness (e.g., English proficiency, subject content knowledge) and group characteristics among the students they typically work with. Results showed the Swedish teachers had the option of translanguaging (i.e., using both Swedish and English) when teaching, an option not available to the expatriate teacher group, thereby impacting their teaching. Another important factor was where these faculty members had completed their doctorate studies, either in an anglo-phone country or in Sweden. Some expatriate teachers had more experience teaching in a variety of international contexts, which seems to have equipped them with various teaching strategies to accommodate EMI students with various proficiency levels.
The project’s three most important results are as follows:
1. EMI teachers’ use and positioning of keywords during lectures differs and can impact potential for student learning;
2. Evolution of linguistic features of EMI teacher output are evident and should be taken into account when addressing various groups of EMI students;
3. EMI teachers comes from a range of domestic and expatriate contexts and bring various individual background traits to their teaching, which can affect performance, perceptions, and student learning.
The worldwide dominance of English in many aspects of society has also affected the education sector, where English medium instruction (EMI) programs have become common (e.g., Bolton et al., 2024; Dafouz & Smit, 2020; Lasagabaster & Doiz, 2021). In EMI, English is used as a second or additional language (L2) by either the teacher and/or at least some of the students in a particular educational setting. Given that many educators and students who participate in EMI use L2 English (i.e., they are or have been English language learners at some point), attention to the quality of teaching and education in such situations is paramount. In other words, the potentially great variations in student and teacher English proficiency may affect learning. Ozer (2020) and Morell et al. (2020), among others, have promoted systematic attention to the quality of EMI and highlight the importance for pedagogic and linguistic training for teachers. Similarly, Cañado (2023) investigated multiple aspects involved in EMI teacher training, analyzing training needs related to theoretical perspectives on EMI, linguistic competence, materials choices, and classroom methodologies at one university in Spain. Meanwhile, Deroey (2023) reviewed EMI training programs in 18 countries and found observation and feedback sessions to be a staple component to support teachers. Such training, however, is often in short supply, as is the monitoring of EMI for quality assurance, teacher performance, and student learning outcomes. Contributions from the present project serve to inform such teacher support in terms of content teaching, language use, and adjusting to various student groups.
The worldwide dominance of English in many aspects of society has also affected the education sector, where English medium instruction (EMI) programs have become common (e.g., Bolton et al., 2024; Dafouz & Smit, 2020; Lasagabaster & Doiz, 2021). In EMI, English is used as a second or additional language (L2) by either the teacher and/or at least some of the students in a particular educational setting. Given that many educators and students who participate in EMI use L2 English (i.e., they are or have been English language learners at some point), attention to the quality of teaching and education in such situations is paramount. In other words, the potentially great variations in student and teacher English proficiency may affect learning. Ozer (2020) and Morell et al. (2020), among others, have promoted systematic attention to the quality of EMI and highlight the importance for pedagogic and linguistic training for teachers. Similarly, Cañado (2023) investigated multiple aspects involved in EMI teacher training, analyzing training needs related to theoretical perspectives on EMI, linguistic competence, materials choices, and classroom methodologies at one university in Spain. Meanwhile, Deroey (2023) reviewed EMI training programs in 18 countries and found observation and feedback sessions to be a staple component to support teachers. Such training, however, is often in short supply, as is the monitoring of EMI for quality assurance, teacher performance, and student learning outcomes. Contributions from the present project serve to inform such teacher support in terms of content teaching, language use, and adjusting to various student groups.
New research avenues
During the project, a central focus was on teacher actions, decisions and backgrounds. This focus means that student perspectives and learning outcomes in relation to teacher performance in EMI have yet to be fully explored. Questions about whether student learning is impacted, for instance, by the frequency and/or framing of keywords during lectures or by any linguistic evolution on the part of a teacher throughout an EMI remain in need of focused research. Ideally, future research will account for both sides of the teacher-student symbiosis in EMI.
International dimensions and dissemination
The international dimension of the project is evident in the sharing of results and generation of dialogue with colleagues in a variety of contexts, including both at university level and at pre-university English for academic purposes preparation courses. This has included work with representatives from the Oxford University EMI Research Group, the British Association of Applied Linguistics’ EMI special interest group, a symposium on Contest Language Policies, Skolporten, and EMI teacher support work with partners in Japan, Italy, the UK, and Taiwan. Another international dimension was engaging with the Manchester Voices project, hosted by Manchester Municipal University and Manchester Central Library. The project's focus on language attitudes, accent perceptions, and linguistic backgrounds is relevant to the RJ sabbatical work.In addition, findings from the project were presented at the conference titled Educational Linguistics 2024: Language(s) from Childhood to Adult Age at Linnea University in Kalmar, Sweden.
The amount of teaching and learning at universities that occurs in second language (L2) English continues to increase rapidly. English medium instruction (EMI) situations like these often lead to classrooms where students and teachers have varying levels of English proficiency. L2 listening and speaking skills impact the quality of education and learning that occurs. As such, students and teachers in English medium education face a number of challenges. This one-year sabbatical allowed for in-depth investigations of data generated from L2 English lectures and related measurements of student learning, along with interviews with university faculty who teach in L2 English. Data included lecture recordings, transcriptions, teacher interviews, student surveys, and student reports collected after lectures. Importantly, the data set provided insights in to the relation between the teacher’s spoken L2 English output and the students’ receptive listening ability and learning. The project allowed for exploration topics such as teacher keyword use and lexical choices along with the influences that teachers’ linguistic backgrounds can have on education.
Data included recordings and transcripts of EMI lectures, student surveys, post-lecture student reports, and EMI teacher interviews. To implement the various analytical procedures, specific tools were used with different types of data. ELAN software was used to investigate questions related to teachers’ rate of speech during different teaching sessions, while AntWordProfiler was used to study teachers’ vocabulary choices while lecturing. These analyses constituted quantitative aspects of the project. In addition, teacher speech in lectures was studied qualitatively using Sinclair and Coulthard’s (1975) framework for understanding and categorizing classroom discourse. These factors were investigated both from a comparative perspectives (i.e., comparing the practices of multiple teachers) and from a longitudinal perspective that monitored the evolution of teaching practices made by the same EMI teacher over a series of classes. Various existing word lists (e.g., Academic Word List (Coxhead, 2000), and Spoken Academic Word List (Dang, et al., 2017) were used to explore the frequency and accessibility of teachers’ lexical choices. Teacher interviews were also analyzed qualitatively, probing questions about linguistic and academic backgrounds as well as teacher views on student preparedness for studying content and general academic work in L2 English.
The first goal was to explore differences between EMI teachers in terms of rate of speech, vocabulary usage, and keyword framing during lectures. Previous work (Siegel, 2022) informed this analysis by indicating two lectures that students had either relative success or particular challenges at understanding; specifically, where students were either able to report high levels of teacher-selected keywords or where students struggled to do so. A deeper, comparative investigation of those two lectures was therefore conducted. This involved ELAN software to analyze rate of speech at evenly-distributed points throughout the lectures. Vocabulary analysis included use of vocabulary lists and lecture transcripts analyzed using AntWordProfiler, which gives data output on the frequency of vocabulary coverage. In other words, this analysis demonstrated how accessible the teachers’ speech was in terms of vocabulary choices and keyword frequency. Framing of keywords (i.e., how teachers presented keywords during EMI) was qualitatively investigated by identifying utterances that contained keywords and employing Sinclair and Coulthard’s (1975) framework to categorize utterances into, for example, informatives, asides, and comments. A sample of the analysis of keywords appearing in various classroom speech acts shows that informatives appeared over 80 times in one lecture but over 50 times in another. Likewise, metastatements were used more than 20 times in one lecture but only two times in another. Findings like these suggest that different teachers uses these speech acts to varying degrees, meaning students have different access to the information.
A second goal was to monitor the evolution of EMI teachers’ output in a longitudinal fashion. This perspective is based on second language acquisition theories that posit teachers use modified spoken output to accommodate listeners in the beginning and that such scaffolding can be gradually retracted. Such accommodation may be related to accent familiarity, rate of speech, vocabulary use, and repetition of keywords in teacher speech. To investigate these aspects, three lectures taught in succession by the same EMI teacher were in focus. Analysis of lecture recordings and transcripts given at three points during a course monitored for any observable changes in teacher speech that might align with second language acquisition and accommodation theories. The study employed ELAN software to monitor rate of speech and AntWordProfiler software to compare the teacher’s word choice with word lists of common usage (e.g., General Service Word List, Academic Word List, etc.). Tables 1 and 2 below provide examples of speech/articulation rate and vocabulary choice, respectively.
Table 1
Speech rate analysis
Syllable/ min Speech Rate Total pause time Articulation Rate in sps
Lecture 1 Averages 177 2.95 21 4.56
Lecture 2 Averages 188 3.14 21.15 4.6
Lecture 3 Averages 210 3.5 16.85 4.87
Table 2
Vocabulary coverage analysis
Lecture 1 Lecture 2 Lecture 3
GSL 1K 86.2% 86.5% 85.8%
GSL 2K 3.2% 3.1% 4.6%
AWL 4.5% 4.1% 4.3%
Off list 6.1% 6.3% 5.3%
Total 100% 100% 100%
Findings from this study suggest some unconscious evolution of language use over the course of a series of EMI lectures, with the teacher increasing speech and articulation rate as the lectures progressed, thereby suggesting more confidence in accent familiarity and listener accommodation as well as comfort level with the teaching style and/or subject matter. Vocabulary coverage generally remained consistent and did not evolve, with approximately 90% of vocabulary coming from the GSL1K and GSL2K in all three lectures, with between 3.1% and 4.6% of words appearing on the AWL. However, keyword appearance did evolve, with a clear reduction in keyword appearance, likely as the lectures shifted from more general to more specific disciplinary knowledge.
The third goal for the project was to explore diversity among EMI teachers and understand how their linguistic and educational backgrounds may impact their views on EMI students as well as their teaching practices. This qualitative angle on the topic involved analysis of a series of interviews with EMI teachers, some Swedish and some expatriate teachers who work in Sweden. Previous research on EMI teachers has generally viewed them as a homogeneous group rather than appreciated the potential heterogeneity that they embody. The study focused on how teacher background factors (e.g., linguistic, academic) can influence their self-perceptions as EMI teachers and how they perceive EMI student preparedness (e.g., English proficiency, subject content knowledge) and group characteristics among the students they typically work with. Results showed the Swedish teachers had the option of translanguaging (i.e., using both Swedish and English) when teaching, an option not available to the expatriate teacher group, thereby impacting their teaching. Another important factor was where these faculty members had completed their doctorate studies, either in an anglo-phone country or in Sweden. Some expatriate teachers had more experience teaching in a variety of international contexts, which seems to have equipped them with various teaching strategies to accommodate EMI students with various proficiency levels.
The project’s three most important results are as follows:
1. EMI teachers’ use and positioning of keywords during lectures differs and can impact potential for student learning;
2. Evolution of linguistic features of EMI teacher output are evident and should be taken into account when addressing various groups of EMI students;
3. EMI teachers comes from a range of domestic and expatriate contexts and bring various individual background traits to their teaching, which can affect performance, perceptions, and student learning.
The worldwide dominance of English in many aspects of society has also affected the education sector, where English medium instruction (EMI) programs have become common (e.g., Bolton et al., 2024; Dafouz & Smit, 2020; Lasagabaster & Doiz, 2021). In EMI, English is used as a second or additional language (L2) by either the teacher and/or at least some of the students in a particular educational setting. Given that many educators and students who participate in EMI use L2 English (i.e., they are or have been English language learners at some point), attention to the quality of teaching and education in such situations is paramount. In other words, the potentially great variations in student and teacher English proficiency may affect learning. Ozer (2020) and Morell et al. (2020), among others, have promoted systematic attention to the quality of EMI and highlight the importance for pedagogic and linguistic training for teachers. Similarly, Cañado (2023) investigated multiple aspects involved in EMI teacher training, analyzing training needs related to theoretical perspectives on EMI, linguistic competence, materials choices, and classroom methodologies at one university in Spain. Meanwhile, Deroey (2023) reviewed EMI training programs in 18 countries and found observation and feedback sessions to be a staple component to support teachers. Such training, however, is often in short supply, as is the monitoring of EMI for quality assurance, teacher performance, and student learning outcomes. Contributions from the present project serve to inform such teacher support in terms of content teaching, language use, and adjusting to various student groups.
The worldwide dominance of English in many aspects of society has also affected the education sector, where English medium instruction (EMI) programs have become common (e.g., Bolton et al., 2024; Dafouz & Smit, 2020; Lasagabaster & Doiz, 2021). In EMI, English is used as a second or additional language (L2) by either the teacher and/or at least some of the students in a particular educational setting. Given that many educators and students who participate in EMI use L2 English (i.e., they are or have been English language learners at some point), attention to the quality of teaching and education in such situations is paramount. In other words, the potentially great variations in student and teacher English proficiency may affect learning. Ozer (2020) and Morell et al. (2020), among others, have promoted systematic attention to the quality of EMI and highlight the importance for pedagogic and linguistic training for teachers. Similarly, Cañado (2023) investigated multiple aspects involved in EMI teacher training, analyzing training needs related to theoretical perspectives on EMI, linguistic competence, materials choices, and classroom methodologies at one university in Spain. Meanwhile, Deroey (2023) reviewed EMI training programs in 18 countries and found observation and feedback sessions to be a staple component to support teachers. Such training, however, is often in short supply, as is the monitoring of EMI for quality assurance, teacher performance, and student learning outcomes. Contributions from the present project serve to inform such teacher support in terms of content teaching, language use, and adjusting to various student groups.
New research avenues
During the project, a central focus was on teacher actions, decisions and backgrounds. This focus means that student perspectives and learning outcomes in relation to teacher performance in EMI have yet to be fully explored. Questions about whether student learning is impacted, for instance, by the frequency and/or framing of keywords during lectures or by any linguistic evolution on the part of a teacher throughout an EMI remain in need of focused research. Ideally, future research will account for both sides of the teacher-student symbiosis in EMI.
International dimensions and dissemination
The international dimension of the project is evident in the sharing of results and generation of dialogue with colleagues in a variety of contexts, including both at university level and at pre-university English for academic purposes preparation courses. This has included work with representatives from the Oxford University EMI Research Group, the British Association of Applied Linguistics’ EMI special interest group, a symposium on Contest Language Policies, Skolporten, and EMI teacher support work with partners in Japan, Italy, the UK, and Taiwan. Another international dimension was engaging with the Manchester Voices project, hosted by Manchester Municipal University and Manchester Central Library. The project's focus on language attitudes, accent perceptions, and linguistic backgrounds is relevant to the RJ sabbatical work.In addition, findings from the project were presented at the conference titled Educational Linguistics 2024: Language(s) from Childhood to Adult Age at Linnea University in Kalmar, Sweden.