1.Introduction

The hegemony of English as a medium of education (EME) around the world is undeniable. The practice has been increasing steadily even in countries that traditionally have had little to no English education, such as the contexts under investigation in the present study: China and Spain (Lasagabaster, 2020Lasagabaster, D. (2022) English-Medium instruction in higher education. Cambridge: Cambridge University Press. Lasagabaster, D. (2022) English-Medium instruction in higher education. Cambridge: Cambridge University Press. ). EME, the denomination used in the present study, also referred to as English-medium instruction (EMI), is the practice of teaching academic subjects through English in contexts where English is not spoken as a first or official language (Macaro, 2022Macaro, E. (2022) English Medium Instruction: What do we know so far, and what do we still need to find out? Language Teaching, 55 (4), 533–546. Macaro, E. (2022) English Medium Instruction: What do we know so far, and what do we still need to find out? Language Teaching, 55(4), 533–546. ).

The creation of the European higher education area in 2010European Higher Education Area (2010, March). Ministerial declarations and communiqués: Budapest-Vienna declaration on the European higher education area. Retrieved September 9, 2024 from https://​ehea​.info​/page​-ministerial​-declarations​-and​-communiquesEuropean Higher Education Area (2010, March). Ministerial declarations and communiqués: Budapest-Vienna declaration on the European higher education area. Retrieved September 9, 2024 from https://​ehea​.info​/page​-ministerial​-declarations​-and​-communiques (EHEA, 2024) led to the massification, marketization and employability of universities (Alves & Tomlinson, 2020Alves, M. G., & Tomlinson, M. (2020) The changing value of higher education in England and Portugal: Massification, marketization and public good. European educational research journal, 20 (2), 176–192. Alves, M. G., & Tomlinson, M. (2020) The changing value of higher education in England and Portugal: Massification, marketization and public good. European educational research journal, 20(2), 176–192. ). Specifically, the Bologna Process, which began in 1999, sparked the shift of higher education in Europe from institutions of knowledge to economic enterprises (Štech, 2011Štech, S. (2011) The Bologna process as a new public management tool in higher education. Journal of Pedagogy, 2 (2), 263–282. Štech, S. (2011) The Bologna process as a new public management tool in higher education. Journal of Pedagogy, 2(2), 263–282. ). In order to secure their place in the global higher education market, Spain, along with many other European countries, began implementing EME to foster the internationalization and collaboration between countries, which has, in turn, led to a body of research questioning the equity, diversity and inclusion of the marketization of universities and the widespread implementation of EME and its consequences (see, for example, Bowles and Murphy, 2020Bowles, H., & Murphy, A. C. (Eds.) (2020) English-Medium Instruction and the Internationalization of Universities. London: Palgrave Macmillan. Bowles, H., & Murphy, A. C. (Eds.) (2020) English-Medium Instruction and the Internationalization of Universities. London: Palgrave Macmillan. , Jaeger, 2023Jæger, K. (2023) Elusive politics: De-internationalizing higher education in the context of international recognition conventions. Higher Education Policy, 36 , 411–429. Jæger, K. (2023) Elusive politics: De-internationalizing higher education in the context of international recognition conventions. Higher Education Policy, 36, 411–429. , and Wilkinson, 2013Wilkinson, R. (2013) English-Medium Instruction at a Dutch University: Challenges and Pitfalls. In A. Doiz, D. Lasagabaster, and J. M. Sierra, J. M. (Eds.), English-medium instruction at universities: Global challenges. Multilingual Matters.Wilkinson, R. (2013) English-Medium Instruction at a Dutch University: Challenges and Pitfalls. In A. Doiz, D. Lasagabaster, and J. M. Sierra, J. M. (Eds.), English-medium instruction at universities: Global challenges. Multilingual Matters.).

In China, the Chinese Ministry of Education is the entity responsible for the top-down EME policy implementation across the country which began around the year 2000. The main reason for its creation was the promotion of bilingual medium education to strengthen undergraduate teaching and to improve education quality (Ministry of Education, 2010Ministry of Education (2010) Guojia Zhongchangqi Jiaoyu Gaige He Fazhan Guihua Gangyao (2010–2020). [The Outline of Medium and Long-Term Education Reform and Development Planning (2010–2020)]. Beijing: Ministry of EducationMinistry of Education (2010) Guojia Zhongchangqi Jiaoyu Gaige He Fazhan Guihua Gangyao (2010–2020). [The Outline of Medium and Long-Term Education Reform and Development Planning (2010–2020)]. Beijing: Ministry of Education). The practice of EME in China has been steadily increasing ever since, and it has been noted that “English proficiency has become the most coveted form of cultural capital in Chinese society” (Hu & Lei, 2014Hu, G., & Lei, J. (2014) English-medium instruction in Chinese higher education: A case study. Higher Education, 67(5), 551–567. Hu, G., & Lei, J. (2014) English-medium instruction in Chinese higher education: A case study. Higher Education, 67(5), 551–567. , p. 564). While EME enables and contributes to the increased diversity, inclusion and equity of higher education in some respects it also has great potential to hinder them (DeCosta et al., 2021De Costa, P. I., Green-Eneix, C., & Li, W. (2021) Embracing diversity, inclusion, equity and access in EMI-TNHE: Towards a social justice-centered reframing of English language teaching. RELC Journal, 52 (2), 227–235. De Costa, P. I., Green-Eneix, C., & Li, W. (2021) Embracing diversity, inclusion, equity and access in EMI-TNHE: Towards a social justice-centered reframing of English language teaching. RELC Journal, 52(2), 227–235. ). More research is needed in this area to tease apart the affordances and drawbacks of this type of education and to uncover where diversity, inclusion and equity may be affected in EME. As Paulsrud et al., (2021)Paulsrud, B., Tian, Z., & Toth, J. (2021) Introduction. In B. Paulsrud, Z. Tian, & J. Toth (Eds.), English-medium instruction and translanguaging. Bristol: Multilingual Matters. Paulsrud, B., Tian, Z., & Toth, J. (2021) Introduction. In B. Paulsrud, Z. Tian, & J. Toth (Eds.), English-medium instruction and translanguaging. Bristol: Multilingual Matters. mention, there is a need to understand the unique needs of the educational context, a need to “understand the affordances available for learning, communicating and building identity” (p. 15) in each context, and to a need to understand the opportunities and practices of each context. Thus, the present study reflects on two different variables in English-medium higher education: context and multilingual identity, and asks if perceptions, multilingual beliefs, and motivations toward EME differ depending on the context and/or on one’s multilingual identity. After reporting the findings, a thorough discussion is provided shedding light on the specific ways in which diversity, inclusion and equity might be enabled or hindered in the two contexts under study, and finally actionable strategies are suggested to address the challenges with an aim to improve the EME experience for the different stakeholders.

2.Theoretical framework

Research on EME contexts has been steadily increasing to keep up with its ever-growing implementation around the world, and while tendencies and trends have been noted within specific contexts, it remains unknown how applicable the findings of each study are to new contexts. This is largely due to the multitude of factors involved in the practice. For instance, EME cannot be studied in isolation and is recognized as interacting with social behaviors, groupings, and beliefs that interact with language policy. In an effort to encompass the wide range of factors involved in EME education, the ROADMAPPING framework (Dafouz & Smit, 2020Dafouz, E., & Smit, U. (2020) ROAD-MAPPING English medium education in the internationalised university. London: Palgrave Macmillan. Dafouz, E., & Smit, U. (2020) ROAD-MAPPING English medium education in the internationalised university. London: Palgrave Macmillan. ) was proposed as a common framework for interpretation and analysis. It takes a holistic approach, theoretically grounded in sociolinguistics, ecolinguistics, and language policy research. This is the perspective from which the present study is anchored. The ROADMAPPING model offers a template for analysis allowing for wider generalizations and applicability to diverse contexts. The framework includes six interconnected dimensions: roles of English (Ro), academic disciplines (AD), (language) management (M), agents (A), practices and purposes (PP), and internationalization and glocalization (ING).

The present study is particularly interested in the (1) roles, which refers to the “communicative functions that language fulfills in the higher education institution” (Dafouz & Smit, 2020Dafouz, E., & Smit, U. (2020) ROAD-MAPPING English medium education in the internationalised university. London: Palgrave Macmillan. Dafouz, E., & Smit, U. (2020) ROAD-MAPPING English medium education in the internationalised university. London: Palgrave Macmillan. , p. 60), (2) practices and purposes which refers to the classroom practices or professional development practices that involve EME, and (3) agents, which refers to the social players that are engaged with EME at sociopolitical, institutional, or hierarchical levels. Applied to the present study, agents are the roles or identities students adopt and project in EME. These three dimensions were chosen because they are thought to be closely intertwined. For example, agency refers to the institutional policies and individual perspectives of the stakeholders. This dimension influences the purposes and practices of English in the learning context which, in turn, influences the role of English in the context. The study targets these dimensions by collecting quantitative data through questionnaires on our participants’ perspectives, multilingual beliefs, and motivations in the two EME settings. Being a relatively new framework for analysis, research applying the ROADMAPPING framework is still scarce. However, one study revealed that learners in different contexts view the role of English as distinct, specifically, Thai students perceived English as a target to be learned whereas Austrian students perceived English as a tool to learn content and that language goals were not to be expected from EME (Baker & Hüttner, 2017Baker, W., & Hüttner, J. (2017) English and more: a multisite study of roles and conceptualizations of language in English medium multilingual universities from Europe to Asia. Journal of Multilingual and Multicultural Development, 30 (6), 501–516. Baker, W., & Hüttner, J. (2017) English and more: a multisite study of roles and conceptualizations of language in English medium multilingual universities from Europe to Asia. Journal of Multilingual and Multicultural Development, 30(6), 501–516. ). This demonstrates the different perspectives learners in different contexts can have towards EME and how it might affect their practices and purposes in the classroom. Studies on these dimensions in different Asian contexts reveal that English holds the role of lingua franca and can be understood as part of multilingual and translanguaging processes as there is a clear role for other language practices (Baker & Tsou, 2021Baker, W., & Tsou, W. (2021) EMI and translanguaging in Asia through the ROAD-MAPPING lens. In W. Baker, & W. Tsou, (Eds.), English-medium instruction translanguaging practices in Asia: Theories, frameworks and implementation in higher education. (pp. 183–195). Springer. Baker, W., & Tsou, W. (2021) EMI and translanguaging in Asia through the ROAD-MAPPING lens. In W. Baker, & W. Tsou, (Eds.), English-medium instruction translanguaging practices in Asia: Theories, frameworks and implementation in higher education. (pp. 183–195). Springer. ). Regarding agents Baker and Tsou (2021)Baker, W., & Tsou, W. (2021) EMI and translanguaging in Asia through the ROAD-MAPPING lens. In W. Baker, & W. Tsou, (Eds.), English-medium instruction translanguaging practices in Asia: Theories, frameworks and implementation in higher education. (pp. 183–195). Springer. Baker, W., & Tsou, W. (2021) EMI and translanguaging in Asia through the ROAD-MAPPING lens. In W. Baker, & W. Tsou, (Eds.), English-medium instruction translanguaging practices in Asia: Theories, frameworks and implementation in higher education. (pp. 183–195). Springer. found that when EMI students are encouraged to use their full language repertoires, they can better develop their agency and multilingual self in the EME context. Beyond these studies little is known about these dimensions across EME contexts, which is one of the unique contributions of the present study.

3.Methodology

4.Results

The data was first analyzed to target RQ1, for which it was split into two groups: Spain or China. The descriptive data for each variable is presented in Table 3. For this data a high mean (closer to 5) shows agreement or alignment with the variable in question and a low mean (closer to 1) shows disagreement or disalignment with the variable in question.

Then, to detect any differences in any of the variables between the two EME contexts, an independent sample t-test was carried out. Firstly, the data was checked for normality and homogeneity, and both were found to be normal for all variables except for perceived difficulty and multilingual identity. Thus, for these two variables the Mann-Whitney U value is reported in place of the student’s t value. The results are summarized in Table 4.

Significant differences were found within each construct investigated. Specifically, regarding perceptions, perceived level and perceived difficulty were significant, showing that the Chinese students felt they had a lower level of English and found EME more difficult than the Spanish students did. Two of the three multilingual belief variables were significant: OMS and FMS. The mean scores show that the Spanish EME students were more open-minded and had a stronger future multilingual self than the Chinese EME students. Then, with respect to motivations, results show that the Spanish students were more intrinsically and instrumentally motivated than their Chinese counterparts and that the Chinese students felt more anxious in their EME classes than the Spanish ones. Finally, significantly more Spanish participants self-identified as being multilingual, whereas most Chinese participants did not feel multilingual or were unsure if they were multilingual. No other significant differences between the groups were detected.

Turning to the differences between groups depending on multilingual identity, the target of RQ2, we found that of the 48 Spanish participants, 41 self-identified as multilingual speakers, mentioning that they felt this because they “can speak three or four languages fluently’, that they ‘can express themselves in three languages”, and that they “can switch between 3 languages with ease”. Four of the Spanish participants said they were not multilingual stating that they “only speak three languages”, that they “do not practice the languages they know”, or that they “are not completely fluent in the languages they know”. The remaining three participants said that they were unsure because they were “not sure how fluent in each domain they must be to be considered multilingual” or that while they “speak different languages the only one they feel comfortable in is their native language”. As for the Chinese participants, nine of them self-identified as being multilingual speakers mainly because they “know and use three languages”. Twenty-eight of them felt they were not multilingual because they “do not speak more than two languages”, they “never use the second or third language in daily life”, and they “do not have any opportunities to use the foreign languages that they know”, or that they “have not mastered the language yet”. Finally, 22 of them reported not knowing if they were multilingual or not because even though they “have studied and know other languages they have not mastered them”, or they felt they “do not use the foreign languages enough”. The descriptive statistics of the data when grouped by multilingual identity are presented in Table 5.

To determine if there were any significant differences between the three groups, those who did identify as multilingual (Yes), those who did not (No) and those who were not sure (Unsure), a one-way ANOVA was carried out. When the data was tested for normality and homogeneity all variables were considered normal except integrative motivation and perceived improvement. For this reason, we report the Welch’s F for these two variables and the Fisher’s F for the other variables. Also, for the normally distributed data the Tukey post-hoc test is reported and for the abnormally distributed data the Games-Howell test is reported. A summary of the findings is reported in Table 6.

The ANOVA results show significant main effects for the variables of perceived difficulty, perceived level, OMS, intrinsic motivation, and anxiety. The post-hoc tests reveal that regarding perceived difficulty, those who perceived themselves as multilingual felt that EME was less difficult than both those who did not feel multilingual (p = < .001) and those who were unsure of their multilingual status (p = < .009). Regarding perceived level, the Yes group was again significantly different from the other two groups, No and Unsure, (p = < .001) and (p = < .001) respectively, revealing that those who felt multilingual also felt they had a higher level of English than those who did not feel multilingual or who were unsure. With respect to the OMS variable, we found that there was a significant difference only between the Yes and No group (p = .018) showing that those who identified as being multilingual were more open-minded than those who did not feel multilingual, but not more so than those who were unsure of their multilingual abilities. As for motivations, results show that the self-identifying multilingual group was more intrinsically motivated than the group who did not feel multilingual (p = < .001) and the group who was unsure (p = .037). Finally, regarding anxiety we found a significant difference between both the Yes group and the No group (p = .002) and the Yes group and the Unsure group (p = .001) revealing that those who identified as being multilingual were less anxious in their EME classes than those who did not or who were unsure. No significant differences were detected between the No and Unsure groups for any variables.

5.Discussion

This study has undertaken an analysis of the effects of context and multilingual identity on perceptions, motivations, and multilingual beliefs. Two EME contexts were examined, an education degree program in China and Spain, with an aim to offer some insights to the dimensions of roles, practices and purposes, and agents in each of these contexts with respect to English under the ROADMAPPING framework.

We first consider the results related to the RQ1 which examined how context affects multilingual identity, perceptions, multilingual beliefs and, motivations. Findings showed that there were significant group differences for all three perception variables: perceived difficulty, perceived level, and perceived improvement. This demonstrates how closely connected these variables are to each other and to language proficiency level. It also demonstrates that the Chinese participants felt they had a lower or insufficient English level for the EME courses and therefore also found the experience of taking EME courses more challenging.

A second finding in relation to RQ1 is that the Spanish group perceived significantly greater improvements than the Chinese group, which had less EME exposure and lower English levels. This contradicts previous research, for example, Ament and Pérez-Vidal (2015)Ament, J. & Pérez-Vidal, C. (2015) Linguistic outcomes of English medium instruction programmes in higher education: A study on economics undergraduates at a Catalan University. Higher Learning Research Communications, 5 (1), 47–68. Ament, J. & Pérez-Vidal, C. (2015) Linguistic outcomes of English medium instruction programmes in higher education: A study on economics undergraduates at a Catalan University. Higher Learning Research Communications, 5(1), 47–68. , who found that when comparing language outcomes in full- and semi-EME programs only the semi-EME students experienced significant gains. In addition to this, lower-level groups are usually reported to make the most significant gains with respect to language, which is usually attributed to having a wider margin for improvement while more advanced levels plateau (Coyle et al., 2010Coyle, D., Hood, P., & Marsh, D. (2010) CLIL: Content and language integrated learning. Cambridge University Press. Coyle, D., Hood, P., & Marsh, D. (2010) CLIL: Content and language integrated learning. Cambridge University Press. ; Wilkinson, 2004Wilkinson, R. (Ed.) (2004) Integrating content and language: meeting the challenge of a multilingual higher education; proceedings of the ICL conference, October 23–25 2003. Universitaire Pers Maastricht.Wilkinson, R. (Ed.) (2004) Integrating content and language: meeting the challenge of a multilingual higher education; proceedings of the ICL conference, October 23–25 2003. Universitaire Pers Maastricht.). One possible explanation for the finding could be that, as some participants mentioned in the open-ended questions, the Chinese participants had fewer opportunities to practice English both in and out of the EME classes and that, perhaps, more effective strategies to encourage interaction and language use in the EME class is needed to enhance language skills (Aizawa et al., 2023Aizawa, I., Rose, H., Thompson, G., & Curle, S. (2023) Beyond the threshold: Exploring English language proficiency, linguistic challenges, and academic language skills of Japanese students in an English medium instruction programme. Language Teaching Research, 27 (4), 837–861. Aizawa, I., Rose, H., Thompson, G., & Curle, S. (2023) Beyond the threshold: Exploring English language proficiency, linguistic challenges, and academic language skills of Japanese students in an English medium instruction programme. Language Teaching Research, 27(4), 837–861. ). This finding, could also be particular to the Chinese context, as it is in line with another Chinese EME study which detected that students’ self-perceived language gains became worse after a one-semester of EME courses. The authors noted, through classroom observation, that the EME teachers had a teacher-centered approach which allowed little time for students to interact and they did little to scaffold the students’ language learning (Zhang & Pladevall-Ballester, 2022Zhang, M., & Pladevall-Ballester, E. (2022) Students’ attitudes and perceptions towards three EMI courses in mainland China. Language, Culture and Curriculum, 35 (2), 200–216. Zhang, M., & Pladevall-Ballester, E. (2022) Students’ attitudes and perceptions towards three EMI courses in mainland China. Language, Culture and Curriculum, 35(2), 200–216. ).

These findings interpreted under the ROADMAPPING framework reveal that the role of English for the Chinese participants on a local level is more of a target than a tool. They also suggest that the Chinese participants have poorer agency in the language and hesitate to interact with it in the EME context. Finally, regarding practices and purposes, EME is difficult to engage with, it is something to be learned and it causes anxiety in the majority of Chinese participants. The opposite is true of the Spanish participants: they seem to use English as a tool to communicate, have agency and participate in the EME classroom. They also seem to have more use for English both in and out of the classroom.

A similar interpretation can be applied to the findings on the effect of context on motivations. Specifically, the Chinese participants, possibly due to the increased perceived difficulty and lower English level, were less intrinsically motivated and experienced higher anxiety in the EME class compared to their Spanish counterparts. This finding is consistent with other studies that have demonstrated a negative correlation between intrinsic motivation and anxiety (Ament et al., 2020Ament, J., Barón-Parés, J., & Pérez-Vidal, C. (2020) Exploring the relationship between motivations, emotions and pragmatic marker use in English-medium instruction learners. Journal of Language Learning in Higher Education, 10 (2), 469–489. Ament, J., Barón-Parés, J., & Pérez-Vidal, C. (2020) Exploring the relationship between motivations, emotions and pragmatic marker use in English-medium instruction learners. Journal of Language Learning in Higher Education, 10(2), 469–489. ). This finding gives more insight to the agency dimension when it comes to English use; the Spanish students tend to have more personal agency reflected through their inner motivation to engage with EME. Another difference noted is that the Spanish students were significantly more instrumentally motivated that the Chinese ones. This finding seems to suggest that the Chinese students do not see an immediate or real need for English in their immediate or near future lives compared to the Spanish students. This reflects that at the local level the role and practices and purposes dimensions are significantly different in the two contexts. For example, the Spanish students have more contact with international students due to being part of the European Union’s Erasmus exchange program. In addition, the Spanish study takes place in Barcelona, not only a popular tourist destination but also a city where a high percentage of foreigners choose to live, making English a common language at some levels of society and daily life. The Chinese participants, however, have little contact with foreigners in daily life and few opportunities to use English outside classrooms.

Motivation results regarding RQ1 also showed that there were important similarities between the two groups regarding extrinsic motivation. This could be because, as Lasagabaster (2016)Lasagabaster, D. (2016) The relationship between motivation, gender, L1 and possible selves in English-medium instruction. International Journal of Multilingualism, 13 (3), 315–332. Lasagabaster, D. (2016) The relationship between motivation, gender, L1 and possible selves in English-medium instruction. International Journal of Multilingualism, 13(3), 315–332. found, EME students tend to have strong family and other societal pressures that drive motivation. Another similarity between the groups worth noting is integrative motivation levels, which suggests that both groups are similarly interested in English language and culture. These two findings demonstrate the similarities of the role and practices and purposes dimensions at the international level i.e., the status of English as a global language widely recognized for its usefulness in business, academia and leisure. As well as the access to valuable cultural capital that speaking English grants (Hu & Lei 2014Hu, G., & Lei, J. (2014) English-medium instruction in Chinese higher education: A case study. Higher Education, 67(5), 551–567. Hu, G., & Lei, J. (2014) English-medium instruction in Chinese higher education: A case study. Higher Education, 67(5), 551–567. ; Machin et al., 2023Machin, E., Ament, J., & Pérez-Vidal, C. (2023) Motivation and investment: Exploring the choice of English-medium instruction for mid-degree undergraduates in Catalonia. Journal of English-Medium Instruction, 2 (1), 79–100. Machin, E., Ament, J., & Pérez-Vidal, C. (2023) Motivation and investment: Exploring the choice of English-medium instruction for mid-degree undergraduates in Catalonia. Journal of English-Medium Instruction, 2(1), 79–100. ).

Results from the multilingual beliefs variables showed that the groups were different according to FMS, which seems to parallel the motivation variables findings. It seems that the Chinese students have difficulty visualizing the usefulness (role and practices and purposes) of their EME studies for their future (integral motivation) and that they also have a less-well developed FMS (agency) compared to the Spanish students. No significant differences were found for the BAM variable, indicating that, at the individual level, the inner beliefs of the EME students in Spain and China do not differ when it comes to the positive views they hold towards multilingualism. What may be surprising is that, despite this similarity in multilingual beliefs, there is a significant difference in OMS. Culturally, the Chinese students may have a stronger sense of ethnic protection as was found in Chen et al.’s study (2016Chen, S. X., Lam, B. C. P., Hui, B. P. H., Ng, J. C. K., Mak, W. W. S., Guan, Y., Buchtel, E. E., Tang, W. C. S., & Lau, V. C. Y. (2016) Conceptualizing psychological processes in response to globalization: Components, antecedents, and consequences of global orientations. Journal of Personality and Social Psychology, 110 (2), 302–331. Chen, S. X., Lam, B. C. P., Hui, B. P. H., Ng, J. C. K., Mak, W. W. S., Guan, Y., Buchtel, E. E., Tang, W. C. S., & Lau, V. C. Y. (2016) Conceptualizing psychological processes in response to globalization: Components, antecedents, and consequences of global orientations. Journal of Personality and Social Psychology, 110(2), 302–331. ), which was found to correlate with less open-mindedness. It may also be the case that the lower proficiency levels and lack of experience abroad and with foreigners cause the Chinese students to be less open to different or conflicting ideas, opinions or experiences. Finally, less Chinese students identified themselves as multilingual speakers compared to the Spanish ones. This finding is in line with the trends noted regarding perspectives, motivations and multilingual beliefs of the Chinese group. It may also be that the Spanish students have more fully developed multilingual selves and therefore also feel a stronger agency when interacting in EME courses. An intriguing finding is that most of the Chinese participants were unsure of their multilingual status, which could indicate that their multilingual identity is being constructed but that they had not fully claimed their agency in English at the time of data collection.

RQ2 asked if multilingual identity had any effect on the perspectives towards EME, motivations or multilingual beliefs. The findings showed that participants who identified as being multilingual had a higher perceived level of English; they felt they had a sufficient English level for the course. While those who did not identify as multilingual felt the opposite and, in turn, felt that the EME courses were more difficult. The findings with respect to motivations show a similar pattern, namely that those who feel they are multilingual feel less anxious in the EME classes and have higher intrinsic motivation than those who do not or are unsure. This suggests that those who identify as multilingual may feel they are users of the language and thus no longer feel the “social relations of power […] that constrain opportunities for language learners to speak” (Darvin & Norton, 2021Darvin, R. & Norton, B. (2021) Identity. In T. Gregerson & S. Mercer (Eds.), Handbook of the Psychology of Language Learning and Teaching. Routledge. Darvin, R. & Norton, B. (2021) Identity. In T. Gregerson & S. Mercer (Eds.), Handbook of the Psychology of Language Learning and Teaching. Routledge. , p. 91). Finally, those who feel multilingual were found to be more open-minded than those who are not. This parallels Sung’s (2016)Sung, C. C. M. (2016) ESL university students’ perceptions of their global identities in English as a lingua franca communication: A case study at an international university in Hong Kong. The Asia-Pacific Education Researcher, 25 , 305–314. Sung, C. C. M. (2016) ESL university students’ perceptions of their global identities in English as a lingua franca communication: A case study at an international university in Hong Kong. The Asia-Pacific Education Researcher, 25, 305–314. and Chen et al.’s (2016)Chen, S. X., Lam, B. C. P., Hui, B. P. H., Ng, J. C. K., Mak, W. W. S., Guan, Y., Buchtel, E. E., Tang, W. C. S., & Lau, V. C. Y. (2016) Conceptualizing psychological processes in response to globalization: Components, antecedents, and consequences of global orientations. Journal of Personality and Social Psychology, 110 (2), 302–331. Chen, S. X., Lam, B. C. P., Hui, B. P. H., Ng, J. C. K., Mak, W. W. S., Guan, Y., Buchtel, E. E., Tang, W. C. S., & Lau, V. C. Y. (2016) Conceptualizing psychological processes in response to globalization: Components, antecedents, and consequences of global orientations. Journal of Personality and Social Psychology, 110(2), 302–331. findings, who found that those who had strong global identities were more tolerant and open to new experiences. In sum, when learners’ identify as multilingual speakers they perceive having a higher language level and experience less difficulty and anxiety in their EME classes. They also experience higher intrinsic motivation towards EME (enjoyment and fulfillment) and open-mindedness is supported. This suggests that the multilingual context created by EME might encourage speaker agency and open-mindedness.

Finally, applying the ROADMAPPING framework to the findings we can see that the perceptions, motivations and multilingual beliefs of the EME students significantly impact the roles, practices and purposes, and agency in the EME context. Specifically, the study demonstrates that when students perceive themselves to have lower levels of English they tend to identify more as learners of the language rather than users of the language. This is evident through the findings on motivations. As learners, rather than users of the language, the students have higher levels of anxiety and low intrinsic motivation which affects the practices and purposes of the students in the classrooms. For example, research shows that anxious students participate less in the classroom which can influence learning outcomes as Yuksel et al., (2023)Yuksel, D., Soruç, A., Horzum, B., & McKinley, J. (2023) Examining the role of English language proficiency, language learning anxiety, and self-regulation skills in EMI students’ academic success. Studies in Second Language Learning and Teaching, 13 (2), 399–426. Yuksel, D., Soruç, A., Horzum, B., & McKinley, J. (2023) Examining the role of English language proficiency, language learning anxiety, and self-regulation skills in EMI students’ academic success. Studies in Second Language Learning and Teaching, 13(2), 399–426. reported. Regarding purpose, when students feel they have a low level they tend to use English as a target to be learned rather than as a tool to communicate as was reported in Baker and Hüttner (2017)Baker, W., & Hüttner, J. (2017) English and more: a multisite study of roles and conceptualizations of language in English medium multilingual universities from Europe to Asia. Journal of Multilingual and Multicultural Development, 30 (6), 501–516. Baker, W., & Hüttner, J. (2017) English and more: a multisite study of roles and conceptualizations of language in English medium multilingual universities from Europe to Asia. Journal of Multilingual and Multicultural Development, 30(6), 501–516. . Finally, this study has also brought to light that if students do not feel capable and confident in their educational setting it can affect their agency. The students do not feel empowered to participate and fully take on a multilingual identity, something that our findings suggest may negatively affect their educational outcomes, their future opportunities (FMS) as well as their open-mindedness.

6.Conclusion

The present study has measured how either context or multilingual identity impacts three complex individual difference variables in two EME contexts. There are two main findings to be noted from the study, one in relation to context (RQ1), and the other in relation to multilingual identity (RQ2). Firstly, concerning RQ1, we did find significant context effects. Specifically, at the individual level EME students in the Chinese context seem to suffer from poor self-perceptions and in turn have little agency when they speak English. They tend to perceive themselves as learners of the language rather than users of it. The Spanish students have an increased tendency to perceive themselves as capable in the EME context and, as such, seem to have claimed their right as speakers of the language. At the local level the context plays an important role with respect to the practices and purposes dimension. This was evident through the attitudes, motivations, and FMS variables. Finally, at the international level there were no significant differences regarding the perspectives towards English and extrinsic and integrative motivation of the two contexts.

Concerning RQ2, we did find significant effects for self reported multilingualism. Specifically, those who identify as multilinguals also have increased agency in the EME context which, in turn, fosters OMS, FMS, enjoyment and engagement in the EME context. What is clear from the findings is that the individual difference variables are independent of the contextual ones. This suggests that perceptions, attitudes, motivation and identity have the potential to affect students regardless of their context.

A limitation of this study may be related to the differences in the two groups with regards to culture which may explain some of the results but is also difficult to control for. In addition to culture, other factors that we were unable to control for were educational system of the country and pedagogical style of the individual teachers. However, the bringing together and comparison of the two cultures, pedagogical styles, and academic systems under the ROADMAPPING framework is also one of the strengths of the study. Another limitation is that the Chinese students began the study with an intermediate English level while most of the Spanish students had an advanced English level, which may be a confounding variable. In fact, the findings suggest that perceived proficiency level is a key factor in perspectives and experiences in EME. We encourage more studies in this area to help tease apart the relationship between language level and perspectives. Another weakness could be that, even though exposure to EME was similar, it was not exactly the same between the groups, and in the future controlling for this variable would help strengthen the results. A contribution of the present study is that it examines three complex individual difference variables and by applying the ROADMAPPING framework it describes how the variables impact roles, practices and purposes as well as agency in two EME contexts. Despite any possible shortcomings, we believe that the study offers valuable insights into the interactions between the sets of variables investigated and how they play out in two different contexts.

Thus, interpreting the findings together, a number of actionable strategies for educators, students and EME program design can be suggested in order to improve experiences and outcomes of EME. Firstly, for EME classrooms to become more diverse, inclusive and equitable, both learners and teachers should be more aware of the specific challenges they face in these contexts (e.g. level, exposure, need, motivation) and strive to find methods of overcoming them. Some ways to achieve this would be to give students surveys or questionnaires to identify how they feel, have group or class discussions identifying challenges and what the possible solutions might be at the local level. It is through becoming aware that the students can then empower themselves with respect to their English language use.

Secondly, a strategy that could benefit lower proficiency EME students would be the integration of translanguaging and multilingual practices. Research shows that when learners are able to translanguage, (use their full-multilingual repertoire) they are empowered and engage more deeply with their studies (Baker & Tsou, 2021Baker, W., & Tsou, W. (2021) EMI and translanguaging in Asia through the ROAD-MAPPING lens. In W. Baker, & W. Tsou, (Eds.), English-medium instruction translanguaging practices in Asia: Theories, frameworks and implementation in higher education. (pp. 183–195). Springer. Baker, W., & Tsou, W. (2021) EMI and translanguaging in Asia through the ROAD-MAPPING lens. In W. Baker, & W. Tsou, (Eds.), English-medium instruction translanguaging practices in Asia: Theories, frameworks and implementation in higher education. (pp. 183–195). Springer. ). EME instructors and students may benefit from specific training on the incorporation of targeted translanguaging educational practices.

Thirdly, we recommend that, rather than implementing entry level tests or minimum proficiency levels in English — which can act as gatekeepers to EME and hinder inclusion and equity — EME courses or programs be taught through a combination of instruction from language and content specialists (Doiz, et al., 2019Doiz, A., Lasagabaster, D., & Pavón, V. (2019) The integration of language and content in English-medium instruction courses: Lecturers’ beliefs and practices. Ibérica, 38 (38), 151–176.Doiz, A., Lasagabaster, D., & Pavón, V. (2019) The integration of language and content in English-medium instruction courses: Lecturers’ beliefs and practices. Ibérica, 38(38), 151–176.; Kletzenbauer, et al. 2022Kletzenbauer, P., Fürstenberg, U., & Reitbauer, M. (2022) Becoming “language-aware” in ICLHE: Creating a space of trust for teachers, teachers trainers and language specialists. CLIL Journal of Innovation and Research in Plurilingual and Pluricultural Education, 5 (2), 53–63. Kletzenbauer, P., Fürstenberg, U., & Reitbauer, M. (2022) Becoming “language-aware” in ICLHE: Creating a space of trust for teachers, teachers trainers and language specialists. CLIL Journal of Innovation and Research in Plurilingual and Pluricultural Education, 5(2), 53–63. ). Adopting this pedagogical approach can scaffold lower-level students and serve to boost inclusivity and diversity in the classroom. The findings from our study point out that if students feel confident and capable in their EME courses it could help shift perspectives and improve outcomes.

Finally, due to the demonstrated contextual differences of the dimensions of role, practices and purposes and agency in the present study it is essential that a needs analysis be carried out before implementing EME to decide how best to address both the individual, local and international needs of the context. EME cannot be a one-size-fits all application and must be implemented taking into consideration the unique context of each university. Perhaps through the adoption of these measures, EME could foster more diversity, equity and inclusion. We encourage more studies that bring together multiple contexts or multiple individual difference variables to gain more insights into the complexity of EME and to corroborate the trends noted in this study.